8 Jun
2013
8 Jun
'13
1:38 a.m.
Hi, All,
Going through stuff on the shelf and I've run across a couple of
486-based fully-integrated LCD/touchscreen machines. One is a SCAN
Corporation SCANtouch Model 3000, the other is a Planar Systems box.
Similarities include limited memory (2x 72-pin SIMM sockets on the
SCANtouch or 4x 30-pin SIMM sockets on the Planar, both with
"double-sided SIMM" support, so 64MB or 32MB max respectively), one
ISA socket (no PCI), onboard NE2000, serial, parallel, 2.5" PATA
disk... these systems will run MS-DOS, of course, and Win95 (the touch
screen on the SCANtouch is old enough that it apparently doesn't work
with Win98 mouse drivers, according to forum posts I dug up). I
really don't care about running DOS or Win95 on a touchscreen/LCD
machine, and the practical alternative is some flavor of Linux or
UNIX. What I'm having difficulty digging up is when the break from
low-mem/pre-Pentium systems happened and what distros are on which
side of the divide. RedHat 7/8/9 require a Pentium and 64MB minimum
from what I can research/remember. The last time I ran
Linux on a
486, it was Yggdrasil (and before that, some early form of Slackware
on a 386).
So... anyone have a "go-to" Linux distro for sticking on a 486? I
know RedHat 5.2 will work - that's what's on my PS/2-E (w/486SLC and
12MB of free mem I was recently discussing). Any other choices? At
one time, when a Dell P-133 Latitude laptop was my main machine, I ran
Solaris 7 on it (because it was a better choice than Linux at the
time) even though it still had some issues (and workarounds) with the
NeoMagic video chipset and the 3Com 3C589 PCMCIA NIC. I suspect this
is likely to have similar "challenges". I've dug up full specs on the
Scantouch 3000 innards - PCM-4890 integrated CPU board, NE2000
(Realtek RTL8019) Ethernet, C&T 65545 video chipset w/800x600 max LCD
resolution, Sharp LM10V33 VGA (640x480) 10.4" color LCD, VIA VT82C496G
chipset, , PC104 sockets... so I have little concern about getting
*something* working with it.
Getting 10-year-old RedHat working on a Pentium-class machine isn't a
real challenge, but it's been long enough since I've really fiddled
with 486s that specific memories of system configuration are getting a
bit fuzzy. At one time, over 15 years ago, it was a daily thing
knowing the ins and outs of what the 486 could and couldn't do, before
CPUs and memory and clock speeds took off like a rocket, and when 4MB
was ordinary, but more than 16MB was uncomfortably expensive for hobby
gear.
Helpful suggestions wanted.
Thanks,
-ethan
8 Jun
8 Jun
7:19 a.m.
On Sat, 8 Jun 2013, Ethan Dicks wrote:
Going through stuff on the shelf and I've run
across a couple of
486-based fully-integrated LCD/touchscreen machines. One is a SCAN
Corporation SCANtouch Model 3000, the other is a Planar Systems box.
Similarities include limited memory (2x 72-pin SIMM sockets on the
SCANtouch or 4x 30-pin SIMM sockets on the Planar, both with
"double-sided SIMM" support, so 64MB or 32MB max respectively), one ISA
socket (no PCI), onboard NE2000, serial, parallel, 2.5" PATA disk...
these systems will run MS-DOS, of course, and Win95 (the touch screen on
the SCANtouch is old enough that it apparently doesn't work with Win98
mouse drivers, according to forum posts I dug up). I really don't care
about running DOS or Win95 on a touchscreen/LCD machine, and the
practical alternative is some flavor of Linux or UNIX. What I'm having
difficulty digging up is when the break from low-mem/pre-Pentium systems
happened and what distros are on which side of the divide. RedHat 7/8/9
require a Pentium and 64MB minimum from what I can research/remember.
The last time I ran Linux on a 486, it was Yggdrasil (and before that,
some early form of Slackware on a 386).
So... anyone have a "go-to" Linux distro for
sticking on a 486? I know
RedHat 5.2 will work - that's what's on my PS/2-E (w/486SLC and 12MB of
free mem I was recently discussing). Any other choices?
This was part of my gripe in the other thread. We just don't have a full
fledged or embedded Linux solution that is currently maintained that will
work well with these type of embedded boards.
RedHat 5.2 was the last RedHat Linux distribution I used on 486 based
boards and IIRC, 5.2 was the last of the pre-GNOME RedHat releases (this
was around the time the Pentium II came out and the ATX form factor came
to market). I later moved all the 486 based stuff I maintained that used
Linux to Debian 2.1 (slink) and 2.2 (potato). I would think that most any
Debian version from 3.0 (woody) up through 5.0 (lenny) would work
moderately well with a 486 based application, as long as you aren't trying
to use something like GNOME or KDE under X11.
A current Debian distribution such as 6.0 (squeeze) or 7.0 (wheezy) should
work on a 486, at least in console mode (I've not even considered trying
to run current X11 on one) but you'll need as much memory as you can
install plus lots of hard drive and a swap partition. Installation would
also likely be very slow, and you'd want to deselect everything in
'tasksel' during installation (unless you have at least 2GB of hard drive
space, but even then it will install a lot of useless stuff) and manually
install whatever packages you need after letting it install the base OS.
Your board with 4 30-pin SIMM sockets likely tops out at 16MB. AFAIK, no
one ever made 8MB 30-pin SIMMs, although at one time there were some 16MB
SIMMs available at great expense (~$400.00/ea or ~$25/megabyte). 99% of
the boards out there did not support 30-pin SIMMs larger than 4MB though.
I do have a few EISA boards that could (8 sockets, 128MB max), but I never
was able to put my hands on any 16MB 60ns 30-pin SIMMs in matched sets of
4.
I've dug up full specs on the Scantouch 3000
innards - PCM-4890
integrated CPU board, NE2000 (Realtek RTL8019) Ethernet, C&T 65545 video
chipset w/800x600 max LCD resolution, Sharp LM10V33 VGA (640x480) 10.4"
color LCD, VIA VT82C496G chipset, , PC104 sockets... so I have little
concern about getting *something* working with it.
The PCM-4890 was made by Advantech and is a pretty good board. I'm
familiar with them and at one time used lots of them. It is a 5.25" sized
"Biscuit PC" form factor. Mostly due to the lack of a decent Linux
distribution to use with them, I swapped out all the remaining 5.25"
Biscuit PC 486 boards I was maintaining with Socket 7 boards about 8-10
years ago. A quick Google search even turns up a pdf file of the user
guide: ftp://ftp.emacinc.com/Archive/m-4890.pdf
I would suggest avoiding "EDO" memory with these and stick with a pair of
32MB 72-pin FPM Parity type SIMMs. EDO support with many of these types of
boards is problematic and was more of an afterthought and marketing ploy.
Samsung FPM SIMMs in particular work very, very well with these types of
boards.
Getting 10-year-old RedHat working on a Pentium-class
machine isn't a
real challenge, but it's been long enough since I've really fiddled with
486s that specific memories of system configuration are getting a bit
fuzzy. At one time, over 15 years ago, it was a daily thing knowing the
ins and outs of what the 486 could and couldn't do, before CPUs and
memory and clock speeds took off like a rocket, and when 4MB was
ordinary, but more than 16MB was uncomfortably expensive for hobby gear.
We just don't have a currently maintained Linux distribution out there
that is intended for these types of applications. I would personally love
to have a distribution that was tailored for low memory embedded-server
applications, but something that could also (optionally) support a GUI
would be very useful too. I still maintain lots of Socket 7 and Socket 370
stuff, and despite the "Just go buy a new PC" responses from many of the
current Linux distribution maintainers, it is cases such as this where you
can't just replace something with a cheap off-the-shelf consumer "PC".
9:04 a.m.
On Jun 8, 2013, at 7:19, Tothwolf <tothwolf at concentric.net> wrote:
Your board with 4 30-pin SIMM sockets likely tops out
at 16MB. AFAIK, no one ever made 8MB 30-pin SIMMs, although at one time there were some
16MB SIMMs available at great expense (~$400.00/ea or ~$25/megabyte).
They're not as expensive now; I just recently put four of them into
my Quadra 700 (which definitely supports them) for about $40.
99% of the boards out there did not support 30-pin
SIMMs larger than 4MB though.
That's the rub. Inasmuch as the 30-pin sense lines were ever "standard",
I don't think there was ever a standard sense configuration for 16MB,
which means that the chipset would have to support the rows/cols
configuration (many didn't) and the BIOS would need to manually
probe the bounds of the memory array. Many Macs supported 16MB 30-
pin SIMMs, but my recollection is that more than a few didn't.
I do have a few EISA boards that could (8 sockets,
128MB max), but I never was able to put my hands on any 16MB 60ns 30-pin SIMMs in matched
sets of 4.
I don't recall if they're 60ns or not, but they're pretty cheap on eBay now.
You can often find them (matched sets, that is) marketed specifically
for Quadra 700s. :-)
- Dave
11:26 a.m.
On Sat, 8 Jun 2013, David Riley wrote:
On Jun 8, 2013, at 7:19, Tothwolf <tothwolf at
concentric.net> wrote:
I'll keep that in mind for later. I still don't see spending $80 for 128MB
of ram for a 486 board though, let alone two of them, even if they are
EISA. IIRC, both boards had developed CPU clock problems too, so I should
probably troubleshoot that first anyway.
OTOH, I did pick up 4x 128MB 72-pin SIMMs awhile back for ~$15/ea for a
Socket 7 board that I'm still using, which is working extremely well now.
Those replaced the 2x 32MB and 4x 16MB modules that it had been using for
close to a decade. Most Socket 7 boards can't handle 512MB of ram (128MB
or 64MB max), but it is an exception to the rule and uses an Intel 430HX
(Triton II) chipset along with enough cache for all 512MB.
I still think Intel got pretty stupid in refusing to support 512MB (and
ECC) with their later Socket 7 chipsets, and especially those with SDRAM
(DIMM) support, since that shortsightedness gave the low-cost third party
chipset companies such as VIA the chance to take over that portion of the
market. Had Intel not done that, there might be far more Socket 7 boards
still in use now too. Still, even without ECC support, chipsets such as
the VIA Apollo VP2 are pretty nice.
Your board with 4 30-pin SIMM sockets likely tops
out at 16MB. AFAIK,
no one ever made 8MB 30-pin SIMMs, although at one time there were some
16MB SIMMs available at great expense (~$400.00/ea or ~$25/megabyte).
They're not as expensive now; I just recently put four of them into my
Quadra 700 (which definitely supports them) for about $40.
99% of the boards out there did not support
30-pin SIMMs larger than
4MB though.
That's the rub. Inasmuch as the 30-pin sense lines were ever "standard",
I don't think there was ever a standard sense configuration for 16MB,
which means that the chipset would have to support the rows/cols
configuration (many didn't) and the BIOS would need to manually probe
the bounds of the memory array. Many Macs supported 16MB 30- pin SIMMs,
but my recollection is that more than a few didn't.
I do have a few EISA boards that could (8
sockets, 128MB max), but I
never was able to put my hands on any 16MB 60ns 30-pin SIMMs in matched
sets of 4.
I don't recall if they're 60ns or not, but they're pretty cheap on eBay
now. You can often find them (matched sets, that is) marketed
specifically for Quadra 700s. :-) 
10:23 a.m.
--- On Sat, 6/8/13, Tothwolf <tothwolf at concentric.net> wrote:
A current Debian distribution such as 6.0 (squeeze) or
7.0
(wheezy) should work on a 486, at least in console mode
(I've not even considered trying to run current X11 on one)
but you'll need as much memory as you can install plus lots
of hard drive and a swap partition. Installation would also
likely be very slow, and you'd want to deselect everything
in 'tasksel' during installation (unless you have at least
2GB of hard drive space, but even then it will install a lot
of useless stuff) and manually install whatever packages you
need after letting it install the base OS.
Isn't that how everyone installs Debian? Just run the installer for base, and install
manually what you want, when you want it? Tasksel is pretty broken and installs all kinds
of useless junk.
I'm running a recent version of Debian (5.0) on a 486/66 with 32 megs of RAM. It runs
great, and it's one of the most stable, constantly on boxes I have. It's mostly
just a utility box, but it does it's job well. Of course, no X, but it's not like
I need X not like I'm going to try to run a browser on a 486, and that's about all
you need X for. I compiled a custom kernel for this machine to trim down the memory
footprint, I'm running 2.6.37, with just what I need added. The root filesystem is a 1
gig partition on a 1.2 gig drive - the rest of which is used for swap. I think I'm
only using 400 megs.
Linux runs fine on older hardware. Sure, it's a bit sluggish - ssh, for instance, is
pretty processor intensive, and is slow to initiate on a machine this old. This is to be
expected. Compiles are also very slow - building the kernel took well over a day. But, by
and large, most stuff runs OK, and the system is very usable. Definitely would not want to
run X, however...
-Ian
11:07 a.m.
On Sat, 8 Jun 2013, Mr Ian Primus wrote:
--- On Sat, 6/8/13, Tothwolf <tothwolf at
concentric.net> wrote:
Nope, many use tasksel and accept the defaults (as inefficient as they
are). I'd be all for scrapping tasksel, but others would scream...
Actually tasksel itself isn't the problem, its the individual "task" files
bundled with tasksel that are poorly constructed. Even the base install
will install some useless cruft though, and a careful audit will find a
number of things that can be uninstalled.
A current Debian distribution such as 6.0
(squeeze) or 7.0 (wheezy)
should work on a 486, at least in console mode (I've not even
considered trying to run current X11 on one) but you'll need as much
memory as you can install plus lots of hard drive and a swap partition.
Installation would also likely be very slow, and you'd want to deselect
everything in 'tasksel' during installation (unless you have at least
2GB of hard drive space, but even then it will install a lot of useless
stuff) and manually install whatever packages you need after letting it
install the base OS.
Isn't that how everyone installs Debian? Just run the installer for
base, and install manually what you want, when you want it? Tasksel is
pretty broken and installs all kinds of useless junk. I'm running a recent version of Debian (5.0) on a
486/66 with 32 megs of
RAM. It runs great, and it's one of the most stable, constantly on boxes
I have. It's mostly just a utility box, but it does it's job well. Of
course, no X, but it's not like I need X not like I'm going to try to
run a browser on a 486, and that's about all you need X for. I compiled
a custom kernel for this machine to trim down the memory footprint, I'm
running 2.6.37, with just what I need added. The root filesystem is a 1
gig partition on a 1.2 gig drive - the rest of which is used for swap. I
think I'm only using 400 megs.
Linux runs fine on older hardware. Sure, it's a bit sluggish - ssh, for
instance, is pretty processor intensive, and is slow to initiate on a
machine this old. This is to be expected. Compiles are also very slow -
building the kernel took well over a day. But, by and large, most stuff
runs OK, and the system is very usable. Definitely would not want to run
X, however...
You can speed up ssh a good bit if you force the use protocol 1 instead of
protocol 2 since protocol 2 is more CPU intensive. Most sshd packages
today have only protocol 2 enabled though so you'd have to add protocol 1
to /etc/ssh/sshd_config (see the sshd_config manpage) and then supply '-1'
to the ssh command. For internal use on a private network, protocol 1 is
probably fine, but for a machine with ssh open to the world, protocol 2
should be enforced.
11:19 a.m.
For internal use on a private network, [ssh] protocol
1 is probably
fine, but for a machine with ssh open to the world, protocol 2
should be enforced.
For internal use on a private network, why not just use rsh?
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
11:55 a.m.
On Sat, 8 Jun 2013, Mouse wrote:
Because ssh protocol 1 still allows the use of a private key? Even on an
internal network I'd use ssh over something like telnet since ssh with a
private key is going to prevent a compromised host or device from brute
forcing access to other stuff.
For internal
use on a private network, [ssh] protocol 1 is probably
fine, but for a machine with ssh open to the world, protocol 2 should
be enforced.
For internal use on a private network, why not just use rsh? 
10 Jun
10 Jun
1:04 a.m.
On 6/8/13 7:23 AM, "Mr Ian Primus" <ian_primus at yahoo.com> wrote:
Linux runs fine on older hardware. Sure, it's a bit sluggish - ssh, for
instance, is pretty processor intensive, and is slow to initiate on a
machine this old. This is to be expected. Compiles are also very slow -
building the kernel took well over a day. But, by and large, most stuff
runs OK, and the system is very usable. Definitely would not want to run
X, however...
My first Linux machine was a 386SX, and I ran X on it. I don't recall how
much working store I had, but I can guarantee you it wasn't much. I used
the FVWM (Feeble Virtual Window Manager) and I thought it was totally
cool. Granted, I wasn't trying to do much with it - I probably had two or
three terminal windows, a clock and xeyes running. But I also had rsh to
a local college and I started playing with networking applications and
utilities.
Yes, kernel builds were slow. But I had so much fun!
The modern distributions are jam-packed with lots of stuff that someone
apparently thinks "everyone" needs and, as a result, they are HUGE. On
one hand, they're trying to meet the expectations of the Windows crowd,
but IMHO they go too far and load up with lots and lots of *stuff* that I
for one never use. There was a message on this thread that reminded me of
installing the "base" and "dev" packages and calling it good,
installing
other tools as I needed them. After all, how hard is it to un-tar and
make?
I think that 386SX may have been a 4MB machine. Slow, but it got there.
-- Ian
10:35 a.m.
On Mon, Jun 10, 2013 at 1:04 AM, Ian King <IanK at vulcan.com> wrote:
On 6/8/13 7:23 AM, "Mr Ian Primus"
<ian_primus at yahoo.com> wrote:
My first Linux machine was a 386SX, and I ran X on it... FVWM...
... two or three terminal windows, a clock and xeyes running...
I think that 386SX may have been a 4MB machine. Slow, but it got there.
Similar to my early Linux experience, but I had a 386DX/40 w/4MB, so
not as slow, but kernel builds still took a while (and I had an AHA1540,
so unless i wanted the largest pre-built kernel, I had to make my own
to trim out all the stuff I _didn't_ need). Back those days, I ran
Slackware, but I couldn't tell you what distro versions (but I remember
the kernel versions - 0.40 was the first I remember trying, then because of
the SCSI issue, I stepped back for a bit and really dove in around 0.95).
-ethan
10:57 a.m.
On 06/10/2013 10:35 AM, Ethan Dicks wrote:
On Mon, Jun 10, 2013 at 1:04 AM, Ian King <IanK at
vulcan.com> wrote:
Same her on Slackware, it was the common distro then. My first linux PC was
a 486sx/33 with 8mb. Never got Xwindows working usefully due to oddball
video.
Next try was a P133 and that ran it fairly well with 16M.
Allsion
On 6/8/13 7:23 AM, "Mr Ian Primus"
<ian_primus at yahoo.com> wrote:
My first Linux machine was a 386SX, and I ran X on it... FVWM...
... two or three terminal windows, a clock and xeyes running...
I think that 386SX may have been a 4MB machine. Slow, but it got there.
Similar
to my early Linux experience, but I had a 386DX/40 w/4MB, so
not as slow, but kernel builds still took a while (and I had an AHA1540,
so unless i wanted the largest pre-built kernel, I had to make my own
to trim out all the stuff I _didn't_ need). Back those days, I ran
Slackware, but I couldn't tell you what distro versions (but I remember
the kernel versions - 0.40 was the first I remember trying, then because of
the SCSI issue, I stepped back for a bit and really dove in around 0.95).
-ethan
11:20 a.m.
On Mon, Jun 10, 2013 at 10:57 AM, allison <ajp166 at verizon.net> wrote:
Same her on Slackware, it was the common distro then.
My first linux PC was
a 486sx/33 with 8mb. Never got Xwindows working usefully due to oddball
video.
Yeah... I do remember that about X... the first time I got it working,
I did it on a 486 by hunting down a specific VLB card with one of the
Cirrus Logic chips on the list of supported hardware, then having to
go out and find more RAM for the video card because it was only
half-populated.
Some parts of the "good old days" I do not miss.
-ethan
11:35 a.m.
On 06/10/2013 11:20 AM, Ethan Dicks wrote:
On Mon, Jun 10, 2013 at 10:57 AM, allison <ajp166
at verizon.net> wrote:
Yes, right at the cusp of PLUG&PRAY hardware.
Foo on that still.
Allison
Same her on Slackware, it was the common distro
then. My first linux PC was
a 486sx/33 with 8mb. Never got Xwindows working usefully due to oddball
video.
Yeah... I do remember that about X... the first time I got it working,
I did it on a 486 by hunting down a specific VLB card with one of the
Cirrus Logic chips on the list of supported hardware, then having to
go out and find more RAM for the video card because it was only
half-populated.
Some parts of the "good old days" I do not miss.
-ethan
11:44 a.m.
New subject: [cctalk] Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 06/10/2013 11:20 AM, Ethan Dicks wrote:
On Mon, Jun 10, 2013 at 10:57 AM, allison <ajp166
at verizon.net> wrote:
I seem to remember doing pretty much exactly the same thing.
My first Linux experimentation was a bit later than yours, I think.
I'm pretty sure I first tried v0.97.
Same her on Slackware, it was the common distro
then. My first linux PC was
a 486sx/33 with 8mb. Never got Xwindows working usefully due to oddball
video.
Yeah... I do remember that about X... the first time I got it working,
I did it on a 486 by hunting down a specific VLB card with one of the
Cirrus Logic chips on the list of supported hardware, then having to
go out and find more RAM for the video card because it was only
half-populated. Some parts of the "good old days" I do not
miss.
Yep. Linux was a "toy" back then; I got real work done on
SPARCstations running SunOS4. But it was pretty impressive to see a
UNIX-like system come up on PC hardware.
The first time I saw it was with 386BSD, whose boot messages...right
down to the formatting of the device probe messages...were damn near
identical to those of SunOS4. I was blown away.
Linux took awhile to grow up. Now all of my desktop work is done
under Linux; I never thought I'd say that. But fast, built-like-tanks
Sun machines running Solaris still do all the server duties here, and
will for a long while yet.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
3:35 p.m.
New subject: [cctalk] Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 10 Jun 2013, at 11:44, "Dave McGuire" <mcguire at neurotica.com>
wrote:
On 06/10/2013 11:20 AM, Ethan Dicks wrote:
Speaking of 386BSD, I need a copy of 1.0. ;)
On Mon, Jun 10, 2013 at 10:57 AM, allison
<ajp166 at verizon.net> wrote:
I seem to remember doing pretty much exactly the same thing.
My first Linux experimentation was a bit later than yours, I think.
I'm pretty sure I first tried v0.97.
Same her on Slackware, it was the common distro
then. My first linux PC was
a 486sx/33 with 8mb. Never got Xwindows working usefully due to oddball
video.
Yeah... I do remember that about X... the first time I got it working,
I did it on a 486 by hunting down a specific VLB card with one of the
Cirrus Logic chips on the list of supported hardware, then having to
go out and find more RAM for the video card because it was only
half-populated. Some parts of the "good old days" I do
not miss.
Yep. Linux was a "toy" back then; I got real work done on
SPARCstations running SunOS4. But it was pretty impressive to see a
UNIX-like system come up on PC hardware.
The first time I saw it was with 386BSD, whose boot messages...right
down to the formatting of the device probe messages...were damn near
identical to those of SunOS4. I was blown away.
Linux took awhile to grow up. Now all of my desktop work is done
under Linux; I never thought I'd say that. But fast, built-like-tanks
Sun machines running Solaris still do all the server duties here, and
will for a long while yet.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
12:31 p.m.
On Mon, 10 Jun 2013, Ethan Dicks wrote:
On Mon, Jun 10, 2013 at 1:04 AM, Ian King <IanK at
vulcan.com> wrote:
I have the official Walnut Creek 2-disc set for Slackware 2.3 (July 1995)
and 4-disc set for Slackware 96 (August 1996) that I can image and make
available if needed. I don't have the bandwidth to keep them online for
/everyone/ to download, but I can make them available to whoever wants to
mirror them for others. The 2.3 set has a 1.2.8 kernel and 96 has a 2.0.0
kernel. Slackware 96 ran just fine on any 386 with 4MB of memory and would
install on a very small ~25MB or so hard drive without gcc, X, etc. If I
had the chance to obtain any of the other older versions I'd make those
available as well, but they don't seem to turn up all that often now.
On 6/8/13 7:23 AM, "Mr Ian Primus"
<ian_primus at yahoo.com> wrote:
My first Linux machine was a 386SX, and I ran X on it... FVWM... ...
two or three terminal windows, a clock and xeyes running...
I think that 386SX may have been a 4MB machine. Slow, but it got
there.
Similar to my early Linux experience, but I had a 386DX/40 w/4MB, so not
as slow, but kernel builds still took a while (and I had an AHA1540, so
unless i wanted the largest pre-built kernel, I had to make my own to
trim out all the stuff I _didn't_ need). Back those days, I ran
Slackware, but I couldn't tell you what distro versions (but I remember
the kernel versions - 0.40 was the first I remember trying, then because
of the SCSI issue, I stepped back for a bit and really dove in around
0.95). 
12:41 p.m.
On 6/10/13 9:31 AM, Tothwolf wrote:
I have the official Walnut Creek 2-disc set for
Slackware 2.3 (July 1995) and 4-disc set for Slackware 96 (August 1996) that I can image
and make available if needed.
I've been collecting walnut creek discs for a while so the Jul 95 iso and cover scan
would be nice.
It wouldn't surprise me if Jason Scott has the whole set already. I've wondered if
the contents of WC's ftp site survived somewhere.
MD5 (199310_Simtel_MSDOS/SIMTEL_1093.iso) = fa0e31a351448b803df17e9d16429f39
MD5 (199403_Hobbes_OS2/HOBBES_0394.iso) = fa7f3c1763315d93d48b0071e1c93558
MD5 (199403_Internet_Info/INET_0394.iso) = a77f353a1677c8e0c0e273e5d4e792ff
MD5 (199411_CPM/CPM_CDROM.iso) = 42571cceef022c582f02dabb4ac68e2c
MD5 (199411_Linux_Yggdrasil/Yggdrasil_Linux_1194.iso) = b5b65933c23de6f6d3b14abfcf98f1b5
MD5 (199501_CDROM_Sampler/CDROM_Sampler_9501.iso) = b23fbe1166983e80405a8e94b9a0bd32
MD5 (199501_CPM/CPM_CDROM.iso) = faa8c5ba9d1184b7bfc34d999a8c42c4
MD5 (199503_Internet_Info/INET_0395.iso) = c31f122d6e1ec6be7f0418b877613852
MD5 (199504_Best_of_Walnut_Creek_CDROM/Best_of_Walnut_Creek_199504.iso) =
a4aa8bda9224e5bc2b8c16ae008d6ac3
MD5 (199504_Giga_Games_2/GIGAGAMES2.iso) = 00d8313dcc151ed306f0bb1365ea51ca
MD5 (199504_POV_RAY/POV_Ray.iso) = 3a642b1d46f57f20ea9cd5d4fbdcb01d
MD5 (199505_Simtel_MSDOS/SIMTEL0595A.iso) = fbee4594e1d596d0e1ee715253af543c
MD5 (199505_Simtel_MSDOS/SIMTEL0595B.iso) = 3a08d918302e7f31591a434c7cc0e640
MD5 (199507_TeX/TEX_0595_V1.iso) = e0db869ae5071a38fc5f2ce18353879a
MD5 (199507_TeX/TEX_0595_V2.iso) = 864f0a8d236a3f4954d7d60cc2f2a23e
MD5 (199508_The_Business_Software_Library/BUS_LIB95.iso) =
59dc633b8f94333d29d9e183d6d5c6e5
MD5 (199509_Simtel_Windows/Simtel_Windows_0995.iso) = 541f67fa7450bd66b7e2e9d2cc812ae0
MD5 (199510_Avalon/AVALON.iso) = 0ee12594c67ed0f5dd3bdace74330fbe
MD5 (199510_Font_Garden/Font Garden_1095.iso) = 0b302e17766bbc72c3e02483a15eb107
MD5 (199511_Linux_Slackware_3.0/SLACK1195_1.iso) = 631dfae9567c9e9f34980af100539b9c
MD5 (199511_Linux_Slackware_3.0/SLAK10_95_2.iso) = 5c9ed883cf1685244d0594863c6e4ffe
MD5 (199602_GNU/GNU_0296.iso) = 5af203f6e24f7d6d629ed1ddc09e508c
MD5 (199602_Task_Buster/Task_Buster_0296.iso) = aef3585efaea78138af00638d4d2cd9e
MD5 (199602_Walnut_Creek_C_UG/CUG_296.iso) = 1774993b8a120fb9f5d0b927c805ff70
MD5 (199603_Newt_For_Windows_NT/NEWT_0396.iso) = 5e2107b87744ba7f6e6e5741ddbb6379
MD5 (199604_Perl/CPAN_0496.iso) = d2dd63618e32749ac5437aa57da465e3
MD5 (199605_BSD_Docs/BSD_DOCS.iso) = adb33000488ad8a18f6f9bbe61f53e28
MD5 (199608_Linux_Slackware/slack96_d1_0896.iso) = cc7d58178f01d8c80219575f9f4ea2ec
MD5 (199608_Linux_Slackware/slack96_d2_0896.iso) = 187a321522f0842b71fa77ef75ff398d
MD5 (199608_Linux_Slackware/slack96_d3_0896.iso) = aa07e28421b255902151f2f542da22ca
MD5 (199608_Linux_Slackware/slack96_d4_0896.iso) = 88cc616e1c12fbcc0e56558ef9c53166
MD5 (199608_MultiMedia_CDROM/MM-CD-ROM.iso) = c12dd6875afcc46f940132cc3dca60e6
MD5 (199610_Internet_Info/INET_996.iso) = 1714609b8afd93b40b6ebccdd21a9072
MD5 (199612_Toolkit_for_Delphi/Delphi_Tlkt.iso) = e5a244cde8c9864fb2ce1ef5b88ba315
MD5 (199702_CICA/CICA2_0297.iso) = babb3b4ea08e2c5d617ada12228f79d8
MD5 (199702_CICA/CICA_0297.iso) = 22c5e535de4eec3860543f3fdfdefe9d
MD5 (199703_Ada/01ada0397.iso) = 60d9f450917371a6cb1ac069eb063503
MD5 (199703_Ada/02ada0397.iso) = 74134d17107c0405050185f76627550a
MD5 (199803_Simtel_MSDOS/Simtel398_1.iso) = 5ab31373f718f55d7328706da2faaf3f
MD5 (199803_Simtel_MSDOS/Simtel398_2.iso) = d0ab2b0ad976a2170c5555f14e97354a
MD5 (199911_Linux_Slackware_7.0/01_Slackware_7.0_Installation.iso) =
6d553749aa71683d31c3064ac32a2c1c
MD5 (199911_Linux_Slackware_7.0/02_Bootable_Slackware_7.0.iso) =
5d4e971a999a818d87894bb419bc8e7b
MD5 (199911_Linux_Slackware_7.0/03_Slackware_7.0_source.iso) =
42be5a88cec743ed4fd7b676eed2b25b
1:13 p.m.
On 6/10/2013 9:41 AM, Al Kossow wrote:
On 6/10/13 9:31 AM, Tothwolf wrote:
I have an 8 CD set entitled "Linux Internet Archives" (Winter 1997) from
Yggdrasil; it includes Debian 1.1, Slackware 3.1, RedHat 4.0, i386 Mach
Linux, very early AXP Linux, Mach-based Hurd 0.0, and a bunch of stuff
I have the official Walnut Creek 2-disc set for
Slackware 2.3 (July
1995) and 4-disc set for Slackware 96 (August 1996) that I can image
and make available if needed.
I've been collecting walnut creek discs for a while so the Jul 95 iso
and cover scan would be nice.
It wouldn't surprise me if Jason Scott has the whole set already. I've
wondered if the contents of WC's ftp site survived somewhere. from sunsite and MIT. If there's any interest I
can make images of
these discs.
(This set was seriously awesome to me back in '97 since downloading this
stuff on my 28.8 modem was rather impractical...)
- Josh
MD5 (199310_Simtel_MSDOS/SIMTEL_1093.iso) =
fa0e31a351448b803df17e9d16429f39
MD5 (199403_Hobbes_OS2/HOBBES_0394.iso) =
fa7f3c1763315d93d48b0071e1c93558
MD5 (199403_Internet_Info/INET_0394.iso) =
a77f353a1677c8e0c0e273e5d4e792ff
MD5 (199411_CPM/CPM_CDROM.iso) = 42571cceef022c582f02dabb4ac68e2c
MD5 (199411_Linux_Yggdrasil/Yggdrasil_Linux_1194.iso) =
b5b65933c23de6f6d3b14abfcf98f1b5
MD5 (199501_CDROM_Sampler/CDROM_Sampler_9501.iso) =
b23fbe1166983e80405a8e94b9a0bd32
MD5 (199501_CPM/CPM_CDROM.iso) = faa8c5ba9d1184b7bfc34d999a8c42c4
MD5 (199503_Internet_Info/INET_0395.iso) =
c31f122d6e1ec6be7f0418b877613852
MD5
(199504_Best_of_Walnut_Creek_CDROM/Best_of_Walnut_Creek_199504.iso) =
a4aa8bda9224e5bc2b8c16ae008d6ac3
MD5 (199504_Giga_Games_2/GIGAGAMES2.iso) =
00d8313dcc151ed306f0bb1365ea51ca
MD5 (199504_POV_RAY/POV_Ray.iso) = 3a642b1d46f57f20ea9cd5d4fbdcb01d
MD5 (199505_Simtel_MSDOS/SIMTEL0595A.iso) =
fbee4594e1d596d0e1ee715253af543c
MD5 (199505_Simtel_MSDOS/SIMTEL0595B.iso) =
3a08d918302e7f31591a434c7cc0e640
MD5 (199507_TeX/TEX_0595_V1.iso) = e0db869ae5071a38fc5f2ce18353879a
MD5 (199507_TeX/TEX_0595_V2.iso) = 864f0a8d236a3f4954d7d60cc2f2a23e
MD5 (199508_The_Business_Software_Library/BUS_LIB95.iso) =
59dc633b8f94333d29d9e183d6d5c6e5
MD5 (199509_Simtel_Windows/Simtel_Windows_0995.iso) =
541f67fa7450bd66b7e2e9d2cc812ae0
MD5 (199510_Avalon/AVALON.iso) = 0ee12594c67ed0f5dd3bdace74330fbe
MD5 (199510_Font_Garden/Font Garden_1095.iso) =
0b302e17766bbc72c3e02483a15eb107
MD5 (199511_Linux_Slackware_3.0/SLACK1195_1.iso) =
631dfae9567c9e9f34980af100539b9c
MD5 (199511_Linux_Slackware_3.0/SLAK10_95_2.iso) =
5c9ed883cf1685244d0594863c6e4ffe
MD5 (199602_GNU/GNU_0296.iso) = 5af203f6e24f7d6d629ed1ddc09e508c
MD5 (199602_Task_Buster/Task_Buster_0296.iso) =
aef3585efaea78138af00638d4d2cd9e
MD5 (199602_Walnut_Creek_C_UG/CUG_296.iso) =
1774993b8a120fb9f5d0b927c805ff70
MD5 (199603_Newt_For_Windows_NT/NEWT_0396.iso) =
5e2107b87744ba7f6e6e5741ddbb6379
MD5 (199604_Perl/CPAN_0496.iso) = d2dd63618e32749ac5437aa57da465e3
MD5 (199605_BSD_Docs/BSD_DOCS.iso) = adb33000488ad8a18f6f9bbe61f53e28
MD5 (199608_Linux_Slackware/slack96_d1_0896.iso) =
cc7d58178f01d8c80219575f9f4ea2ec
MD5 (199608_Linux_Slackware/slack96_d2_0896.iso) =
187a321522f0842b71fa77ef75ff398d
MD5 (199608_Linux_Slackware/slack96_d3_0896.iso) =
aa07e28421b255902151f2f542da22ca
MD5 (199608_Linux_Slackware/slack96_d4_0896.iso) =
88cc616e1c12fbcc0e56558ef9c53166
MD5 (199608_MultiMedia_CDROM/MM-CD-ROM.iso) =
c12dd6875afcc46f940132cc3dca60e6
MD5 (199610_Internet_Info/INET_996.iso) =
1714609b8afd93b40b6ebccdd21a9072
MD5 (199612_Toolkit_for_Delphi/Delphi_Tlkt.iso) =
e5a244cde8c9864fb2ce1ef5b88ba315
MD5 (199702_CICA/CICA2_0297.iso) = babb3b4ea08e2c5d617ada12228f79d8
MD5 (199702_CICA/CICA_0297.iso) = 22c5e535de4eec3860543f3fdfdefe9d
MD5 (199703_Ada/01ada0397.iso) = 60d9f450917371a6cb1ac069eb063503
MD5 (199703_Ada/02ada0397.iso) = 74134d17107c0405050185f76627550a
MD5 (199803_Simtel_MSDOS/Simtel398_1.iso) =
5ab31373f718f55d7328706da2faaf3f
MD5 (199803_Simtel_MSDOS/Simtel398_2.iso) =
d0ab2b0ad976a2170c5555f14e97354a
MD5 (199911_Linux_Slackware_7.0/01_Slackware_7.0_Installation.iso) =
6d553749aa71683d31c3064ac32a2c1c
MD5 (199911_Linux_Slackware_7.0/02_Bootable_Slackware_7.0.iso) =
5d4e971a999a818d87894bb419bc8e7b
MD5 (199911_Linux_Slackware_7.0/03_Slackware_7.0_source.iso) =
42be5a88cec743ed4fd7b676eed2b25b
1:35 p.m.
On Mon, 10 Jun 2013, Al Kossow wrote:
On 6/10/13 9:31 AM, Tothwolf wrote:
I suppose I'll image the Slackware 2.3 discs and then also see if the
Slackware 96 discs I have match those MD5s. I found that there were two
releases that were both called "Slackware 96", but since my set has
"August 1996" printed on the cover, it may very well be the same 4-disc
set as the 4 discs above.
I've hunted for the old Walnut Creek archives for quite some time but
never found all that much. From what I remember, it was pretty much all
junked after they changed names and moved away from Slackware and got way
more involved with FreeBSD. Somewhere I have a lot of links saved away but
I don't know when I'll have the time to go hunt that stuff down.
On the subject of old shareware type cdroms, I have a /ton/ of stuff that
needs to be made available for archive.org (some of which are still in my
cdrom jukebox), but I haven't a clue what the process of getting that
stuff online will entail.
I have the official Walnut Creek 2-disc set for
Slackware 2.3 (July
1995) and 4-disc set for Slackware 96 (August 1996) that I can image
and make available if needed.
I've been collecting walnut creek discs for a while so the Jul 95 iso
and cover scan would be nice. It wouldn't surprise me if Jason Scott has
the whole set already. I've wondered if the contents of WC's ftp site
survived somewhere.
MD5 (199608_Linux_Slackware/slack96_d1_0896.iso) =
cc7d58178f01d8c80219575f9f4ea2ec
MD5 (199608_Linux_Slackware/slack96_d2_0896.iso) =
187a321522f0842b71fa77ef75ff398d
MD5 (199608_Linux_Slackware/slack96_d3_0896.iso) =
aa07e28421b255902151f2f542da22ca
MD5 (199608_Linux_Slackware/slack96_d4_0896.iso) =
88cc616e1c12fbcc0e56558ef9c53166
3:29 p.m.
On 6/10/13 10:35 AM, Tothwolf wrote:
On the subject of old shareware type cdroms, I have a
/ton/ of stuff that needs to be made available for archive.org (some of which are still in
my cdrom jukebox), but I haven't a clue what the
process of getting that stuff online will entail.
just contact Jason Scott jason at textfiles.com
6:18 p.m.
On Mon, 10 Jun 2013, Tothwolf wrote:
I have the official Walnut Creek 2-disc set for
Slackware 2.3 (July 1995) and
4-disc set for Slackware 96 (August 1996) that I can image and make available
if needed. I don't have the bandwidth to keep them online for /everyone/ to
download, but I can make them available to whoever wants to mirror them for
others. The 2.3 set has a 1.2.8 kernel and 96 has a 2.0.0 kernel. Slackware
96 ran just fine on any 386 with 4MB of memory and would install on a very
small ~25MB or so hard drive without gcc, X, etc. If I had the chance to
obtain any of the other older versions I'd make those available as well, but
they don't seem to turn up all that often now.
I have Yggdrasil Linux circa-1994 on CD. It will definitely run on a 386
with 4M of memory. Kernel 1.0.something IIRC.
Is anyone hosting images of the these old chestnuts?
--
1:18 p.m.
On Mon, 10 Jun 2013, Ian King wrote:
On 6/8/13 7:23 AM, "Mr Ian Primus"
<ian_primus at yahoo.com> wrote:
A 386DX-20 and 4MB would run X with FVWM and I used just that combination
for several years. With 8MB and later 16MB, Netscape worked just fine on a
386DX under X. For comparison purposes, Netscape (the 16 bit version) also
worked fine on Windows 3.1 on a 286 with 4 to 8MB of ram. This of course
assumes you have a swap partition or paging file.
The issue of modern distributions using so much in the way of resources is
caused by several factors. At a lower level, you have userspace developers
today writing very very inefficient code. In some cases they overuse stuff
like malloc() and free() which are more CPU intensive but can potentially
use less memory, and in other cases developers are using far too many
static buffers, which while using far less CPU, can pre-allocate and eat
up a lot more memory. Those combined with just far less efficient code in
general are the worst offenders when it comes to CPU and memory usage.
As for disk space, a lot of that is due to the distribution maintainers.
Software packages today are not compiled to be space/memory/CPU efficient,
but rather to work on the largest subset of what the package maintainers
/personally/ consider to be "modern" CPUs (which right now, are multi-core
AMD and Intel CPUs). Those packages eat up a lot more disk space (and
often memory) and it is now the norm for most distributions to build
everything with debugging symbols. The thinking is that debug symbols
somehow help debug large programs that crash, however it really hurts the
performance of most software, and is very very silly for basic UNIX-like
command line utilities.
I've both developed and maintained software, as well as packaged all sorts
of software for all sorts of distributions including those that used the
RPM (RedHat) and DEB (Debian) package formats (and others), so I have some
experience in seeing this stuff from all sides. I found that any time I
tried to push for more efficient code, I got pushback from other
developers and package maintainers because "fixing" and making stuff more
CPU/memory/space efficient was considered to be a /far/ less important
task than rushing to push out the next bleeding-edge version of something.
...and don't even get me started on some of the Debian maintainers adding
buggy third-party patches to some of the software they distribute...
Dealing with upstream bug reports from users complaining about bugs and
crashes introduced by such patches wastes a lot of developer time. Your
average package maintainer does /not/ know the software better than the
upstream developers, and those that try to be clever and "fix" stuff that
doesn't really need to be fixed tend to create even more work for the
upstream developers when users begin complaining.
I'm slowing coming to the opinion that a first year C developer should be
/forced/ to develop on a CPU and memory constrained platform, such as a
386 with 4-8MB of memory vs a modern multi-core CPU with multiple
gigabytes of memory, so that they will learn first hand how to write more
efficient C code. I wonder how many of the current userspace developers
and package maintainers have ever even touched a 386 based machine, let
alone something even more resource limited?
Linux runs fine on older hardware. Sure, it's
a bit sluggish - ssh, for
instance, is pretty processor intensive, and is slow to initiate on a
machine this old. This is to be expected. Compiles are also very slow -
building the kernel took well over a day. But, by and large, most stuff
runs OK, and the system is very usable. Definitely would not want to
run X, however...
My first Linux machine was a 386SX, and I ran X on it. I don't recall
how much working store I had, but I can guarantee you it wasn't much.
I used the FVWM (Feeble Virtual Window Manager) and I thought it was
totally cool. Granted, I wasn't trying to do much with it - I probably
had two or three terminal windows, a clock and xeyes running. But I
also had rsh to a local college and I started playing with networking
applications and utilities.
Yes, kernel builds were slow. But I had so much fun!
The modern distributions are jam-packed with lots of stuff that someone
apparently thinks "everyone" needs and, as a result, they are HUGE. On
one hand, they're trying to meet the expectations of the Windows crowd,
but IMHO they go too far and load up with lots and lots of *stuff* that
I for one never use. There was a message on this thread that reminded
me of installing the "base" and "dev" packages and calling it good,
installing other tools as I needed them. After all, how hard is it to
un-tar and make?
I think that 386SX may have been a 4MB machine. Slow, but it got there.
2:34 p.m.
On Jun 10, 2013, at 1:18 PM, Tothwolf <tothwolf at concentric.net> wrote:
As for disk space, a lot of that is due to the
distribution maintainers. Software packages today are not compiled to be space/memory/CPU
efficient, but rather to work on the largest subset of what the package maintainers
/personally/ consider to be "modern" CPUs (which right now, are multi-core AMD
and Intel CPUs). Those packages eat up a lot more disk space (and often memory) and it is
now the norm for most distributions to build everything with debugging symbols. The
thinking is that debug symbols somehow help debug large programs that crash, however it
really hurts the performance of most software, and is very very silly for basic UNIX-like
command line utilities.
In what sense does it hurt performance? I didn't think the symbols got
loaded into RAM unless you were actually opening something up under GDB
or something that needed to read the symbol table. Or are you talking
about startup performance when the symbol table is used to bind the
dynamic libs?
I'm slowing coming to the opinion that a first
year C developer should be /forced/ to develop on a CPU and memory constrained platform,
such as a 386 with 4-8MB of memory vs a modern multi-core CPU with multiple gigabytes of
memory, so that they will learn first hand how to write more efficient C code. I wonder
how many of the current userspace developers and package maintainers have ever even
touched a 386 based machine, let alone something even more resource limited?
Or a reasonably-sized embedded ARM SBC, which are a little easier to
come by these days. I'm specifically talking about something like an
ARM7/9 that can run Linux, but an example with an MMU-less embedded OS
would work as well, provided that it's self-hosting; there are also
plenty of developers who would do well to slim down their compile times
as well, so being required to compile on the same machine you run your
code on is a decent requirement as well (if not always reflective of
reality).
That, and small 8-bits, especially the PIC (which, as much as I dislike
it as a CPU, is a great example because of its rather parsimonious
instruction set and resource allotment). It REALLY emphasizes the
difference between C and assembly for embedded work that so many just
don't get (one of our customers asked us to do an AES engine on a PIC
in BASIC, no joke).
- Dave
11 Jun
11 Jun
12:17 a.m.
(one of our customers asked us to do an AES engine on
a PIC in BASIC,
no joke).
I once did IDEA in MUF (a language rather like FORTH with all the real
elegance ripped out). I feel confident I could do AES in BASIC, though
whether the result fits in a PIC depends on the PIC and the BASIC.
And, of course, the result isn't exactly going to be fast.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
10 Jun
10 Jun
9:12 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 10/06/13 1:18 PM, Tothwolf wrote:
...
I'm slowing coming to the opinion that a first year C developer should
be /forced/ to develop on a CPU and memory constrained platform, such as
Agree 100%. It wouldn't help if they *continued* to develop on machines
a few years behind the curve, as well.
--T
a 386 with 4-8MB of memory vs a modern multi-core CPU
with multiple
gigabytes of memory, so that they will learn first hand how to write
more efficient C code. I wonder how many of the current userspace
developers and package maintainers have ever even touched a 386 based
machine, let alone something even more resource limited?
9:22 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
We develop our vertical client/server app on powerful development machines,
then use load testing software to see how it scales on multiple user loads
and slower client machines.
Using one machine is a non-starter, as we have different hardware
requirements depending on what the user is doing. If you're doing basic
accounting, you don't need a lot of machine. If you are doing statistical
analysis on millions of data points, you do.
On Mon, Jun 10, 2013 at 9:12 PM, Toby Thain <toby at telegraphics.com.au>wrote:
On 10/06/13 1:18 PM, Tothwolf wrote:
...
I'm slowing coming to the opinion that a first year C developer should
be /forced/ to develop on a CPU and memory constrained platform, such as
Agree 100%. It wouldn't help if they *continued* to develop on machines a
few years behind the curve, as well.
--T
a 386 with 4-8MB of memory vs a modern multi-core CPU with multiple
gigabytes of memory, so that they will learn
first hand how to write
more efficient C code. I wonder how many of the current userspace
developers and package maintainers have ever even touched a 386 based
machine, let alone something even more resource limited?
9:37 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 10/06/13 9:22 PM, Jason McBrien wrote:
We develop our vertical client/server app on powerful
development machines,
then use load testing software to see how it scales on multiple user loads
and slower client machines.
Using one machine is a non-starter, as we have different hardware
requirements depending on what the user is doing. If you're doing basic
accounting, you don't need a lot of machine. If you are doing statistical
analysis on millions of data points, you do.
Right, every situation is different. For code that runs on client
machines, bringing the developer face to face with older spec is
probably a good idea. For back-end code, it doesn't necessarily apply.
--Toby
On Mon, Jun 10, 2013 at 9:12 PM, Toby Thain<toby at telegraphics.com.au>wrote:
On 10/06/13 1:18 PM, Tothwolf wrote:
...
I'm slowing coming to the opinion that a first year C developer should
be /forced/ to develop on a CPU and memory constrained platform, such as
Agree 100%. It wouldn't help if they *continued* to develop on machines a
few years behind the curve, as well.
--T
a 386 with 4-8MB of memory vs a modern multi-core CPU with multiple
gigabytes of memory, so that they will learn
first hand how to write
more efficient C code. I wonder how many of the current userspace
developers and package maintainers have ever even touched a 386 based
machine, let alone something even more resource limited?
12 Jun
12 Jun
6:19 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Mon, Jun 10, 2013 at 09:12:19PM -0400, Toby Thain wrote:
On 10/06/13 1:18 PM, Tothwolf wrote:
Software is rarely written by a single developer, but part of a team effort
with a client or management dictating the trade-offs of hardware costs versus
developer effort on performance improvements, reducing the defect rate, and/or
adding more features. It's hardly fair to punish the developer for the
capriciousness of their employer.
As it so happens, outside of the willy-waving from Google et al about how well
they treat their developers, most developers are treated like dirt and get
whatever cast-off hardware the IT department deigns to let them have. I have
fired a number of clients for taking the piss in this regard. No, if you want
me to write Unix software for you, insisting that I must only use a ratty old
Pentium 4 with 512MB of RAM running a locked-down Windows XP that won't even
let me install PuTTY or an editor is not acceptable.
I'm slowing coming to the opinion that a
first year C developer should be
/forced/ to develop on a CPU and memory constrained platform, such as
Agree 100%.
It wouldn't help if they *continued* to develop on machines a few
years behind the curve, as well. 
7:09 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 6/12/2013 4:19 PM, Peter Corlett wrote:
As it so happens, outside of the willy-waving from
Google et al about how well
they treat their developers, most developers are treated like dirt and get
whatever cast-off hardware the IT department deigns to let them have. I have
fired a number of clients for taking the piss in this regard. No, if you want
me to write Unix software for you, insisting that I must only use a ratty old
Pentium 4 with 512MB of RAM running a locked-down Windows XP that won't even
let me install PuTTY or an editor is not acceptable.
Umm... How do you write Unix software if you are running windows?
Ben.
PS I keep forgetting that the concept of 'a single user UNIX' was never
a selling point so the PC world never got a UNIX like product for the
86's/286's to replace DOS. OS/9 (6809) did come close but a bit to late
for unix like product.
7:54 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 06/12/2013 04:09 PM, ben wrote:
.
PS I keep forgetting that the concept of 'a single
user UNIX' was never
a selling point so the PC world never got a UNIX like product for the
86's/286's to replace DOS. OS/9 (6809) did come close but a bit to late
for unix like product.
If you look at the release notes for MS-DOS 2.0, Microsoft was pretty
clear in stating that at some point, the development path for DOS and
XENIX was supposed to merge.
I guess that objective went the way of new versions of Microsoft OS/2...
--Chuck
14 Jun
14 Jun
7:14 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Wed, Jun 12, 2013 at 05:09:54PM -0600, ben wrote:
[...]
Umm... How do you write Unix software if you are
running windows?
I pick the first off this list based on what the client and/or their
bureaucracy permits me to do:
1) Ignore the client-provided Windows box and do everything on my own Mac
laptop.
2) Wipe said Windows box and install a real operating system.
3) Check my bank balance and/or the job market to decide whether I am in a
position to fire a client that's being silly.
4) Install virtualisation software and boot up a real operating system.
5) Install PuTTY and ssh to a real computer.
6) Install Cygwin and/or other development tools, and hope I don't accidentally
write something non-portable.
7) Give up, because it is impossible to write software now that the client has
vetoed all possible means, so just tread water until my finances and/or the
job market improves then fire the client.
Ben.
PS I keep forgetting that the concept of 'a single user UNIX' was never
a selling point so the PC world never got a UNIX like product for the
86's/286's to replace DOS. OS/9 (6809) did come close but a bit to late
for unix like product.
BeOS was sort of a single user Unix: it was sufficiently POSIXy for me to fire
up a shell prompt and feel at home, although it did not support multiple users
at all.
These days, MacOS X is commonly used as a single-user workstation, even though
it *does* support multiple users.
9:13 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 6/14/2013 5:14 PM, Peter Corlett wrote:
On Wed, Jun 12, 2013 at 05:09:54PM -0600, ben wrote:
[...]
THAT IS NOT RUNNING WINDOWS*
* 6 almost works. The whole point is if you are writing unix software
why are you on a windows machine? While a PC does count as real computer
I/O, CPU, Memory what was a *real machine* back then and what CPU.
Ben.
Umm... How do you write Unix software if you are
running windows?
I pick the first off this list based on what the client and/or their
bureaucracy permits me to do:
1) Ignore the client-provided Windows box and do everything on my own Mac
laptop.
2) Wipe said Windows box and install a real operating system.
3) Check my bank balance and/or the job market to decide whether I am in a
position to fire a client that's being silly.
4) Install virtualisation software and boot up a real operating system.
5) Install PuTTY and ssh to a real computer.
6) Install Cygwin and/or other development tools, and hope I don't accidentally
write something non-portable.
7) Give up, because it is impossible to write software now that the client has
vetoed all possible means, so just tread water until my finances and/or the
job market improves then fire the client.
9:24 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 15 June 2013 02:13, ben <bfranchuk at jetnet.ab.ca> wrote:
THAT IS NOT RUNNING WINDOWS*
* 6 almost works. The whole point is if you are writing unix software why
are you on a windows machine? While a PC does count as real computer
Yeah, but the trouble is, in a day job, one typically does not get the
choice of what one runs. One gets what one is given.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
9:48 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
* 6 almost works. The whole point is if you are
writing unix software
why are you on a windows machine? While a PC does count as real computer
If Microsoft could do Xenix development under MS-DOS, why can't we do it
under Windoze :-?
10:20 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 6/14/2013 7:48 PM, Fred Cisin wrote:
I am still trying to figure what windows was developed on.I would not
surprise me in Microsoft has a used a *better* machine than a DOS PC.
Ben.
* 6 almost
works. The whole point is if you are writing unix software
why are you on a windows machine? While a PC does count as real computer
If Microsoft could do Xenix development under MS-DOS, why can't we do it
under Windoze :-?
11:24 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, 15 Jun 2013, ben wrote:
--
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
On 6/14/2013 7:48 PM, Fred Cisin wrote:
I am still trying to figure what windows was developed on.I would not
surprise me in Microsoft has a used a *better* machine than a DOS PC.
Ben.
Punched cards, obviously! (or a mac for maximum irony)
* 6
almost works. The whole point is if you are writing unix software
why are you on a windows machine? While a PC does count as real computer
If Microsoft could do Xenix development under MS-DOS, why can't we do it
under Windoze :-?
15 Jun
15 Jun
3:26 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 15/06/2013 04:24, Cory Smelosky wrote:
On Fri, 15 Jun 2013, ben wrote:
Depends on which
windows...
... do you think they actually used OS/2 for any windows development?
On 6/14/2013 7:48 PM, Fred Cisin wrote:
I am still trying to figure what windows was developed on.I would not
surprise me in Microsoft has a used a *better* machine than a DOS PC.
Ben.
Punched cards, obviously! (or a mac for maximum irony) * 6
almost works. The whole point is if you are writing unix software
why are you on a windows machine? While a PC does count as real
computer
If Microsoft could do Xenix development under MS-DOS, why can't we
do it
under Windoze :-?
14 Jun
14 Jun
10:01 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 06/14/2013 09:24 PM, Liam Proven wrote:
On 15 June 2013 02:13, ben <bfranchuk at
jetnet.ab.ca> wrote:
...then one needs to get a job working for decent people. These
days, at least in the USA, beggars can't be choosers, sure, but there
are GOOD jobs around. If I can find one, anyone can.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
THAT IS NOT RUNNING WINDOWS*
* 6 almost works. The whole point is if you are writing unix software why
are you on a windows machine? While a PC does count as real computer
Yeah, but the trouble is, in a day job, one typically does not get the
choice of what one runs. One gets what one is given.
10:29 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 6/14/2013 8:01 PM, Dave McGuire wrote:
...then one needs to get a job working for decent
people. These
days, at least in the USA, beggars can't be choosers, sure, but there
are GOOD jobs around. If I can find one, anyone can.
That is why I got out of computing... New job.
-Dave
BTW. How many people really have 486's kicking around. I live in a
an apartment and have no room to keep *lots* of stuff.
Ben's poll:
A) Never used a PC
B) Power user must *have the best gaming system*
C) Upgrades ever 5 years as the old system is too outdated.
D) Upgrades after major smoke and fire.
E) The thrift shop has the newer model.
F) It works for my needs, why change.
11:32 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, 15 Jun 2013, ben wrote:
--
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
On 6/14/2013 8:01 PM, Dave McGuire wrote:
I mostly fit in that category.
...then one needs to get a job working for
decent people. These
days, at least in the USA, beggars can't be choosers, sure, but there
are GOOD jobs around. If I can find one, anyone can.
That is why I got out of computing... New job.
-Dave
BTW. How many people really have 486's kicking around. I live in a
an apartment and have no room to keep *lots* of stuff.
Ben's poll:
A) Never used a PC
B) Power user must *have the best gaming system*
C) Upgrades ever 5 years as the old system is too outdated.
D) Upgrades after major smoke and fire. E) The thrift shop has the newer model.
Define "newer"...I upgrade in reverse chronological order.
F) It works for my needs, why change.
I apply this to a lot of things, too.
15 Jun
15 Jun
3:50 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What
BTW. How many people really have 486's kicking
around. I live in a
Not so much 'kicking around' as 'my most powrful PC'
an apartment and have no room to keep *lots* of
stuff.
Ben's poll:
A) Never used a PC
B) Power user must *have the best gaming system*
C) Upgrades ever 5 years as the old system is too outdated.
D) Upgrades after major smoke and fire.
E) The thrift shop has the newer model.
F) It works for my needs, why change.
F. This machine is still a 5170 at heart. OK, there's a 486 kldugeboard
in the 286 socket. And the only part I don't have a schemtic for is the
hard drive...
-tony
14 Jun
14 Jun
10:32 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 15 June 2013 03:01, Dave McGuire <mcguire at neurotica.com> wrote:
...then one needs to get a job working for decent
people. These days, at
least in the USA, beggars can't be choosers, sure, but there are GOOD jobs
around. If I can find one, anyone can.
Good for you.
Have I mentioned recently the roughly *3,000 job applications* I've
made around the world since the end of my last F/T role in 2008?
Those are all for positions I'm well-qualified for with abundant
experience, natch. 25Y in IT, I've supported more than a dozen OSes
professionally; I know every OS, server and desktop, in current
business use, and most of the big apps.
I average 2-3 interviews a year, if that.
It may be only moderately bad in the USA. Elsewhere in the world, it's
/terrible./ I believe there are roughly 20-25? more people looking for
work than jobs in the UK currently; the /average/ number of applicants
is around 18-23 per job, depending who you ask, and that includes the
really crappy ones.
In professional roles, it ranges from around 65-70 applicants per job
to over 900 per job for desirable positions; I have head of jobs
getting several thousand applicants as a matter of routine.
The last 2 positions I landed lasted 8 days and 6 weeks, respectively.
I am still on 2 months' unpaid leave from the last one (end 2012) -
I've been on those 2 months leave since the 2nd week of December. And
that was switching to a whole new industry for me.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
15 Jun
15 Jun
7:32 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 06/14/2013 10:32 PM, Liam Proven wrote:
Umm.
I'm sorry you're going through that. I wish you better luck in the
future. (seriously)
You could come here, but you might value freedom and privacy, which
we don't really "do" anymore.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
...then one
needs to get a job working for decent people. These days, at
least in the USA, beggars can't be choosers, sure, but there are GOOD jobs
around. If I can find one, anyone can.
Good for you.
Have I mentioned recently the roughly *3,000 job applications* I've
made around the world since the end of my last F/T role in 2008?
Those are all for positions I'm well-qualified for with abundant
experience, natch. 25Y in IT, I've supported more than a dozen OSes
professionally; I know every OS, server and desktop, in current
business use, and most of the big apps.
I average 2-3 interviews a year, if that.
It may be only moderately bad in the USA. Elsewhere in the world, it's
/terrible./ I believe there are roughly 20-25? more people looking for
work than jobs in the UK currently; the /average/ number of applicants
is around 18-23 per job, depending who you ask, and that includes the
really crappy ones.
In professional roles, it ranges from around 65-70 applicants per job
to over 900 per job for desirable positions; I have head of jobs
getting several thousand applicants as a matter of routine.
The last 2 positions I landed lasted 8 days and 6 weeks, respectively.
I am still on 2 months' unpaid leave from the last one (end 2012) -
I've been on those 2 months leave since the 2nd week of December. And
that was switching to a whole new industry for me.
10:01 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 16 June 2013 00:32, Dave McGuire <mcguire at neurotica.com> wrote:
Umm.
I'm sorry you're going through that. I wish you better luck in the
future. (seriously)
You could come here, but you might value freedom and privacy, which we
don't really "do" anymore.
It's been the case since the beginning of the credit crunch. Almost
all the IT journos I know personally are out of work; a few hang on
doing forum moderation or writing for ?10/article.
As for techies, well, things are slightly better than they were 2-3Y
ago. Around 2009-2010, pretty much everyone I knew was out of work
except the actual genius-level guys or the handful who hang on in the
City (UK-speak for "Wall St.") Now, I guess only about half to
two-thirds of the IT people I know are out of work; basically, only
the young, the very keen, the very talented, or those who are still in
positions that are >5Y old, indeed, 10+Y old, are still employed.
/Good/ Linux & FOSS techies are OK. Windows people are fungible and
disposable now. Sadly, I am mainly a Windows guy, with sidelines in
Mac and Linux, but I'm not a specialist. I don't know Perl or any
other scripting language, I don't do regular expressions, I'm not a
programmer or a devop -- primarily, I was a desktop support guy. I'm
good at that, but that business is more or less a minimum-wage
helldesk job now, staffed mainly from remote places, via call queues,
under immense pressure. I've done a bit of it; it's /horrid./ I was a
2nd/3rd level desktop support guy, but you can't get into those roles
now without multiple MSCEs etc., which I don't have and can't afford
to do.
Anyway, I'm tired of "reskilling". I've been doing it for 25Y. I
started on VAX-VMS and CP/M and have switched to new OSes & new sets
of apps about 15 times now. I'm sick of it. Problems that were sorted
and had gone away are re-appearing, together with vast realms of new
complexity, coupled with an attitude of "if it works, leave it", "if
it doesn't work, reimage it or deploy a new VM and hope for the best".
There's little real deep troubleshooting any more and nobody has the
breadth of coverage I used to do. It's all nuke and reload from orbit.
If that doesn't work, give it more RAM or wait for a new version.
Enough. Screw IT. Bored now, as "Bad Willow" said.
I'm currently waiting to hear about a technical-writer position in
Berlin in Germany and if that fails I've applied for English-teaching
positions in both England and Japan.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
16 Jun
16 Jun
3:18 a.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 16/06/2013 00:32, Dave McGuire wrote:
On 06/14/2013 10:32 PM, Liam Proven wrote:
Umm.
I'm sorry you're going through that. I wish you better luck in the
future. (seriously)
You could come here, but you might value freedom and privacy, which
we don't really "do" anymore.
We don't seem to do that over here (in the UK) either........
...then one needs to get a job working for decent people. These
days, at
least in the USA, beggars can't be choosers, sure, but there are
GOOD jobs
around. If I can find one, anyone can.
Good for you.
Have I mentioned recently the roughly *3,000 job applications* I've
made around the world since the end of my last F/T role in 2008?
Those are all for positions I'm well-qualified for with abundant
experience, natch. 25Y in IT, I've supported more than a dozen OSes
professionally; I know every OS, server and desktop, in current
business use, and most of the big apps.
I average 2-3 interviews a year, if that.
It may be only moderately bad in the USA. Elsewhere in the world, it's
/terrible./ I believe there are roughly 20-25? more people looking for
work than jobs in the UK currently; the /average/ number of applicants
is around 18-23 per job, depending who you ask, and that includes the
really crappy ones.
In professional roles, it ranges from around 65-70 applicants per job
to over 900 per job for desirable positions; I have head of jobs
getting several thousand applicants as a matter of routine.
The last 2 positions I landed lasted 8 days and 6 weeks, respectively.
I am still on 2 months' unpaid leave from the last one (end 2012) -
I've been on those 2 months leave since the 2nd week of December. And
that was switching to a whole new industry for me.
-Dave
I also sympathise with Liam. Getting an IT based job in the UK is hard,
and in the North of the UK where I live I would say very hard. I spent
two years teaching Maths and never enjoyed it. I managed to sneak my way
back into IT via the back door. However I know of many folk who are
struggling to find work. I would really like to work for a few more
years yet (I am 59) but I shall be leaving my current post in just over
a year on enhanced terms but as its part of government cut backs I
won't be replaced. And whilst panic hasn't set in yet the folks who will
be left behind are wondering how they will cope...
Sadly I had hoped to spend more time at the Museum of Science and
Industry in Manchester (www.mosi.org.uk) but despite the petitions I
expect that to close soon...
http://www.manchestereveningnews.co.uk/news/greater-manchester-news/mosi-30…
http://www.change.org/en-GB/petitions/science-museum-group-save-the-museum-…
So if I could get a VISA moving to the US does have its attractions........
Dave
G4UGM
14 Jun
14 Jun
9:23 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 15 June 2013 00:14, Peter Corlett <abuse at cabal.org.uk> wrote:
BeOS was sort of a single user Unix: it was
sufficiently POSIXy for me to fire
up a shell prompt and feel at home, although it did not support multiple users
at all.
For my money -- as someone whose hobby is OSes, basically, and who has
played with as many as I can lay hands upon for about 30y now -- BeOS
is the most important OS for home/personal computers in the last few
decades.
Which is, naturally, why it's obscure, dead & forgotten.
It had many of the best aspects of xNix -- it was POSIX-like, had a
familiar shell, etc. -- but was free of all the decades of cruft
around Unix. I mean, Mac OS X is a gorgeous OS, but it's huge, vastly
complex, not very flexible or customisable, and it's only quick
because it runs on massively powerful hardware.
I don't need a network-transparent GUI. I'm sure it's a great thing,
but in some 20y of using systems with X.11 available, I've /never/
needed that.
I don't need a multiuser PC. It's mine. I'm the only user. I only need
multiuser servers. It's great but it's baggage. Security, sure, some,
but not a minicomputer/server OS on my desktop.
I don't need paging/swapping any more either, TBH.
I don't need 80% of everything that Unix or indeed Linux gives me, in
fact. Don't need it, don't want it.
I was an Archimedes user, and I'm nostalgic for it, but while I'd like
RISC OS on a Raspberry Pi as a toy, I don't /miss/ it. I miss some of
its features, but not the OS.
I miss some of the features of classic MacOS.
I have a QL, Amiga and ST, and they all had strengths and good aspects
-- but I don't miss anything from them, partly because I was never a
full-time user of any of them. I got them all late, as a collector.
Sure, it was impressive to see how much AmigaOS could do in 512kB of
RAM on a pair of floppy drives, but it's 2013 -- I don't need that
now. I don't want the constraints. The QNX demo floppy was impressive
too, come to that.
I miss my ZX Spectrum, but I'm glad I'm not using one any more!
I want a solid, multitasking, multithreading OS, running on lots of
cores, delivering a fast, responsive experience. I want rich media. I
want productivity apps, but Linux and C21 FOSS delivers all I want and
more, really. I don't need anything proprietary. I want flexible
networking and plug-and-play hardware. I don't need a huge fat
dual-slot GPU -- I just want 2 or 3 smallish cheap monitors.
BeOS delivered all I really wanted /out of the OS/ and the only
problems would have been resolved if there had been LibreOffice,
Firefox, Thunderbird, Pidgin & a few other modern Linux apps for it.
(And perhaps a hypervisor. Back then, DOS boxes would have done.)
/And it *didn't* have a couple of gigabytes of cruft I *didn't* need
and don't need now./
That was why it was lovely.
The snappy performance but cheap standard components of an ST, the
multitasking power of an Amiga, the responsiveness of an Archimedes --
and all on commodity PC hardware. My PC never felt as powerful, before
or since. My Core 2 Quad Extreme with Ubuntu is less responsive than a
133MHz Pentium-1 was under BeOS.
I really hope that Haiku one day delivers on that promise again.
Linux is a great server OS, but it's not a great desktop. Never was.
It's better than it used to be but it's still not great. It's good
enough, no more.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
10:01 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 06/14/2013 09:23 PM, Liam Proven wrote:
I don't need a network-transparent GUI. I'm
sure it's a great thing,
but in some 20y of using systems with X.11 available, I've /never/
needed that.
But don't assume that's true of everyone. I've used it DAILY, all
day, every day, no joke, since about 1991. My workflows for nearly
everything would be severely impacted if that ability were to suddenly
go away. (which, thankfully, it won't)
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
10:05 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, 15 Jun 2013, Dave McGuire wrote:
On 06/14/2013 09:23 PM, Liam Proven wrote:
That is one really neat feature of X11 that I've always liked.
I don't need a network-transparent GUI.
I'm sure it's a great thing,
but in some 20y of using systems with X.11 available, I've /never/
needed that.
But don't assume that's true of everyone. I've used it DAILY, all
day, every day, no joke, since about 1991. My workflows for nearly
everything would be severely impacted if that ability were to suddenly
go away. (which, thankfully, it won't)
-Dave
--
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
10:07 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 06/14/2013 10:05 PM, Cory Smelosky wrote:
That is one really neat feature of X11 that I've always liked.
And that's the thing. It's not a "feature", nor is it some
"copying"
scheme layered atop a windowing system. (like RDP, VNC, etc) It is
*central to the design of the windowing system*. It ALWAYS makes a
network/socket connection to put a window up somewhere. It just happens
to be on localhost much of the time.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
I don't need a network-transparent GUI. I'm
sure it's a great thing,
but in some 20y of using systems with X.11 available, I've /never/
needed that.
But don't assume that's true of everyone. I've used it DAILY, all
day, every day, no joke, since about 1991. My workflows for nearly
everything would be severely impacted if that ability were to suddenly
go away. (which, thankfully, it won't)
10:44 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, 15 Jun 2013, Dave McGuire wrote:
On 06/14/2013 10:05 PM, Cory Smelosky wrote:
That is one really neat feature of X11 that I've always liked.
And that's the thing. It's not a "feature", nor is it some
"copying"
scheme layered atop a windowing system. (like RDP, VNC, etc) It is
*central to the design of the windowing system*. It ALWAYS makes a
network/socket connection to put a window up somewhere. It just happens
to be on localhost much of the time.
Errr, right. But I have a bit of a different definition of "feature". ;)
I often consider a system booting to be a feature. ;)
I don't need a network-transparent GUI. I'm
sure it's a great thing,
but in some 20y of using systems with X.11 available, I've /never/
needed that.
But don't assume that's true of everyone. I've used it DAILY, all
day, every day, no joke, since about 1991. My workflows for nearly
everything would be severely impacted if that ability were to suddenly
go away. (which, thankfully, it won't) -Dave
--
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
10:51 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 06/14/2013 10:44 PM, Cory Smelosky wrote:
And that's the thing. It's not a "feature", nor is it some
"copying"
scheme layered atop a windowing system. (like RDP, VNC, etc) It is
*central to the design of the windowing system*. It ALWAYS makes a
network/socket connection to put a window up somewhere. It just happens
to be on localhost much of the time.
Errr, right. But I have a bit of a different definition of "feature". ;)
I often consider a system booting to be a feature. ;)
Knowing your fascination with NT, I'm not surprised. ;)
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
> I
don't need a network-transparent GUI. I'm sure it's a great thing,
> but in some 20y of using systems with X.11 available, I've /never/
> needed that.
But don't assume that's true of everyone. I've used it DAILY, all
day, every day, no joke, since about 1991. My workflows for nearly
everything would be severely impacted if that ability were to suddenly
go away. (which, thankfully, it won't)
That is one really neat feature of X11 that I've always liked.
15 Jun
15 Jun
12:59 a.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, 15 Jun 2013, Dave McGuire wrote:
On 06/14/2013 10:44 PM, Cory Smelosky wrote:
;) NT 4 is a solid...well...it's a solid NT 4. Let's stick with that.
I often consider a system booting to be a
feature. ;)
Knowing your fascination with NT, I'm not surprised. ;)
-Dave
--
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
14 Jun
14 Jun
10:14 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 15 June 2013 03:01, Dave McGuire <mcguire at neurotica.com> wrote:
But don't assume that's true of everyone.
I've used it DAILY, all day,
every day, no joke, since about 1991. My workflows for nearly everything
would be severely impacted if that ability were to suddenly go away. (which,
thankfully, it won't)
That's fine. Long may it be available to you, but for my money, for
Linux on *personal* computers -- which soon enough is going to mean
tablet-type things -- the sooner X.11 disappears, the better. Roll on
Wayland or Mir or whatever.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
10:51 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, 15 Jun 2013, Liam Proven wrote:
--
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
On 15 June 2013 03:01, Dave McGuire <mcguire at neurotica.com> wrote:
X is not suited for the personal "KILL ALL THE BLUE PEOPLE" computer. ;)
But don't assume that's true of
everyone. I've used it DAILY, all day,
every day, no joke, since about 1991. My workflows for nearly everything
would be severely impacted if that ability were to suddenly go away. (which,
thankfully, it won't)
That's fine. Long may it be available to you, but for my money, for
Linux on *personal* computers -- which soon enough is going to mean
tablet-type things -- the sooner X.11 disappears, the better. Roll on
Wayland or Mir or whatever.
15 Jun
15 Jun
12:40 a.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Sat, 15 Jun 2013, Cory Smelosky wrote:
Now wait just a minute. You mean someone has a purpose-built, Smurf
killing computer? I want one. RIGHT. NOW.
:)
g.
--
Proud owner of F-15C 80-0007
http://www.f15sim.com - The only one of its kind.
http://www.diy-cockpits.org/coll - Go Collimated or Go Home.
Some people collect things for a hobby. Geeks collect hobbies.
ScarletDME - The red hot Data Management Environment
A Multi-Value database for the masses, not the classes.
http://www.scarletdme.org - Get it _today_!
That's
fine. Long may it be available to you, but for my money, for
Linux on *personal* computers -- which soon enough is going to mean
tablet-type things -- the sooner X.11 disappears, the better. Roll on
Wayland or Mir or whatever.
X is not suited for the personal "KILL ALL THE BLUE PEOPLE" computer. ;)
6:16 a.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
Now wait just a minute. You mean someone has a
purpose-built, Smurf
killing computer? I want one. RIGHT. NOW.
Yep, it is called Cruel (the cat of Gargamel)
:o)
14 Jun
14 Jun
10:36 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
I don't
need a network-transparent GUI.
But don't assume that's true of everyone.
I've used it DAILY, all
day, every day, no joke, since about 1991. 
20 Jun
20 Jun
7:18 a.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, Jun 14, 2013 at 10:36:21PM -0400, Mouse wrote:
Someone was recently telling me about some people trying to design a
successor to X. I was interested (X really does suck in a few
respects),
Yes, that sounds good so far - there are certainly enough warts in X.
until I heard it was abandoning the networkability
aspect;
that alone makes it not worth even finding out more about, to me.
And that kills it. Non network aware replacement for X? Not interested
at all, may your idea die a quick death.
Kind regards,
Alex.
--
"Opportunity is missed by most people because it is dressed in overalls and
looks like work." -- Thomas A. Edison
10:25 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 20 June 2013 12:18, Alexander Schreiber <als at thangorodrim.de> wrote:
On Fri, Jun 14, 2013 at 10:36:21PM -0400, Mouse
wrote:
Too late!
http://www.omgubuntu.co.uk/2013/03/ubuntus-new-display-server-mir-gets-demo…
Also:
http://wayland.freedesktop.org/
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
Someone was recently telling me about some people trying to design a
successor to X. I was interested (X really does suck in a few
respects),
Yes, that sounds good so far - there are certainly enough warts in X.
until I heard it was abandoning the
networkability aspect;
that alone makes it not worth even finding out more about, to me.
And that kills it. Non network aware replacement for X? Not interested
at all, may your idea die a quick death.
14 Jun
14 Jun
10:17 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 6/14/2013 7:23 PM, Liam Proven wrote:
Linux is a great server OS, but it's not a great desktop. Never was.
It's better than it used to be but it's still not great. It's good
enough, no more.
I never could get X-Windows to work well when I played with Linux.
Have they fixed the apps to work yet with PC screen formats.
Ben.
PS. I play the odd game so I need a PC.
11:02 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 15 June 2013 03:17, ben <bfranchuk at jetnet.ab.ca> wrote:
I never could get X-Windows to work well when I played
with Linux.
Have they fixed the apps to work yet with PC screen formats.
I don't know what that means.
If it helps, I have a fast, fluid, 3D-composited, accelerated desktop
across 2 19" LCDs running at 3840*1024 from an old nVidia GeForce 230
here. Everything works and it plays Portal nicely, too.
PS. I play the odd game so I need a PC.
I always point people at a modern console for an easy life. Failing
that, dual-boot Windows.
But more & more Linux games are coming. Steam runs natively now.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
15 Jun
15 Jun
1:23 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 6/14/2013 9:02 PM, Liam Proven wrote:
Back to X-windows, Just what was the typical Unix screen size and depth
for about 1990's.
On 15 June 2013 03:17, ben <bfranchuk at
jetnet.ab.ca> wrote:
That is good. As Mr Weak Eyes here, I have trouble with 800 x 600 screens.
I never could get X-Windows to work well when I
played with Linux.
Have they fixed the apps to work yet with PC screen formats.
I don't know what that means.
If it helps, I have a fast, fluid, 3D-composited, accelerated desktop
across 2 19" LCDs running at 3840*1024 from an old nVidia GeForce 230
here. Everything works and it plays Portal nicely, too.
PS. I play the
odd game so I need a PC.
I always point people at a modern console for an easy life. Failing
that, dual-boot Windows.
But more & more Linux games are coming. Steam runs
natively now.
I play the games I like, not what sells. Windows is good for 'click'
for your favorite item and games come under that heading. I have 'Cards'
for Solitaire , the better version. Linux is free so dirt cheap for
the new apps; that does not make Linux better.
Ben.
6:06 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
It was thus said that the Great ben once stated:
On 6/14/2013 9:02 PM, Liam Proven wrote:
Hey, don't knock those high resolution screens. I'm sitting in front of a
24" 1920x1200 LCD screen and I've got a 86x23 character terminal open that
you can see clear across the room (it's something like a 48pt font).
-spc (But normally I use a font size that gives me 315x81 character
terminal---I like seeing *lots* of text ... )
On 15 June 2013 03:17, ben <bfranchuk at
jetnet.ab.ca> wrote:
That is good. As Mr Weak Eyes here, I have trouble with 800 x 600 screens. I never could get X-Windows to work well when I
played with Linux.
Have they fixed the apps to work yet with PC screen formats.
I don't know what that means.
If it helps, I have a fast, fluid, 3D-composited, accelerated desktop
across 2 19" LCDs running at 3840*1024 from an old nVidia GeForce 230
here. Everything works and it plays Portal nicely, too.
3:42 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 06/15/2013 06:06 AM, Sean Conner wrote:
It was thus said that the Great ben once stated:
That's the only way to do it. :)
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
On 6/14/2013 9:02 PM, Liam Proven wrote:
Hey, don't knock those high resolution screens. I'm sitting in front of a
24" 1920x1200 LCD screen and I've got a 86x23 character terminal open that
you can see clear across the room (it's something like a 48pt font).
-spc (But normally I use a font size that gives me 315x81 character
terminal---I like seeing *lots* of text ... ) On 15 June 2013 03:17, ben <bfranchuk at
jetnet.ab.ca> wrote:
That is good. As Mr Weak Eyes here, I have trouble with 800 x 600 screens. I never could get X-Windows to work well when I
played with Linux.
Have they fixed the apps to work yet with PC screen formats.
I don't know what that means.
If it helps, I have a fast, fluid, 3D-composited, accelerated desktop
across 2 19" LCDs running at 3840*1024 from an old nVidia GeForce 230
here. Everything works and it plays Portal nicely, too.
9:31 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 15 June 2013 06:23, ben <bfranchuk at jetnet.ab.ca> wrote:
Back to X-windows, Just what was the typical Unix screen size and depth for
about 1990's.
I want to cavil over "X-windows" but sod it.
It depends when in the 1990s!
Personally, with Lasermoon Linux-FT on my work desktop in about 1996,
I started off with a 17" CRT at around 1024?768 or 1152?864. It could
do 1280?1024 and it looked snazzy and "power-user" like, with lots of
tiny windows, but it was too small to comfortably read.
When I went freelance later that year, my home system started off with
twin 14" screens, I think, and I gradually went to twin 15" then twin
17". IIRC, my preference was 1152?864 or thereabouts - an obscure res,
but widely-supported. My eyes don't mind low refresh rates and most
17" CRTs could do 1152?864 at about 70-75Hz, which is good enough for
me.
By early C21, I was using twin 19" CRTs which is about when I switched
to 1280?1024, at first on two separate Matrox cards, and later on a
series of single cards with dual outputs. I switched to nVidia by
about 2005-2006 or so.
I don't actually understand your point here.
I only have one Steam game - Portal. It's the first new game in about
6-7y to appeal to me. The last one was Guitar Hero. Before that, Quake
III Arena; before that, Quake 1, and before that, Doom. That takes us
back to nearly 20y ago.
It took me a few years to get my own copy of Portal - I play so little
that it's a very low priority for me. Portal itself is about 6Y old
now.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
On 6/14/2013 9:02 PM, Liam Proven wrote:
Firstly, I made a mistake - my desktop is 2560?1024, not 3480 across.
That is, in other words, 2 19" monitors running at 1280?1024. For
2013, that is really low-resolution. These are roughly decade-old
TFTs, given to me for nothing by a client.
I also tend to zoom in the text in my web browsers, terminals and
wordprocessor windows. I am quite short-sighted myself (approximately
-6 dioptre); I need glasses to use a computer.
The client who gave me these screens now has a 24" Thunderbolt display
on his MacBook Pro. I have dropped the resolution on that to about
1280?768 or something for him, so that text is large, bold and fairly
easy to read. (He's in his late 70s.)
In summary: there are two ways to go. Me, I favour using old,
trailing-edge hardware -- it's part of the reason that I'm here -- and
I find that older, lower-res screens are easier on my eyes. It rankles
with me to run a modern hi-res display at much under its native
resolution - it seems a waste and you lose the sharpness of a digital
TFT display.
OTOH, with devices like the Macbook Pro with "retina display", it is
normal to run it at very high resolutions with /huuuuuge/ fonts, for
pin-sharp text. Few people have the eyesight to use it at its native
res with normal-size fonts and widgets designed for a 72dpi screen.
As such, modern GUIs designed to cope with high-DPI displays, with
scalable screen furniture, cope well with arbitrary, non-native
resolutions. I know a number of people with low vision or other
eyesight problems and they seem to like modern very-high-DPI displays,
because they can scale stuff and it remains sharp.
Scaled displays on older, low-DPI flatscreens look blurred and fuzzy
and are harder on weak eyes or poor vision.
When 15" laptops started going towards 1400+ pixels across the screen,
my eyes gave up the battle. I could no longer comfortably read
native-size text. That's when I started using GUI-scaling features.
So I would have to argue that of the two routes I suggest - either use
displays from before when resolutions exceeded your personal comfort
level, or, bite the bullet, make the jump to 150-200+ dpi screens,
scale the GUI and enjoy - the latter is significantly better if you
have real problems with your vision.
On 15 June 2013 03:17, ben <bfranchuk at jetnet.ab.ca> wrote:
That is good. As Mr Weak Eyes here, I have trouble with 800 x 600 screens.
I never could get X-Windows to work well when I played with Linux.
Have they fixed the apps to work yet with PC screen formats.
I don't know what that means.
If it helps, I have a fast, fluid, 3D-composited, accelerated desktop
across 2 19" LCDs running at 3840*1024 from an old nVidia GeForce 230
here. Everything works and it plays Portal nicely, too. PS. I play the odd game so I need a PC.
I always point people at a modern console for an easy life. Failing
that, dual-boot Windows.
But more &
more Linux games are coming. Steam runs natively now.
I play the games I like, not what sells. Windows is good for 'click'
for your favorite item and games come under that heading. I have 'Cards'
for Solitaire , the better version. Linux is free so dirt cheap for
the new apps; that does not make Linux better. 
11:26 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of
Back to X-windows, Just what was the typical Unix
screen size and depth
for about 1990's.
Don't know about typical, but the Sun lab we set up in 1991 had mostly
CG3 cards with a couple of CG12s. IIRC, they were 1152x900x8 and
1280x1024x24, respectively. And I remember those Trinitron monitors
as being quite nice. Speaking of which, I've got a few (of the
monitors) I wouldn't mind unloading.
BLS
3:42 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 06/15/2013 01:23 AM, ben wrote:
That is good. As Mr Weak Eyes here, I have trouble with 800 x 600 screens.
I would too! I get eye fatigue at anything less than about 1280x1024.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
I never could get X-Windows to work well when I played
with Linux.
Have they fixed the apps to work yet with PC screen formats.
I don't know what that means.
If it helps, I have a fast, fluid, 3D-composited, accelerated desktop
across 2 19" LCDs running at 3840*1024 from an old nVidia GeForce 230
here. Everything works and it plays Portal nicely, too.
4:17 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 15 June 2013 20:42, Dave McGuire <mcguire at neurotica.com> wrote:
I would too! I get eye fatigue at anything less
than about 1280x1024.
Um. Are you joking, or did you miss the point here?
If you're kidding, um, well, er, hey, I am not one to attack poor
taste humour generally, but, er... well...
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
5:19 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 06/15/2013 04:17 PM, Liam Proven wrote:
I didn't miss the point. (at least I don't think I did!)
Ben is stuck in Windows land. Windows users are accustomed to reducing the
resolution of the graphics system in order to get larger text. This has, of
course, been ludicrous since day one. Just increase the friggin' font size.
Apparently that's "hard" in Windows land.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
I would too!
I get eye fatigue at anything less than about 1280x1024.
Um. Are you joking, or did you miss the point here?
If you're kidding, um, well, er, hey, I am not one to attack poor
taste humour generally, but, er... well...
6:26 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 15/06/2013 22:19, Dave McGuire wrote:
On 06/15/2013 04:17 PM, Liam Proven wrote:
I didn't miss the point. (at
least I don't think I did!)
Ben is stuck in Windows land. Windows users are accustomed to reducing the
resolution of the graphics system in order to get larger text. This has, of
course, been ludicrous since day one. Just increase the friggin' font size.
Apparently that's "hard" in Windows land.
Oddly that takes us round a full circle. Whilst it is easy to change the
font sizes in Windoze, The problem is that many applications don't cope
well when you do it.....and of course it requires a re-boot
So I guess Devs should test on slow machines and with a range of font
sizes....
I
would too! I get eye fatigue at anything less than about 1280x1024.
Um. Are you
joking, or did you miss the point here?
If you're kidding, um, well, er, hey, I am not one to attack poor
taste humour generally, but, er... well... -Dave
Dave...
7:07 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Sat, 15 Jun 2013, Dave wrote:
Oddly that takes us round a full circle. Whilst it is
easy to change the font
sizes in Windoze, The problem is that many applications don't cope well when
you do it.....and of course it requires a re-boot
Actually, you change the scaling factor, not the font size. Doesn't
require a reboot either.
g.
--
Proud owner of F-15C 80-0007
http://www.f15sim.com - The only one of its kind.
http://www.diy-cockpits.org/coll - Go Collimated or Go Home.
Some people collect things for a hobby. Geeks collect hobbies.
ScarletDME - The red hot Data Management Environment
A Multi-Value database for the masses, not the classes.
http://www.scarletdme.org - Get it _today_!
7:31 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 06/15/2013 06:26 PM, Dave wrote:
I didn't miss the point. (at
least I don't think I did!)
Ben is stuck in Windows land. Windows users are accustomed to
reducing the
resolution of the graphics system in order to get larger text. This
has, of
course, been ludicrous since day one. Just increase the friggin' font
size.
Apparently that's "hard" in Windows land.
Oddly that takes us round a full circle. Whilst it is easy to change the
font sizes in Windoze, The problem is that many applications don't cope
well when you do it.....and of course it requires a re-boot
So I guess Devs should test on slow machines and with a range of font
sizes....
ROFL!!
--
Dave McGuire, AK4HZ
New Kensington, PA
I would
too! I get eye fatigue at anything less than about
1280x1024.
Um. Are you joking, or did you miss the point here?
If you're kidding, um, well, er, hey, I am not one to attack poor
taste humour generally, but, er... well...
6:32 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 15/06/2013 22:19, Dave McGuire wrote:
On 06/15/2013 04:17 PM, Liam Proven wrote:
I didn't miss the point. (at
least I don't think I did!)
Ben is stuck in Windows land. Windows users are accustomed to reducing the
resolution of the graphics system in order to get larger text. This has, of
course, been ludicrous since day one. Just increase the friggin' font size.
Apparently that's "hard" in Windows land.
-Dave
Tried it on this Windows/7 box and you no longer need to re-boot. You do
have to log out and it though so the session can restart...
I
would too! I get eye fatigue at anything less than about 1280x1024.
Um. Are you
joking, or did you miss the point here?
If you're kidding, um, well, er, hey, I am not one to attack poor
taste humour generally, but, er... well...
6:32 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 6/15/2013 3:19 PM, Dave McGuire wrote:
On 06/15/2013 04:17 PM, Liam Proven wrote:
I didn't miss the point. (at least I don't think I did!)
Ben is stuck in Windows land. Windows users are accustomed to reducing the
resolution of the graphics system in order to get larger text. This has, of
course, been ludicrous since day one. Just increase the friggin' font size.
Apparently that's "hard" in Windows land.
Why? if I set the screen font size to what I can see, I am back to the
same screen
size as before. I have small computer desk, so getting a bigger screen
is not a option. I just grumble that *24x80* text size (full monitor)
are not a Windows option under windows. If had good text emulation then
I would be using VIM more often for editing. Is there a way to set the
fonts for that option under windows rather than dancing toasters.
I
would too! I get eye fatigue at anything less than about 1280x1024.
Um. Are you joking, or did you miss the point here?
If you're kidding, um, well, er, hey, I am not one to attack poor
taste humour generally, but, er... well... -Dave
I love how True-type fonts no longer seem to work for the video display
and printer.
Ben.
7:30 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 06/15/2013 06:32 PM, ben wrote:
On 6/15/2013 3:19 PM, Dave McGuire wrote:
Why what?
On 06/15/2013 04:17 PM, Liam Proven wrote:
I didn't miss the point. (at least I don't think I did!)
Ben is stuck in Windows land. Windows users are accustomed to
reducing the
resolution of the graphics system in order to get larger text. This
has, of
course, been ludicrous since day one. Just increase the friggin' font
size.
Apparently that's "hard" in Windows land.
Why? I
would too! I get eye fatigue at anything less than about
1280x1024.
Um. Are you joking, or did you miss the point here?
If you're kidding, um, well, er, hey, I am not one to attack poor
taste humour generally, but, er... well... if I set the screen font size to what I can see, I am
back to the
same screen
size as before. I have small computer desk, so getting a bigger screen
is not a option. I just grumble that *24x80* text size (full monitor)
are not a Windows option under windows. If had good text emulation then
I would be using VIM more often for editing. Is there a way to set the
fonts for that option under windows rather than dancing toasters.
My point was, the rest of the world doesn't do this insane switching
of video modes resolutions crap that the PC world still does. A
graphics subsystem should run at its highest resolution. Beyond a
what-should-be-impenetrable abstraction layer is font rendering. The
latter is what determines how big the characters are on your screen.
Anything less is a design limitation and a bug that needs to be fixed.
IMnsHO.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
9:50 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 15 June 2013 22:19, Dave McGuire <mcguire at neurotica.com> wrote:
I didn't miss the point. (at least I don't
think I did!)
Ben is stuck in Windows land.
What indicates that?
Windows users are accustomed to reducing the
resolution of the graphics system in order to get larger text.
Only incompetent ones. Windows has had on-screen feature scaling since
the 3.1 days and built-in to the GDI since NT 3.
This has, of
course, been ludicrous since day one. Just increase the friggin' font size.
Apparently that's "hard" in Windows land.
No, it isn't, not at all. Hasn't been since 1993 or so.
I think that what Ben was saying was that he had low vision or other
serious eyesight problems and that he therefore cannot read small
text. You appeared to be mocking his disability.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
10:14 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 06/15/2013 09:50 PM, Liam Proven wrote:
Nothing from his post, but I've been talking to Ben off and on for about
ten years, and I know that he runs Windows.
That figures. ;)
Not at all; that's the last thing I'd do. Ben appeared to be saying that
he needs larger text (which many people do) and got it by reducing his
display resolution. I may have misinterpreted what he said, and would be
happy to stand corrected. (Ben?)
However, that exact scenario is something I've known Windows people to do a
lot, for a long time now. I'm pleased to learn that it isn't actually
*necessary*.
-Dave
--
Dave McGuire, AK4HZ
New Kensington, PA
I didn't
miss the point. (at least I don't think I did!)
Ben is stuck in Windows land.
What indicates that? Windows users
are accustomed to reducing the
resolution of the graphics system in order to get larger text.
Only incompetent ones. Windows has had on-screen feature scaling since
the 3.1 days and built-in to the GDI since NT 3. This has, of
course, been ludicrous since day one. Just increase the friggin' font size.
Apparently that's "hard" in Windows land.
No, it isn't, not at all. Hasn't been since 1993 or so.
I think that what Ben was saying was that he had low vision or other
serious eyesight problems and that he therefore cannot read small
text. You appeared to be mocking his disability.
11:08 p.m.
New subject: [cctalk] Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
> I didn't miss the point. (at least I
don't think I did!)
> Ben is stuck in Windows land.
On Sun, 16 Jun 2013, Liam Proven wrote:
What indicates that?
You would think so.
On the [Windoze XP] machines at the college, with most adjustments LOCKED
OUT, the default screen resolution did not have a suitable sized font.
BUT, if I dropped it down a couple of notches, to 1280x1024, I was able to
get a font size that I could stand.
Windows users are accustomed to reducing the
resolution of the graphics system in order to get larger text.
Only incompetent
ones. Windows has had on-screen feature scaling since
the 3.1 days and built-in to the GDI since NT 3. No, it isn't, not at all. Hasn't been since
1993 or so.
But, the college managed to resurrect the problem.
I think that what Ben was saying was that he had low
vision or other
serious eyesight problems and that he therefore cannot read small
text. You appeared to be mocking his disability.
25 years ago, I could [and did] read microfiche without a viewer.
For my personal use, I created my own LaserJet font with 9x12 pixel
characters (and one with 7xs9 and one with 5x7). I loved my
Wyse700/Amdek1280 monitors with 1280x800 pixels that could give me
160 characters x 50 lines text modes.
Presbyopia hit me. HARD.
Now, I can not even read a newspaper without >+2 reading glasses,
+2 for computer screen, and >+3 for OQO and/or 7" Android tablet.
But, worst of all, I ended up constantly battling with my cow-
orkers over screen resolution on shared computers - I had to drop
the display to 1280x1024 in order to get access to a suitably
sized font.
IN THEORY, I should have had access to larger fonts without
changing the display resolution. Would defenestrating some
college administrators have helped? Would it have helped with
font size?
--
Grumpy Ol' Fred cisin at xenosoft.com
14 Jun
14 Jun
11:04 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, 15 Jun 2013, ben wrote:
On 6/14/2013 7:23 PM, Liam Proven wrote:
I've never had it work well for very long. I've decided I'm done with
modern PC hardware...I'm gonna go to a VAXstation or an Indy or similar.
;)
Linux is a great server OS, but it's not a great desktop. Never was.
It's better than it used to be but it's still not great. It's good
enough, no more.
I never could get X-Windows to work well when I played with Linux.
Have they fixed the apps to work yet with PC screen formats.
Ben.
PS. I play the odd game so I need a PC.
--
--
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects
20 Jun
20 Jun
7:21 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, Jun 14, 2013 at 08:17:13PM -0600, ben wrote:
On 6/14/2013 7:23 PM, Liam Proven wrote:
Just Works (TM) here:
- at home: fast, 3D accelerated 1920x1080 desktop
- at work: fast, 3D accelerated 3830x1200 desktop over two displays
The last time I remember actually having to fight to get X to work on
Linux was back before 2000, when one still had to fiddle with modelines
under certain circumstances.
Kind regards,
Alex.
--
"Opportunity is missed by most people because it is dressed in overalls and
looks like work." -- Thomas A. Edison
Linux is a great server OS, but it's not a great desktop. Never was.
It's better than it used to be but it's still not great. It's good
enough, no more.
I never could get X-Windows to work well when I played with Linux.
Have they fixed the apps to work yet with PC screen formats.
10:13 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 20 June 2013 12:21, Alexander Schreiber <als at thangorodrim.de> wrote:
Just Works (TM) here:
- at home: fast, 3D accelerated 1920x1080 desktop
- at work: fast, 3D accelerated 3830x1200 desktop over two displays
So if all you use is xterms, what's the point? :?)
The last time I remember actually having to fight to
get X to work on
Linux was back before 2000, when one still had to fiddle with modelines
under certain circumstances.
Agreed. Such problems are usually history now.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
22 Jun
22 Jun
9:28 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, Jun 21, 2013 at 03:13:35AM +0100, Liam Proven wrote:
On 20 June 2013 12:21, Alexander Schreiber <als at
thangorodrim.de> wrote:
Well, I do occasionally fire up Google Earth, which is the number one
program I use that really _needs_ decent 3D acceleration. ;-)
Kind regards,
Alex.
--
"Opportunity is missed by most people because it is dressed in overalls and
looks like work." -- Thomas A. Edison
Just Works (TM) here:
- at home: fast, 3D accelerated 1920x1080 desktop
- at work: fast, 3D accelerated 3830x1200 desktop over two displays
So if all you use is xterms, what's the point? :?)
20 Jun
20 Jun
7:16 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Sat, Jun 15, 2013 at 02:23:44AM +0100, Liam Proven wrote:
On 15 June 2013 00:14, Peter Corlett <abuse at
cabal.org.uk> wrote:
I use that regularly. It is an incredibly useful detail of the way the
X Window system is designed. So you can, e.g. run the process on the
machine where the big pile of data lives and have it displayed on your
workstation. Or run the process on the workstation inside the corporate
security perimeter, but have the display on your laptop in the hotel,
with the connection piped back through ssh.
BeOS was sort of a single user Unix: it was
sufficiently POSIXy for me to fire
up a shell prompt and feel at home, although it did not support multiple users
at all.
For my money -- as someone whose hobby is OSes, basically, and who has
played with as many as I can lay hands upon for about 30y now -- BeOS
is the most important OS for home/personal computers in the last few
decades.
Which is, naturally, why it's obscure, dead & forgotten.
It had many of the best aspects of xNix -- it was POSIX-like, had a
familiar shell, etc. -- but was free of all the decades of cruft
around Unix. I mean, Mac OS X is a gorgeous OS, but it's huge, vastly
complex, not very flexible or customisable, and it's only quick
because it runs on massively powerful hardware.
I don't need a network-transparent GUI. I'm sure it's a great thing,
but in some 20y of using systems with X.11 available, I've /never/
needed that. I don't need a multiuser PC. It's mine.
I'm the only user. I only need
multiuser servers. It's great but it's baggage. Security, sure, some,
but not a minicomputer/server OS on my desktop.
Again, very useful: allows separating access privileges. Quite a bit
of the stuff even on my workstation needs neither full system access
(i.e. root) nor access to my user data.
I want a solid, multitasking, multithreading OS,
running on lots of
cores, delivering a fast, responsive experience. I want rich media. I
want productivity apps, but Linux and C21 FOSS delivers all I want and
more, really. I don't need anything proprietary. I want flexible
networking and plug-and-play hardware. I don't need a huge fat
dual-slot GPU -- I just want 2 or 3 smallish cheap monitors.
Linux is a great server OS, but it's not a great desktop. Never was.
It's better than it used to be but it's still not great. It's good
enough, no more.
Works for and gives me all I need. But then my idea of a "graphical desktop
environment" has been "a couple browser windows and 20-50 xterms" for the
last 15+ years ;-)
As usual: different people, different use cases and needs ;-)
Kind regards,
Alex.
--
"Opportunity is missed by most people because it is dressed in overalls and
looks like work." -- Thomas A. Edison
10:22 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 20 June 2013 12:16, Alexander Schreiber <als at thangorodrim.de> wrote:
Conceded, yes. But Windows shows that this is possible on a
single-user OS with no multiuser support: it's one user at a time, per
machine. Yes, there are hacks to make it multiuser -- Terminal Server,
Winframe, Citrix, and desktop solutions such as Softxpand or BeTwin.
But as standard, it isn't. No vconsoles, no ssh or telnet, nothing.
Absolutely. But trying to be all things to all men leads to bloat.
To invoke the Pareto Principle, the way to slim, elegant S/W is to
work out the 20% of the features that are all that 80% of users need.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
On Sat, Jun 15, 2013 at 02:23:44AM +0100, Liam Proven
wrote:
I am sure it is very useful /if you need it./ However, I submit that
the majority of personal computer users do not need it.
Consider: the most popular desktop Unix by an order of magnitude (or
2) is Mac OS X. It's outsold all other commercial Unix variants put
together.
No X.11 and no networkable GUI.
Consider: the most popular Linux by a similar proportion, with many
hundreds of millions of users is Android. Its GUI is based on OpenGL
on the framebuffer, with no networkable GUI.
In fact it is fair to say that /only/ deskop/portable/handheld Unix
variants that *don't* have X.11 have been commercial successes.
On 15 June 2013 00:14, Peter Corlett <abuse at
cabal.org.uk> wrote:
I use that regularly. It is an incredibly useful detail of the way the
X Window system is designed. So you can, e.g. run the process on the
machine where the big pile of data lives and have it displayed on your
workstation. Or run the process on the workstation inside the corporate
security perimeter, but have the display on your laptop in the hotel,
with the connection piped back through ssh. BeOS was sort of a single user Unix: it was
sufficiently POSIXy for me to fire
up a shell prompt and feel at home, although it did not support multiple users
at all.
For my money -- as someone whose hobby is OSes, basically, and who has
played with as many as I can lay hands upon for about 30y now -- BeOS
is the most important OS for home/personal computers in the last few
decades.
Which is, naturally, why it's obscure, dead & forgotten.
It had many of the best aspects of xNix -- it was POSIX-like, had a
familiar shell, etc. -- but was free of all the decades of cruft
around Unix. I mean, Mac OS X is a gorgeous OS, but it's huge, vastly
complex, not very flexible or customisable, and it's only quick
because it runs on massively powerful hardware.
I don't need a network-transparent GUI. I'm sure it's a great thing,
but in some 20y of using systems with X.11 available, I've /never/
needed that. I don't
need a multiuser PC. It's mine. I'm the only user. I only need
multiuser servers. It's great but it's baggage. Security, sure, some,
but not a minicomputer/server OS on my desktop.
Again, very useful: allows separating access privileges. Quite a bit
of the stuff even on my workstation needs neither full system access
(i.e. root) nor access to my user data. I want a
solid, multitasking, multithreading OS, running on lots of
cores, delivering a fast, responsive experience. I want rich media. I
want productivity apps, but Linux and C21 FOSS delivers all I want and
more, really. I don't need anything proprietary. I want flexible
networking and plug-and-play hardware. I don't need a huge fat
dual-slot GPU -- I just want 2 or 3 smallish cheap monitors.
Linux is a great server OS, but it's not a great desktop. Never was.
It's better than it used to be but it's still not great. It's good
enough, no more.
Works for and gives me all I need. But then my idea of a "graphical desktop
environment" has been "a couple browser windows and 20-50 xterms" for the
last 15+ years ;-)
As usual: different people, different use cases and needs ;-)
21 Jun
21 Jun
12:09 a.m.
New subject: Pile on X11 - Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 20/06/13 10:22 PM, Liam Proven wrote:
...
Absolutely. But trying to be all things to all men leads to bloat.
You're not seriously trying to tell us that X11 - which was designed to
run, and ran, on machines older than some people subscribed to this list
- is "bloated"?
--T
9:46 a.m.
New subject: Pile on X11 - Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Jun 21, 2013, at 12:09 AM, Toby Thain <toby at telegraphics.com.au> wrote:
On 20/06/13 10:22 PM, Liam Proven wrote:
Well, X.org certainly is. There are X implementations that don't
try to be all things to all people that aren't.
The modern standard is pretty thick, too; it's like saying that
the modern C++ standard is bloated. Just because it's bloated
doesn't mean you can't implement something that runs fast on small
or old hardware, it just means you need to work with a subset.
- Dave
...
Absolutely. But trying to be all things to all men leads to bloat.
You're not seriously trying to tell us that X11 - which was designed to run, and ran,
on machines older than some people subscribed to this list - is "bloated"?
10:56 p.m.
New subject: Pile on X11 - Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 21/06/13 9:46 AM, David Riley wrote:
On Jun 21, 2013, at 12:09 AM, Toby Thain<toby at
telegraphics.com.au> wrote:
I expected somebody to say that. But is it really true? Xorg is
specifically (afaik) highly modular, which implies that bloat can be
effectively attacked.
Is there some inherent architectural problem with Xorg I'm not aware of?
(I run Xorg on pretty fast hardware, by my standards - a P4 3.2 HT - so
I obviously can't tell from that experience whether it's "bloated" or
not.)
--Toby
On 20/06/13 10:22 PM, Liam Proven wrote:
Well, X.org certainly is. ...
Absolutely. But trying to be all things to all men leads to bloat.
You're not seriously trying to tell us that X11 - which was designed to run, and ran,
on machines older than some people subscribed to this list - is "bloated"?
There are X implementations that don't
try to be all things to all people that aren't.
The modern standard is pretty thick, too; it's like saying that
the modern C++ standard is bloated. Just because it's bloated
doesn't mean you can't implement something that runs fast on small
or old hardware, it just means you need to work with a subset.
- Dave
11:25 p.m.
New subject: Pile on X11 - Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
Is there some inherent architectural problem with Xorg
I'm not aware
of?
Probably. In particular, it looks to me as though it's caught the
XFree86 disease of doing stuff that ought to be the kernel's job.
Bus enumeration is the most egregious exmaple I'm aware of. Even on
NetBSD/sparc, on which there's no excuse for it, it insists on walking
the SBus itself looking for framebuffers. Or at least it did when I
looked at it (and promptly held my nose and went back to using the MIT
server - I use X.org/XFree86 only on machines which are saddled with
the interface disaster that is peecee video hardware).
I also haven't managed to find enough documentation to figure out a
bunch of its internal interfaces, but it could be that I just haven't
looked hard enough.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
22 Jun
22 Jun
10:33 a.m.
New subject: Pile on X11 - Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 21/06/13 11:25 PM, Mouse wrote:
That sounds silly, but it's only one data point. Is there a *systemic*
misdesign? (Remembering that the context is "bloat".)
--T
Is there some
inherent architectural problem with Xorg I'm not aware
of?
Probably. In particular, it looks to me as though it's caught the
XFree86 disease of doing stuff that ought to be the kernel's job.
Bus enumeration is the most egregious exmaple I'm aware of. Even on
NetBSD/sparc, on which there's no excuse for it, it insists on walking
the SBus itself looking for framebuffers. Or at least it did when I
looked at it (and promptly held my nose and went back to using the MIT
server - I use X.org/XFree86 only on machines which are saddled with
the interface disaster that is peecee video hardware).
I also haven't managed to find enough documentation to figure out a
bunch of its internal interfaces, but it could be that I just haven't
looked hard enough.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
21 Jun
21 Jun
9:59 a.m.
New subject: Pile on X11 - Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 21 June 2013 05:09, Toby Thain <toby at telegraphics.com.au> wrote:
On 20/06/13 10:22 PM, Liam Proven wrote:
/Modern/ X.11 as used on modern Linux? Yes. It's indirectly-composited
via OpenGL and a plethora of layers.
Versus Wayland...
http://www.phoronix.com/scan.php?page=article&item=x_wayland_situation&…
http://askubuntu.com/questions/11537/why-is-wayland-better
Versus Mir...
http://www.theregister.co.uk/2013/03/05/canonical_mir_announcement/
http://www.phoronix.com/scan.php?page=news_item&px=MTMzMTQ
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
...
Absolutely. But trying to be all things to all men leads to bloat.
You're not seriously trying to tell us that X11 - which was designed to run,
and ran, on machines older than some people subscribed to this list - is
"bloated"?
11:04 p.m.
New subject: Pile on X11 - Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 21/06/13 9:59 AM, Liam Proven wrote:
On 21 June 2013 05:09, Toby Thain<toby at
telegraphics.com.au> wrote:
I haven't suffered by this inefficiency myself, but I'll take your word
for it. (Say, don't you spend more time using Windows than Linux?)
So basically they've decided to do what NEXTSTEP (1989) and OS X (GPU
acceleration since 10.2: 2001) did.
--Toby
On 20/06/13 10:22 PM, Liam Proven wrote:
/Modern/ X.11 as used on modern Linux? Yes. It's indirectly-composited
via OpenGL and a plethora of layers.
...
Absolutely. But trying to be all things to all men leads to bloat.
You're not seriously trying to tell us that X11 - which was designed to run,
and ran, on machines older than some people subscribed to this list - is
"bloated"?
23 Jun
23 Jun
9:07 a.m.
New subject: Pile on X11 - Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 22 June 2013 04:04, Toby Thain <toby at telegraphics.com.au> wrote:
On 21/06/13 9:59 AM, Liam Proven wrote:
It works quite well for me. I merely report what others are saying.
On 21 June 2013 05:09, Toby Thain<toby at telegraphics.com.au> wrote:
I haven't suffered by this inefficiency myself, but I'll take your word for
it.
On 20/06/13 10:22 PM, Liam Proven wrote:
/Modern/ X.11 as used on modern Linux? Yes. It's indirectly-composited
via OpenGL and a plethora of layers. ...
Absolutely. But trying to be all things to all men leads to bloat.
You're not seriously trying to tell us that X11 - which was designed to
run,
and ran, on machines older than some people subscribed to this list - is
"bloated"? (Say, don't you spend more time using Windows than
Linux?)
Me? Good gracious, no. My main PC doesn't even have a working copy of
Windows on it and neither did the previous one. My notebooks do,
mainly for BIOS flashing but occasionally for reproducing clients'
issues.
These days, I pretty much only use Windows if someone is paying me to.
So basically they've decided to do what NEXTSTEP (1989) and OS X (GPU
acceleration since 10.2: 2001) did.
A fair summary, yes.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
22 Jun
22 Jun
7:27 a.m.
New subject: Pile on X11 - Re: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 21 Jun 2013, at 05:09, Toby Thain <toby at telegraphics.com.au> wrote:
On 20/06/13 10:22 PM, Liam Proven wrote:
I refer you to chapter 7 of the Unix-Hater's Handbook. While the writing style may jar
and they are prone to a certain amount of hyperbole, the criticisms seem reasonable
enough.
X11 is an utterly complex kitchen-sink and runs completely counter to the Unix philosophy
of doing one thing, and doing it well. It was the thing that finally tipped me over the
edge into buying a Mac and dumping my Linux desktop. The Linux box remains as a server,
but it doesn't have X11 installed.
...
Absolutely. But trying to be all things to all men leads to bloat.
You're not
seriously trying to tell us that X11 - which was designed to run, and ran, on machines
older than some people subscribed to this list - is "bloated"? 
21 Jun
21 Jun
5:40 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, 21 Jun 2013, Liam Proven wrote:
On 20 June 2013 12:16, Alexander Schreiber <als at
thangorodrim.de> wrote:
I submit, that a majority of Desktop *nix users, also administrate
remote systems with X (*nix, VMS?), at which point firing an Xsession
back at yourself over ssh is a critical piece of functionality.
Though I do hear of a lot of people on this list running X on Windows for
the above. When I got to the point, where I had more PuTTY sessions and
remote X windows on my desktop than native Windows apps, I relaized I was
just using the wrong tool for my job and made the jump to Linux for my primary
workstation.
When I had my Mac, one of the first things I did was install the X11.app.
Indeed, many of my coworkers have jumped ship to using a Mac as their
primary workstation - pretty sure they all have X11 installed as well.
I don't think android (nor iOS) on phones really applies here. I don't
expect, nor need, the same type on windowing environment on a toy as I do
on a machine I use to get work done.
Yes, I acknowledge that different user populations have different needs,
however, to declare networkable X to be unneeded because *you've*
needed it or because it doesn't come natively bundled with Mac or handheld
devices? I could declare that Microsoft should stop selling Windows,
because I don't have a need for it -- but that would be ridiculous.
--
Jason Howe
Sent from my DEC 3000 running VMS 8.4
On Sat, Jun 15, 2013 at 02:23:44AM +0100, Liam
Proven wrote:
I am sure it is very useful /if you need it./ However, I submit that
the majority of personal computer users do not need it.
Consider: the most popular desktop Unix by an order of magnitude (or
2) is Mac OS X. It's outsold all other commercial Unix variants put
together.
No X.11 and no networkable GUI.
Consider: the most popular Linux by a similar proportion, with many
hundreds of millions of users is Android. Its GUI is based on OpenGL
on the framebuffer, with no networkable GUI.
In fact it is fair to say that /only/ deskop/portable/handheld Unix
variants that *don't* have X.11 have been commercial successes.
On 15 June 2013 00:14, Peter Corlett <abuse at
cabal.org.uk> wrote:
I use that regularly. It is an incredibly useful detail of the way the
X Window system is designed. So you can, e.g. run the process on the
machine where the big pile of data lives and have it displayed on your
workstation. Or run the process on the workstation inside the corporate
security perimeter, but have the display on your laptop in the hotel,
with the connection piped back through ssh. BeOS was sort of a single user Unix: it was
sufficiently POSIXy for me to fire
up a shell prompt and feel at home, although it did not support multiple users
at all.
For my money -- as someone whose hobby is OSes, basically, and who has
played with as many as I can lay hands upon for about 30y now -- BeOS
is the most important OS for home/personal computers in the last few
decades.
Which is, naturally, why it's obscure, dead & forgotten.
It had many of the best aspects of xNix -- it was POSIX-like, had a
familiar shell, etc. -- but was free of all the decades of cruft
around Unix. I mean, Mac OS X is a gorgeous OS, but it's huge, vastly
complex, not very flexible or customisable, and it's only quick
because it runs on massively powerful hardware.
I don't need a network-transparent GUI. I'm sure it's a great thing,
but in some 20y of using systems with X.11 available, I've /never/
needed that. 
7:13 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 21 June 2013 11:40, Jason Howe <jason at smbfc.net> wrote:
On Fri, 21 Jun 2013, Liam Proven wrote:
On 20 June 2013 12:16, Alexander Schreiber <als at thangorodrim.de> wrote:
I submit, that a majority of Desktop *nix users, also administrate remote
systems with X (*nix, VMS?), at which point firing an Xsession back at
yourself over ssh is a critical piece of functionality.
Though I do hear of a lot of people on this list running X on Windows for
the above. When I got to the point, where I had more PuTTY sessions and
remote X windows on my desktop than native Windows apps, I relaized I was
just using the wrong tool for my job and made the jump to Linux for my
primary workstation.
When I had my Mac, one of the first things I did was install the X11.app.
Indeed, many of my coworkers have jumped ship to using a Mac as their
primary workstation - pretty sure they all have X11 installed as well.
I don't think android (nor iOS) on phones really applies here. I don't
expect, nor need, the same type on windowing environment on a toy as I do
on a machine I use to get work done.
Yes, I acknowledge that different user populations have different needs,
however, to declare networkable X to be unneeded because *you've* needed it
or because it doesn't come natively bundled with Mac or handheld devices?
I could declare that Microsoft should stop selling Windows, because I
don't have a need for it -- but that would be ridiculous.
--
Jason Howe
Sent from my DEC 3000 running VMS 8.4
I use remote X on occasion. But I would not mind ever having to do so.
Why? While the concept is great, the implementation (on a
protocol/conceptual basis) is crap. Individual windows managed by the
local display manager? This just invites a mess of different special cases
that need to be handled (eg, widgets, pop-ups, modality, pointer
capture..). Oh, and then there are the issues of security, network
interruptions (or changes, such as if you have a laptop), and audio. These
are not insurmountable problems, but adding fixes to X to deal with these
is akin to adding airbags to a horseless carriage.
Far better - and the common solution nowadays - is to have a virtual
persistent framebuffer if you need network transparency. This is what RDP
and VNC do, and from my perspective, do it pretty well; any issues you may
observe in current systems tend to be due to the implementation.
So I say, thanks X11, but it's time to move on.
Joachim.
--
Joachim Thiemann :: http://jthiem.bitbucket.org ::
http://signalsprocessed.blogspot.com
I use that regularly. It is an incredibly useful detail of the way the
X Window system is designed. So you can, e.g. run the process on the
machine where the big pile of data lives and have it displayed on your
workstation. Or run the process on the workstation inside the corporate
security perimeter, but have the display on your laptop in the hotel,
with the connection piped back through ssh.
I am sure it is very useful /if you need it./ However, I submit that
the majority of personal computer users do not need it.
[...]
9:43 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Jun 21, 2013, at 7:13 AM, Joachim Thiemann <joachim.thiemann at gmail.com>
wrote:
I use remote X on occasion. But I would not mind ever
having to do so.
Why? While the concept is great, the implementation (on a
protocol/conceptual basis) is crap. Individual windows managed by the
local display manager? This just invites a mess of different special cases
that need to be handled (eg, widgets, pop-ups, modality, pointer
capture..). Oh, and then there are the issues of security, network
interruptions (or changes, such as if you have a laptop), and audio. These
are not insurmountable problems, but adding fixes to X to deal with these
is akin to adding airbags to a horseless carriage.
Far better - and the common solution nowadays - is to have a virtual
persistent framebuffer if you need network transparency. This is what RDP
and VNC do, and from my perspective, do it pretty well; any issues you may
observe in current systems tend to be due to the implementation.
Yes and no; for graphic-heavy things, a remote framebuffer is ideal, since
you'll essentially be duplicating that over X. But for other things,
especially text-heavy UIs, I'll go with X, especially over a slower
connection. The protocol is a lot more lightweight in the general sense.
I've seen some OK optimizations; Windows Remote Desktop is actually a lot
more speedy than I'd expect, even over a distance of several states.
- Dave
10:21 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 21 June 2013 15:43, David Riley <fraveydank at gmail.com> wrote:
On Jun 21, 2013, at 7:13 AM, Joachim Thiemann
<joachim.thiemann at gmail.com>
wrote:
Maybe - I'm actually curious if this really is the case. In the dark days,
I have done remote X over a 9600 baud modem with full Blinkenlights; there
is a significant amount of negotiation every time a window is created or
even a transient object like a pull-down menu (or the pop-up menu of
xterm). Implementations are typically synchronous as well - request
properties, wait for reply, request window to be drawn, wait for success,
etc...) so in addition to bandwidth, latency can be an issue.
Where does rendering of the text in an xterm happen? I remember xfs being
involved in some way, but it's all a bit hazy; but text-heavy programs
should work equivalently IFF rendering happens on the "client" (in X
terminology). In either case, (X or remote FB) the client sends the bitmap
deltas to the server (display).
The only circumstance where I could see the X concept be "better" (in the
responsiveness criterium) is large scrolled widows, under the assumption
that the entire underlying area is stored (pre-rendered) on the display
server, with the appropriate portion to be exposed under the control of the
server. However, I have no idea if that is actually the case with the
common X11 implementations, or if they just send "expose" events back to
the client which renders on demand.
Or, of course, one can make the argument that if your interfaces are
text-heavy, remote X11 is the wrong tool for the job and the task would be
better served with local xterm-equivalents with ssh connections and GNU
screen or tmux to provide persistence in case of network problems.
Cheers,
Joachim.
--
Joachim Thiemann :: http://jthiem.bitbucket.org ::
http://signalsprocessed.blogspot.com
Far better - and the common solution nowadays -
is to have a virtual
persistent framebuffer if you need network transparency. This is what
RDP
and VNC do, and from my perspective, do it pretty
well; any issues you
may
observe in current systems tend to be due to the
implementation.
Yes and no; for graphic-heavy things, a remote framebuffer is ideal, since
you'll essentially be duplicating that over X. But for other things,
especially text-heavy UIs, I'll go with X, especially over a slower
connection. The protocol is a lot more lightweight in the general sense.
I've seen some OK optimizations; Windows Remote Desktop is actually a lot
more speedy than I'd expect, even over a distance of several states.
11:20 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 21 June 2013 15:21, Joachim Thiemann <joachim.thiemann at gmail.com> wrote:
Or, of course, one can make the argument that if your
interfaces are
text-heavy, remote X11 is the wrong tool for the job and the task would be
better served with local xterm-equivalents with ssh connections and GNU
screen or tmux to provide persistence in case of network problems.
That is what I was thinking, as well.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
12:01 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Jun 21, 2013, at 10:21 AM, Joachim Thiemann <joachim.thiemann at gmail.com>
wrote:
Or, of course, one can make the argument that if your
interfaces are
text-heavy, remote X11 is the wrong tool for the job and the task would be
better served with local xterm-equivalents with ssh connections and GNU
screen or tmux to provide persistence in case of network problems.
What I meant as "text-heavy" was large dialogs (where the graphical
interface is important) full of text, which is rendered on the server.
As opposed to something heavily graphical, like a web browser. Not,
obviously, command-line interfaces; those are clearly better served by
something meant for command-line interfaces, like SSH. :-)
- Dave
7:37 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
> Consider: the most popular desktop Unix by an
order of magnitude (or
> 2) is Mac OS X. It's outsold all other commercial Unix variants put
> together.
Those two statements do not have as much to do with one another as you
seem to think; without installation statistics on noncommercial Unix
variants, such as Linux, we have no grounds for anything but wild
guesses as to whether OS X is "the most popular desktop Unix". At
least not for values of "Unix" relevant to this thread.
> No X.11 and no networkable GUI.
That's odd. The OS X install I know best - on my ex's iMac - has X.
Two of the programs most often running there are games I wrote, which
definitely use X. Perhaps you're thinking of an outdated version?
> Consider: the most popular Linux by a similar
proportion, with many
> hundreds of millions of users is Android.
> In fact it is fair to say that /only/
deskop/portable/handheld Unix
> variants that *don't* have X.11 have been commercial successes.
What does "commercial success" have to do with it?
I, for example, am running X under a Unix variant on at least a
half-dozen machines that show up in nobody's statistics because
everything on them is noncommercial open-source software, showing up in
no company's sales statistics and not reported to anyone. Nor am I
alone; at one of my workplaces there are more machines running such
systems than those running commercial OSes and at the other it's close
enough to half that I'm not sure which has the majority.
Nobody knows how those stack up against commercial variants, because
the nature of the open-source noncommercial world means the necessary
statistics do not exist. Not only is no-one keeping them, no-one is in
a position to keep them.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
9:44 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 21 June 2013 12:37, Mouse <mouse at rodents-montreal.org> wrote:
There are evidence-backed estimates.
http://askubuntu.com/questions/80379/how-many-ubuntu-users-are-there-worldw…
12-20 million.
Various other measures estimate around 0.8% to 1.25% of the market:
http://en.wikipedia.org/wiki/Usage_share_of_operating_systems
http://en.wikipedia.org/wiki/Linux_uptake#Measuring_desktop_adoption
http://www.starryhope.com/ubuntu-most-popular-linux-distro/
Fedora claim way more - 20 to 50 million:
https://fedoraproject.org/wiki/Statistics
However it seems fewer copies are actually in use:
http://w3techs.com/technologies/comparison/os-fedora,os-ubuntu
Mac usage is /way/ higher.
http://www.numberof.net/number%C2%A0of%C2%A0mac%C2%A0users/
http://techcrunch.com/2012/06/11/apple-there-are-now-66-million-mac-users-4…
http://idaconcpts.com/2008/12/05/a-not-so-simple-question-how-many-mac-user…
http://osxdaily.com/2011/03/18/mac-market-share-around-the-world-usa-15-can…
Mac OS X has around 8%-10% of the computer market.
There are well over a billion PCs out there, nearly 2 billion:
http://www.worldometers.info/computers/
Let's say 1? billion PCs. 10% of that is 150 million; 8% of it is 120 million.
So, something over 100 million Mac users.
That means that, if there are really only 20 million or so users of
desktop Linux, there are about 5-6? more Mac OS X users than desktop
Linux users.
That's a pretty big margin, especially considering that Linux is free
and runs on cheap commodity hardware while OS X is commercial and runs
on expensive single-vendor hardware.
It's an optional extra. I do a fair bit of Mac support. In my
experience, it's very rarely installed or used. When it is, it's for
local Unix apps.
It's a good marker of how many people are using something.
> Consider:
the most popular desktop Unix by an order of magnitude (or
> 2) is Mac OS X. It's outsold all other commercial Unix variants put
> together.
Those two statements do not have as much to do with one another as you
seem to think; without installation statistics on noncommercial Unix
variants, such as Linux, we have no grounds for anything but wild
guesses as to whether OS X is "the most popular desktop Unix". At
least not for values of "Unix" relevant to this thread. > No X.11
and no networkable GUI.
That's odd. The OS X install I know best - on my ex's iMac - has X.
Two of the programs most often running there are games I wrote, which
definitely use X. Perhaps you're thinking of an outdated version? > Consider:
the most popular Linux by a similar proportion, with many
> hundreds of millions of users is Android.
> In fact it is fair to say that /only/
deskop/portable/handheld Unix
> variants that *don't* have X.11 have been commercial successes.
What does "commercial success" have to do with it? I, for example, am running X under a Unix variant on
at least a
half-dozen machines that show up in nobody's statistics because
everything on them is noncommercial open-source software, showing up in
no company's sales statistics and not reported to anyone. Nor am I
alone; at one of my workplaces there are more machines running such
systems than those running commercial OSes and at the other it's close
enough to half that I'm not sure which has the majority.
Do you update your OSes? If you do, your update downloads are counted.
Nobody knows how those stack up against commercial
variants, because
the nature of the open-source noncommercial world means the necessary
statistics do not exist. Not only is no-one keeping them, no-one is in
a position to keep them.
Sure, stats exist. Mac sales != Mac OS X users; some run Linux, some
run Windows.
Linux downloads != Linux users: lots of people just play and never
switch to it. OTOH, those who do use it, install multiple machines (or
VMs) from a single disk image.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
10:21 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, 21 Jun 2013, Liam Proven wrote:
On 21 June 2013 12:37, Mouse <mouse at
rodents-montreal.org> wrote:
There are evidence-backed estimates.
http://askubuntu.com/questions/80379/how-many-ubuntu-users-are-there-worldw…
12-20 million.
Various other measures estimate around 0.8% to 1.25% of the market:
http://en.wikipedia.org/wiki/Usage_share_of_operating_systems
http://en.wikipedia.org/wiki/Linux_uptake#Measuring_desktop_adoption
http://www.starryhope.com/ubuntu-most-popular-linux-distro/
>
Consider: the most popular desktop Unix by an order of magnitude (or
> 2) is Mac OS X. It's outsold all other commercial Unix variants put
> together.
Those two statements do not have as much to do with one another as you
seem to think; without installation statistics on noncommercial Unix
variants, such as Linux, we have no grounds for anything but wild
guesses as to whether OS X is "the most popular desktop Unix". At
least not for values of "Unix" relevant to this thread. Fedora claim way more - 20 to 50 million:
https://fedoraproject.org/wiki/Statistics
However it seems fewer copies are actually in use:
http://w3techs.com/technologies/comparison/os-fedora,os-ubuntu
Most servers do not show up in browser and http server estimates. Web
browser and server statistics can not and do not account for the majority
of the computers out there because huge numbers of them are not being used
for web browsing or as web servers. Even those that are used as web
servers cannot be accurately measured because of round robin DNS, reverse
proxies, caching proxies, http gateways, load balancing, etc.
Similar for web browsing, since many of the larger ISPs are using
transparent caching proxies of their own. That isn't even taking into
account that many Linux users end up configuring their browsers to report
false user agent strings in order to force the numerous broken websites out
there to still work with browsers other than Firefox and Internet Explorer
under MS Windows.
Lies, damned lies, and statistics...
11:19 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 21 June 2013 15:21, Tothwolf <tothwolf at concentric.net> wrote:
Most servers do not show up in browser and http server
estimates
Who's talking about servers? We all know Linux is really strong in
servers. No argument there.
Web
browser and server statistics can not and do not account for the majority of
the computers out there because huge numbers of them are not being used for
web browsing
I submit that most desktops *are*.
Similar for web browsing, since many of the larger
ISPs are using
transparent caching proxies of their own.
If those modified the browser user agent, then they'd probably
/inflate/ user figures. But I don't think that they do.
Either way, such things would affect /all/ users on all platforms, no?
That isn't even taking into
account that many Linux users end up configuring their browsers to report
false user agent strings
Very very rare, IME.
in order to force the numerous broken websites out
there to still work with browsers other than Firefox and Internet Explorer
under MS Windows.
That way something the hardcore did in the 1990s. It's unheard-of now.
I see a broken site a few times a /decade./
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
22 Jun
22 Jun
1:29 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, 21 Jun 2013, Liam Proven wrote:
Those that tamper with the headers more often than not either strip the
useragent or replace it with their own useragent.
No, it isn't unheard of, rather it is actually quite a common practice due
to bonehead server administrators running software such as 'mod_security'
along with rules that only allow browsers with "popular" useragents to
access sites (on the false premise that the other useragents are faked
and/or are "hackers" trying to break into their website, and that
"blocking" them makes their website more secure). Assuming that this is a
rare 1990s practice and that no-one really does this anymore would be
rather foolish. In fact, I ran into this issue with Consumer Reports
website last year and took the time to try to educate them. Their reply?
"We outsource our web administration and they can't fix it."
Moreover, these browser-based usage share meters all share another common
fault -- they only make one report per IP address per sample window, which
is usually 24 hours (although some only take one sample per IP every 7
days). This means if you have a NAT router (who doesn't?) and have say a
Mac, and Windows PC, and a Linux workstation, only the first used in that
sample window is going to be recorded, and there is no real way of knowing
which website you visit is doing the sampling. If you also have a
smartphone connected to your NAT via WiFi, then that smartphone's
useragent is going to be the one recorded for your IP during the sample
window. This affects usage numbers for the "less common" operating systems
more than the "more popular" operating systems, because operating systems
such as Microsoft Windows have a larger market share.
On 21 June 2013 15:21, Tothwolf <tothwolf at
concentric.net> wrote:
We are talking about usage share. Just as you cannot accurately measure
the number of servers out there that make use Linux, you similarly cannot
/accurately/ measure the number of workstations which make use of Linux.
Most servers do not show up in browser and http
server estimates
Who's talking about servers? We all know Linux is really strong in
servers. No argument there. Web browser
and server statistics can not and do not account for the
majority of the computers out there because huge numbers of them are
not being used for web browsing
I submit that most desktops *are*.
Similar for web browsing, since many of the
larger ISPs are using
transparent caching proxies of their own.
If those modified the browser user agent, then they'd probably
/inflate/ user figures. But I don't think that they do. Either way, such things would affect /all/ users on
all platforms, no?
Yes, however given Microsoft's market share, it is going to skew the
numbers for operating systems such as Linux more than it would for
Microsoft Windows.
That isn't
even taking into account that many Linux users end up
configuring their browsers to report false user agent strings
Very very rare, IME.
in order to force the numerous broken websites
out
there to still work with browsers other than Firefox and Internet Explorer
under MS Windows.
That way something the hardcore did in the 1990s. It's unheard-of now.
I see a broken site a few times a /decade./
7:19 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
>> [...] the numerous broken websites out there
to still work with
>> browsers other than Firefox and Internet Explorer under MS Windows.
> That way something the hardcore did in the 1990s.
It's unheard-of
> now. I see a broken site a few times a /decade./
Tell that to Facebook, which still refuses to serve up any kind of
content to me when I try to script-fetch an image someone links me to
or some such, instead returning a redirect to a "you're using an
unsupported browser, use one of this half-dozen" (all, of course, are
thoroughly bloated monstrosities). I haven't bothered trying providing
a lying User-Agent: header, largely because it would be fairly
inconvenient for me to capture a sample header to base my forgery on.
I think Facebook hardly counts as "unheard-of".
I ran into this issue with Consumer Reports website
last year and
took the time to try to educate them. Their reply? "We outsource
our web administration and they can't fix it."
"You need to fire them and hire _competent_ web admins." (Yes, I
realize that's preaching to the choir. :/)
This means if you have a NAT router (who doesn't?)
I don't. Well, I do have NAT set up, but I don't have anything in the
NATted range except for a SIP phone.
and have say a Mac, and Windows PC, and a Linux
workstation, only the
first used in that sample window is going to be recorded, [...] This
affects usage numbers for the "less common" operating systems more
than the "more popular" operating systems, because operating systems
such as Microsoft Windows have a larger market share.
Actually, if you think of it as measuring "number of uses" rather than
"number of installed hosts", it's measuring exactly what it should be
measuring. (Measuring it imprecisely, because of the "first hit within
the sample window" effect, but that's what it's measuring.)
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
23 Jun
23 Jun
9:18 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 23 June 2013 00:19, Mouse <mouse at rodents-montreal.org> wrote:
[1] You do realise that you are over-trimming (*again*) and replying
to 2 different people in a single message here, don't you?
In this one, you're answering me, via an answer to someone else who
quoted my text.
[2] You're trying to do something which FB is specifically trying to
block -- IOW you're not using a browser, you're trying to scrape
images directly via a script -- and you're complaining that FB's
efforts to prevent /exactly the sort of thing that you're doing/ are
successful?
[Mocking laughter] Yeah, right.
All this proves is that you're a weirdo who likes to do things the
hard way. As was evinced earlier in your reply. No news there.
A fair point. (Also, in saying so, I hope that I am showing that I am
not being biased or taking an ad-hominem approach.)
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
>> [...]
the numerous broken websites out there to still work with
>> browsers other than Firefox and Internet Explorer under MS Windows.
> That way something the hardcore did in the
1990s. It's unheard-of
> now. I see a broken site a few times a /decade./
Tell that to Facebook, which still refuses to serve up any kind of
content to me when I try to script-fetch an image someone links me to
or some such, instead returning a redirect to a "you're using an
unsupported browser, use one of this half-dozen" (all, of course, are
thoroughly bloated monstrosities). I haven't bothered trying providing
a lying User-Agent: header, largely because it would be fairly
inconvenient for me to capture a sample header to base my forgery on.
I think Facebook hardly counts as "unheard-of". I ran into
this issue with Consumer Reports website last year and
took the time to try to educate them. Their reply? "We outsource
our web administration and they can't fix it."
"You need to fire them and hire _competent_ web admins." (Yes, I
realize that's preaching to the choir. :/)
This means if you have a NAT router (who
doesn't?)
I don't. Well, I do have NAT set up, but I don't have anything in the
NATted range except for a SIP phone. and have say a
Mac, and Windows PC, and a Linux workstation, only the
first used in that sample window is going to be recorded, [...] This
affects usage numbers for the "less common" operating systems more
than the "more popular" operating systems, because operating systems
such as Microsoft Windows have a larger market share.
Actually, if you think of it as measuring "number of uses" rather than
"number of installed hosts", it's measuring exactly what it should be
measuring. (Measuring it imprecisely, because of the "first hit within
the sample window" effect, but that's what it's measuring.)
9:15 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 22 June 2013 18:29, Tothwolf <tothwolf at concentric.net> wrote:
You'll have to walk me through that reasoning. I do not understand or
see what you mean at all.
No, it isn't unheard of, rather it is actually quite a common practice due
to bonehead server administrators running software such as 'mod_security'
along with rules that only allow browsers with "popular" useragents to
access sites (on the false premise that the other useragents are faked
and/or are "hackers" trying to break into their website, and that
"blocking"
them makes their website more secure). Assuming that this is a rare 1990s
practice and that no-one really does this anymore would be rather foolish.
In fact, I ran into this issue with Consumer Reports website last year and
took the time to try to educate them. Their reply? "We outsource our web
administration and they can't fix it."
Again, [[citation needed]]. I've been a desktop Linux user full-time
for a decade now, and part-time for about 6-7y before that. I have
encountered this problem so very, *very* seldom it is not a factor and
has not been this century.
On Fri, 21 Jun 2013, Liam Proven wrote:
Are we?
I thought I was talking about networkable X.11, which by and large
doesn't really apply to servers. I don't know what you feel that we're
talking about.
On 21 June 2013 15:21, Tothwolf <tothwolf at concentric.net> wrote:
We are talking about usage share. Most servers do not show up in browser and http
server estimates
Who's talking about servers? We all know Linux is really strong in
servers. No argument there. Just as you cannot accurately measure the
number of servers out there that make use Linux, you similarly cannot
/accurately/ measure the number of workstations which make use of Linux.
Definte "accurately", with your idea of acceptable error bars, and
compare those to the accuracy of figures for Macs and Windows.
*I* find the figures acceptable, myself.
Those that tamper with the headers more often than not
either strip the
useragent or replace it with their own useragent.
[[Citation needed]]
More to the point, some kind of evidence that this is a significant
factor. TBH it sounds like special pleading to me: "we don't know how
many desktop Linux users are, we can't know, but probably there are
more than reported and logged because lots of them are _hiding_."
Either way,
such things would affect /all/ users on all platforms, no?
Yes, however given Microsoft's market share, it is going to skew the numbers
for operating systems such as Linux more than it would for Microsoft
Windows. That isn't even taking into account that many
Linux users end up
configuring their browsers to report false user agent strings
Very very rare, IME.
in order to force the numerous broken websites
out
there to still work with browsers other than Firefox and Internet
Explorer
under MS Windows.
That way something the hardcore did in the 1990s. It's unheard-of now.
I see a broken site a few times a /decade./ Moreover, these browser-based usage share meters all
share another common
fault -- they only make one report per IP address per sample window, which
is usually 24 hours (although some only take one sample per IP every 7
days). This means if you have a NAT router (who doesn't?) and have say a
Mac, and Windows PC, and a Linux workstation, only the first used in that
sample window is going to be recorded, and there is no real way of knowing
which website you visit is doing the sampling. If you also have a smartphone
connected to your NAT via WiFi, then that smartphone's useragent is going to
be the one recorded for your IP during the sample window. This affects usage
numbers for the "less common" operating systems more than the "more
popular"
operating systems, because operating systems such as Microsoft Windows have
a larger market share.
That one is interesting, I'll give you. I've not heard that before and
you do have a possible point there, although once again, I'd like to
see some actual evidence that it was occurring.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
11:52 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Sun, 23 Jun 2013, Liam Proven wrote:
Look it up yourself. I've already done so (many years ago -- Wikipedia
doesn't write itself) and I certainly don't feel the need to look it up
for you. Hint: AT&T, Comcast, COX, Time Warner, and :gasp: AOL have all
done this at various times and various other ISPs (transparent proxies)
and some webhosts (load balancing, caching) /still/ do.
You'll have to walk me through that reasoning. I do not understand or
see what you mean at all.
Let me get this straight...you don't understand how the percentages are
calculated?
No, it isn't unheard of, rather it is actually quite a common practice due
to bonehead server administrators running software such as 'mod_security'
along with rules that only allow browsers with "popular" useragents to
access sites (on the false premise that the other useragents are faked
and/or are "hackers" trying to break into their website, and that
"blocking"
them makes their website more secure). Assuming that this is a rare 1990s
practice and that no-one really does this anymore would be rather foolish.
In fact, I ran into this issue with Consumer Reports website last year and
took the time to try to educate them. Their reply? "We outsource our web
administration and they can't fix it."
Again, [[citation needed]]. I've been a desktop Linux user full-time
for a decade now, and part-time for about 6-7y before that. I have
encountered this problem so very, *very* seldom it is not a factor and
has not been this century.
You must not get around the web very much, then.
[5 Aug 2012]
: "Our host operates a security perimeter that performs this and other filtering, the
rules simply aren't under our control. The Comodo browser, based on the Chromium
codebase, also sports a u-a that the security perimeter does not allow through."
:
: "I apologize again for the inconvenience and hope you'll be able to stay with
us."
I still don't know why he mentioned Comodo, since that wasn't what I was
using or asking about. I had reported to them that Firefox (nightly
builds) and Lynx (all versions) were not working with their site.
http://yousefourabi.com/blog/2007/10/blocking-bots-with-mod-security/
: "There are two approaches: 1) Block everything but a pre-approved white list, 2)
Allow everything except a pre-denied black list."
The first approach may not be a workable solution, and as stupid as it
might it, that doesn't stop people from _trying_ anyway.
http://www.google.com/ ...
On 22 June 2013 18:29, Tothwolf <tothwolf at
concentric.net> wrote:
It certainly seems like you were, until you just now claimed claimed
otherwise...
: In-Reply-To: <201306211137.HAA10837 at Chip.Rodents-Montreal.ORG>
: Date: Fri, 21 Jun 2013 14:44:00 +0100
: Message-ID: <CAMTenCGtZVXaRstyrQ6_uRT7+vtdUSY1Nd-deXeNHg2PeE57sA at
mail.gmail.com>
: Subject: Re: Devs should be testing on slow machines - Re: What versions of Linux (or
UNIX) are good on old 486 boxes?
: From: Liam Proven <lproven at gmail.com>
:
[SNIP]
: There are evidence-backed estimates.
: http://askubuntu.com/questions/80379/how-many-ubuntu-users-are-there-worldw…
: 12-20 million.
: Various other measures estimate around 0.8% to 1.25% of the market:
: http://en.wikipedia.org/wiki/Usage_share_of_operating_systems
: http://en.wikipedia.org/wiki/Linux_uptake#Measuring_desktop_adoption
[SNIP]
: Mac usage is /way/ higher.
: http://www.numberof.net/number%C2%A0of%C2%A0mac%C2%A0users/
:
http://techcrunch.com/2012/06/11/apple-there-are-now-66-million-mac-users-4…
[SNIP]
On Fri, 21 Jun 2013, Liam Proven wrote:
Are we?
I thought I was talking about networkable X.11, which by and large
doesn't really apply to servers. I don't know what you feel that we're
talking about. On 21 June 2013 15:21, Tothwolf <tothwolf at
concentric.net> wrote:
We are talking about usage share. Most servers do not show up in browser and http
server estimates
Who's talking about servers? We all know Linux is really strong in
servers. No argument there. Just as you
cannot accurately measure the
number of servers out there that make use Linux, you similarly cannot
/accurately/ measure the number of workstations which make use of Linux.
Definte "accurately", with your idea of acceptable error bars, and
compare those to the accuracy of figures for Macs and Windows.
*I* find the figures acceptable, myself.
Those that tamper with the headers more often
than not either strip the
useragent or replace it with their own useragent.
[[Citation needed]] More to the point, some kind of evidence that this is
a significant
factor. TBH it sounds like special pleading to me: "we don't know how
many desktop Linux users are, we can't know, but probably there are
more than reported and logged because lots of them are _hiding_."
Straw Man.
[[Insert giant ASCII middle finger]]
Either way, such things would affect /all/ users on
all platforms, no?
Yes, however given Microsoft's market share, it is going to skew the numbers
for operating systems such as Linux more than it would for Microsoft
Windows. That isn't
even taking into account that many Linux users end up
configuring their browsers to report false user agent strings
Very very rare, IME.
in order to force the numerous broken websites
out
there to still work with browsers other than Firefox and Internet
Explorer
under MS Windows.
That way something the hardcore did in the 1990s. It's unheard-of now.
I see a broken site a few times a /decade./ Moreover,
these browser-based usage share meters all share another common
fault -- they only make one report per IP address per sample window, which
is usually 24 hours (although some only take one sample per IP every 7
days). This means if you have a NAT router (who doesn't?) and have say a
Mac, and Windows PC, and a Linux workstation, only the first used in that
sample window is going to be recorded, and there is no real way of knowing
which website you visit is doing the sampling. If you also have a smartphone
connected to your NAT via WiFi, then that smartphone's useragent is going to
be the one recorded for your IP during the sample window. This affects usage
numbers for the "less common" operating systems more than the "more
popular"
operating systems, because operating systems such as Microsoft Windows have
a larger market share.
That one is interesting, I'll give you. I've not heard that before and
you do have a possible point there, although once again, I'd like to
see some actual evidence that it was occurring.
12:58 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 23 June 2013 16:52, Tothwolf <tothwolf at concentric.net> wrote:
On Sun, 23 Jun 2013, Liam Proven wrote:
Look it up yourself. I've already done so (many years ago -- Wikipedia
doesn't write itself) and I certainly don't feel the need to look it up for
you. Hint: AT&T, Comcast, COX, Time Warner, and :gasp: AOL have all done
this at various times and various other ISPs (transparent proxies) and some
webhosts (load balancing, caching) /still/ do.
You'll have to walk me through that reasoning. I do not understand or
see what you mean at all.
Let me get this straight...you don't understand how the percentages are
calculated?
Again, [[citation needed]]. I've been a desktop Linux user full-time
for a decade now, and part-time for about 6-7y before that. I have
encountered this problem so very, *very* seldom it is not a factor and
has not been this century.
You must not get around the web very much, then.
[5 Aug 2012]
: "Our host operates a security perimeter that performs this and other
filtering, the rules simply aren't under our control. The Comodo browser,
based on the Chromium codebase, also sports a u-a that the security
perimeter does not allow through."
:
: "I apologize again for the inconvenience and hope you'll be able to stay
with us."
I still don't know why he mentioned Comodo, since that wasn't what I was
using or asking about. I had reported to them that Firefox (nightly builds)
and Lynx (all versions) were not working with their site.
http://yousefourabi.com/blog/2007/10/blocking-bots-with-mod-security/
: "There are two approaches: 1) Block everything but a pre-approved white
list, 2) Allow everything except a pre-denied black list."
The first approach may not be a workable solution, and as stupid as it might
it, that doesn't stop people from _trying_ anyway.
http://www.google.com/ ...
So once again, the big brave person with the pretend name refuses to
come up with any evidence whatsoever to support their argument, calls
me stupid, wifully misunderstands statements (e.g. I don't understand
your reasoning) and offers pointedly stupid counters (e.g. what, you
don't know how to calculate percentages?).
Sod it. I haven't got time to play games.
You've lost the debate by attempting to play childish games. I remain
not even slightly convinced that you have a point and you make no
effort to back up your arguments.
No, telling your opponent to do *your* research is not how it works.
Go troll someone else.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
On 22 June 2013 18:29, Tothwolf <tothwolf at concentric.net> wrote:
It certainly seems like you were, until you just now claimed claimed
otherwise...
: In-Reply-To: <201306211137.HAA10837 at Chip.Rodents-Montreal.ORG>
: Date: Fri, 21 Jun 2013 14:44:00 +0100
: Message-ID:
<CAMTenCGtZVXaRstyrQ6_uRT7+vtdUSY1Nd-deXeNHg2PeE57sA at mail.gmail.com>
: Subject: Re: Devs should be testing on slow machines - Re: What versions
of Linux (or UNIX) are good on old 486 boxes?
: From: Liam Proven <lproven at gmail.com>
:
[SNIP]
: There are evidence-backed estimates.
:
http://askubuntu.com/questions/80379/how-many-ubuntu-users-are-there-worldw…
: 12-20 million.
: Various other measures estimate around 0.8% to 1.25% of the market:
: http://en.wikipedia.org/wiki/Usage_share_of_operating_systems
: http://en.wikipedia.org/wiki/Linux_uptake#Measuring_desktop_adoption
[SNIP]
: Mac usage is /way/ higher.
: http://www.numberof.net/number%C2%A0of%C2%A0mac%C2%A0users/
:
http://techcrunch.com/2012/06/11/apple-there-are-now-66-million-mac-users-4…
[SNIP]
On Fri, 21 Jun 2013, Liam Proven wrote:
Are we?
I thought I was talking about networkable X.11, which by and large doesn't
really apply to servers. I don't know what you feel that we're talking
about.
On 21 June 2013 15:21, Tothwolf <tothwolf at concentric.net> wrote:
> Most servers do not show up in browser and http server estimates
Who's talking about servers? We all know Linux is really strong in
servers. No argument there.
We are talking about usage share. Just as
you cannot accurately measure the
number of servers out there that make use Linux, you similarly cannot
/accurately/ measure the number of workstations which make use of Linux.
Definte "accurately", with your idea of acceptable error bars, and
compare those to the accuracy of figures for Macs and Windows.
*I* find the figures acceptable, myself.
Those that tamper with the headers more often
than not either strip the
useragent or replace it with their own useragent.
[[Citation needed]] More to the point, some kind of evidence that
this is a significant
factor. TBH it sounds like special pleading to me: "we don't know how
many desktop Linux users are, we can't know, but probably there are
more than reported and logged because lots of them are _hiding_."
Straw Man.
[[Insert giant ASCII middle finger]]
Either way, such things would affect /all/ users on
all platforms, no?
Yes, however given Microsoft's market share, it is going to skew the
numbers
for operating systems such as Linux more than it would for Microsoft
Windows. > That isn't even taking into account that many
Linux users end up
> configuring their browsers to report false user agent strings
Very very rare, IME.
> in order to force the numerous broken websites out
> there to still work with browsers other than Firefox and Internet
> Explorer
> under MS Windows.
That way something the hardcore did in the 1990s. It's unheard-of now.
I see a broken site a few times a /decade./
No, it isn't unheard of, rather it is actually quite a common practice
due
to bonehead server administrators running software such as 'mod_security'
along with rules that only allow browsers with "popular" useragents to
access sites (on the false premise that the other useragents are faked
and/or are "hackers" trying to break into their website, and that
"blocking"
them makes their website more secure). Assuming that this is a rare 1990s
practice and that no-one really does this anymore would be rather
foolish.
In fact, I ran into this issue with Consumer Reports website last year
and
took the time to try to educate them. Their reply? "We outsource our web
administration and they can't fix it." Moreover,
these browser-based usage share meters all share another common
fault -- they only make one report per IP address per sample window,
which
is usually 24 hours (although some only take one sample per IP every 7
days). This means if you have a NAT router (who doesn't?) and have say a
Mac, and Windows PC, and a Linux workstation, only the first used in that
sample window is going to be recorded, and there is no real way of
knowing
which website you visit is doing the sampling. If you also have a
smartphone
connected to your NAT via WiFi, then that smartphone's useragent is going
to
be the one recorded for your IP during the sample window. This affects
usage
numbers for the "less common" operating systems more than the "more
popular"
operating systems, because operating systems such as Microsoft Windows
have
a larger market share.
That one is interesting, I'll give you. I've not heard that before and
you do have a possible point there, although once again, I'd like to
see some actual evidence that it was occurring.
4:13 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Sun, 23 Jun 2013, Liam Proven wrote:
On 23 June 2013 16:52, Tothwolf <tothwolf at
concentric.net> wrote:
[C&C warning]
Oh wow, more personal attacks, including a rather snarky one which is
completely unrelated to the topic at hand. Looks like your typical
response cycle. Rinse, repeat.
Sorry, but I wasn't debating anything, nor do I troll. I do however call
it as I see it, and you sir, are a bona fide troll. Pot, Kettle. In the
past I've gone up against some of the absolute worst of the worst when it
comes to online trolls (some of which would actually make for some rather
interesting articles), and while you aren't anywhere near the worst, you
are still a troll. On a scale of one to ten, with ten being the worst type
of troll, you are about a three.
I've already done the research for my own purposes and it isn't my problem
that you are just too damn lazy to go look stuff up for yourself with a
basic search with Google and Wikipedia. Given that you refuse to do your
own research, and given way you tend to "respond" when people disagree
with you (I'm now your "opponent"?!), I can now see _exactly_ why Dave and
others take issue with some of your posts and get all riled up. You seem
to feed on the attention though, so I suppose you'll be smiling and
laughing at my response here too.
For that fact, if you can't be bothered to go look stuff up then what the
hell kind of "journalist" are you? Its no wonder some of your articles
tend to be biased and present an incomplete picture to readers. In the
past I've even politely (offlist) tried to point you in the direction of
some stuff that you overlooked or missed (including links and all), but
you gave me the brush off and basically said I didn't know what I was
talking about (I was there dude, I remember this stuff). I already /know/
that I'm a better researcher than you are and _I_ honestly don't feel the
need to "prove" /anything/ to you.
You know what, Fuck you.
EOT
[Sorry to the other list members who've had to wade through Liam's straw
man arguments and nonsensical responses in this thread, but after the
personal attacks, I'm done with this. Liam can go "debate" with himself in
the mirror for all I care. I've got more important projects to tend to.]
On Sun, 23 Jun 2013, Liam Proven wrote:
Look it up yourself. I've already done so (many years ago -- Wikipedia
doesn't write itself) and I certainly don't feel the need to look it up
for you. Hint: AT&T, Comcast, COX, Time Warner, and :gasp: AOL have all
done this at various times and various other ISPs (transparent proxies)
and some webhosts (load balancing, caching) /still/ do.
Let me get this straight...you don't understand how the percentages are
calculated?
You must not get around the web very much, then.
[5 Aug 2012]
: "Our host operates a security perimeter that performs this and other
filtering, the rules simply aren't under our control. The Comodo
browser, based on the Chromium codebase, also sports a u-a that the
security perimeter does not allow through."
:
: "I apologize again for the inconvenience and hope you'll be able to
stay with us."
I still don't know why he mentioned Comodo, since that wasn't what I
was using or asking about. I had reported to them that Firefox (nightly
builds) and Lynx (all versions) were not working with their site.
http://yousefourabi.com/blog/2007/10/blocking-bots-with-mod-security/
: "There are two approaches: 1) Block everything but a pre-approved
white list, 2) Allow everything except a pre-denied black list."
The first approach may not be a workable solution, and as stupid as it
might it, that doesn't stop people from _trying_ anyway.
http://www.google.com/ ...
So once again, the big brave person with the pretend name refuses to
come up with any evidence whatsoever to support their argument, calls me
stupid, wifully misunderstands statements (e.g. I don't understand your
reasoning) and offers pointedly stupid counters (e.g. what, you don't
know how to calculate percentages?).
Sod it. I haven't got time to play games.
You've lost the debate by attempting to play childish games. I remain
not even slightly convinced that you have a point and you make no effort
to back up your arguments.
No, telling your opponent to do *your* research is not how it works.
Go troll someone else. On 22 June 2013 18:29, Tothwolf <tothwolf at
concentric.net> wrote:
It certainly seems like you were, until you just now claimed claimed
otherwise...
: In-Reply-To: <201306211137.HAA10837 at Chip.Rodents-Montreal.ORG>
: Date: Fri, 21 Jun 2013 14:44:00 +0100
: Message-ID: <CAMTenCGtZVXaRstyrQ6_uRT7+vtdUSY1Nd-deXeNHg2PeE57sA at
mail.gmail.com>
: Subject: Re: Devs should be testing on slow machines - Re: What versions of Linux (or
UNIX) are good on old 486 boxes?
: From: Liam Proven <lproven at gmail.com>
[SNIP]
: There are evidence-backed estimates.
: http://askubuntu.com/questions/80379/how-many-ubuntu-users-are-there-worldw…
: 12-20 million.
: Various other measures estimate around 0.8% to 1.25% of the market:
: http://en.wikipedia.org/wiki/Usage_share_of_operating_systems
: http://en.wikipedia.org/wiki/Linux_uptake#Measuring_desktop_adoption
[SNIP]
: Mac usage is /way/ higher.
: http://www.numberof.net/number%C2%A0of%C2%A0mac%C2%A0users/
:
http://techcrunch.com/2012/06/11/apple-there-are-now-66-million-mac-users-4…
[SNIP]
On Fri, 21 Jun 2013, Liam Proven wrote:
> On 21 June 2013 15:21, Tothwolf <tothwolf at concentric.net> wrote:
> Who's talking about servers? We all know Linux is really strong in
> servers. No argument there.
We are talking about usage share.
Are we?
I thought I was talking about networkable X.11, which by and large
doesn't really apply to servers. I don't know what you feel that we're
talking about. Just as
you cannot accurately measure the number of servers out there
that make use Linux, you similarly cannot /accurately/ measure the
number of workstations which make use of Linux.
Definte "accurately", with your idea of acceptable error bars, and
compare those to the accuracy of figures for Macs and Windows.
*I* find the figures acceptable, myself.
Those that tamper with the headers more often
than not either strip
the useragent or replace it with their own useragent.
[[Citation needed]] More to the point, some kind of evidence that
this is a significant
factor. TBH it sounds like special pleading to me: "we don't know how
many desktop Linux users are, we can't know, but probably there are
more than reported and logged because lots of them are _hiding_."
Straw Man.
[[Insert giant ASCII middle finger]]
>
Either way, such things would affect /all/ users on all platforms,
> no?
Yes, however given Microsoft's market share, it is going to skew the
numbers for operating systems such as Linux more than it would for
Microsoft Windows.
You'll have to walk me through that reasoning. I do not understand or
see what you mean at all. >>
That isn't even taking into account that many Linux users end up
>> configuring their browsers to report false user agent strings
>
> Very very rare, IME.
>
>> in order to force the numerous broken websites out there to still
>> work with browsers other than Firefox and Internet Explorer under
>> MS Windows.
>
> That way something the hardcore did in the 1990s. It's unheard-of
> now. I see a broken site a few times a /decade./
No, it isn't unheard of, rather it is actually quite a common
practice due to bonehead server administrators running software such
as 'mod_security' along with rules that only allow browsers with
"popular" useragents to access sites (on the false premise that the
other useragents are faked and/or are "hackers" trying to break into
their website, and that "blocking" them makes their website more
secure). Assuming that this is a rare 1990s practice and that no-one
really does this anymore would be rather foolish. In fact, I ran into
this issue with Consumer Reports website last year and took the time
to try to educate them. Their reply? "We outsource our web
administration and they can't fix it."
Again, [[citation needed]]. I've been a desktop Linux user full-time
for a decade now, and part-time for about 6-7y before that. I have
encountered this problem so very, *very* seldom it is not a factor and
has not been this century. Moreover,
these browser-based usage share meters all share another
common fault -- they only make one report per IP address per sample
window, which is usually 24 hours (although some only take one sample
per IP every 7 days). This means if you have a NAT router (who
doesn't?) and have say a Mac, and Windows PC, and a Linux
workstation, only the first used in that sample window is going to be
recorded, and there is no real way of knowing which website you visit
is doing the sampling. If you also have a smartphone connected to
your NAT via WiFi, then that smartphone's useragent is going to be
the one recorded for your IP during the sample window. This affects
usage numbers for the "less common" operating systems more than the
"more popular" operating systems, because operating systems such as
Microsoft Windows have a larger market share.
That one is interesting, I'll give you. I've not heard that before and
you do have a possible point there, although once again, I'd like to
see some actual evidence that it was occurring.
21 Jun
21 Jun
9:21 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 21 June 2013 10:40, Jason Howe <jason at smbfc.net> wrote:
On Fri, 21 Jun 2013, Liam Proven wrote:
I submit that the majority of desktop *nix users are on Macs and
95%-99% of Mac users don't know what X.11 /is/ and have never used it.
On 20 June 2013 12:16, Alexander Schreiber
<als at thangorodrim.de> wrote:
I submit, that a majority of Desktop *nix users, also administrate remote
systems with X (*nix, VMS?), at which point firing an Xsession back at
yourself over ssh is a critical piece of functionality.
On Sat, Jun 15, 2013 at 02:23:44AM +0100, Liam Proven wrote:
I am sure it is very useful /if you need it./ However, I submit that
the majority of personal computer users do not need it.
Consider: the most popular desktop Unix by an order of magnitude (or
2) is Mac OS X. It's outsold all other commercial Unix variants put
together.
No X.11 and no networkable GUI.
Consider: the most popular Linux by a similar proportion, with many
hundreds of millions of users is Android. Its GUI is based on OpenGL
on the framebuffer, with no networkable GUI.
In fact it is fair to say that /only/ deskop/portable/handheld Unix
variants that *don't* have X.11 have been commercial successes.
On 15 June 2013 00:14, Peter Corlett <abuse at cabal.org.uk> wrote:
>
> BeOS was sort of a single user Unix: it was sufficiently POSIXy for me
> to fire
> up a shell prompt and feel at home, although it did not support
> multiple users
> at all.
For my money -- as someone whose hobby is OSes, basically, and who has
played with as many as I can lay hands upon for about 30y now -- BeOS
is the most important OS for home/personal computers in the last few
decades.
Which is, naturally, why it's obscure, dead & forgotten.
It had many of the best aspects of xNix -- it was POSIX-like, had a
familiar shell, etc. -- but was free of all the decades of cruft
around Unix. I mean, Mac OS X is a gorgeous OS, but it's huge, vastly
complex, not very flexible or customisable, and it's only quick
because it runs on massively powerful hardware.
I don't need a network-transparent GUI. I'm sure it's a great thing,
but in some 20y of using systems with X.11 available, I've /never/
needed that.
I use that regularly. It is an incredibly useful detail of the way the
X Window system is designed. So you can, e.g. run the process on the
machine where the big pile of data lives and have it displayed on your
workstation. Or run the process on the workstation inside the corporate
security perimeter, but have the display on your laptop in the hotel,
with the connection piped back through ssh. Though I do hear of a lot of people on this list
running X on Windows for
the above. When I got to the point, where I had more PuTTY sessions and
remote X windows on my desktop than native Windows apps, I relaized I was
just using the wrong tool for my job and made the jump to Linux for my
primary workstation.
When I had my Mac, one of the first things I did was install the X11.app.
Indeed, many of my coworkers have jumped ship to using a Mac as their
primary workstation - pretty sure they all have X11 installed as well.
Most techies do, in my experience. *I* do on *my* Macs, although mine
are all PowerPC-based & don't get used much any more. However, of
those installed copies of X11.app, I bet that the majority are used
for running /local/ X apps ported from other Unices.
I don't think android (nor iOS) on phones really
applies here. I don't
expect, nor need, the same type on windowing environment on a toy as I do on
a machine I use to get work done.
Conceded, yes. However, we are right in the middle of an explosion in
tablet sales and I think that within a few years (i.e. 3 at max) they
will significantly outnumber all desktops/notebooks put together.
Yes, I acknowledge that different user populations
have different needs,
however, to declare networkable X to be unneeded because *you've* needed it
or because it doesn't come natively bundled with Mac or handheld devices? I
could declare that Microsoft should stop selling Windows, because I don't
have a need for it -- but that would be ridiculous.
I'm not saying it's because /I/ don't need it. I'm saying it's
because
the majority of users don't and that is why the 2 best-selling *nix
OSes have discarded it and why the #1 desktop/notebook Linux distro is
working hard to get rid of it too.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
22 Jun
22 Jun
9:40 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Fri, Jun 21, 2013 at 03:22:55AM +0100, Liam Proven wrote:
On 20 June 2013 12:16, Alexander Schreiber <als at
thangorodrim.de> wrote:
Well, I _did_ try to use a MacBook as a laptop at work once. I gave it
back in disgust after two months because I couldn't get _work_ done with
the darn thing. Returned to my Thinkpad running Linux and work got done
again ;-)
On Sat, Jun 15, 2013 at 02:23:44AM +0100, Liam
Proven wrote:
I am sure it is very useful /if you need it./ However, I submit that
the majority of personal computer users do not need it.
Consider: the most popular desktop Unix by an order of magnitude (or
2) is Mac OS X. It's outsold all other commercial Unix variants put
together.
No X.11 and no networkable GUI.
I don't need a network-transparent GUI. I'm sure it's a great thing,
but in some 20y of using systems with X.11 available, I've /never/
needed that.
I use that regularly. It is an incredibly useful detail of the way the
X Window system is designed. So you can, e.g. run the process on the
machine where the big pile of data lives and have it displayed on your
workstation. Or run the process on the workstation inside the corporate
security perimeter, but have the display on your laptop in the hotel,
with the connection piped back through ssh. Consider: the most popular Linux by a similar
proportion, with many
hundreds of millions of users is Android. Its GUI is based on OpenGL
on the framebuffer, with no networkable GUI.
But Android solves a different problem: a mobile phone/tablet OS, not
a workstation OS.
Kind regards,
Alex.
--
"Opportunity is missed by most people because it is dressed in overalls and
looks like work." -- Thomas A. Edison
23 Jun
23 Jun
9:20 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 23 June 2013 02:40, Alexander Schreiber <als at thangorodrim.de> wrote:
Well, I _did_ try to use a MacBook as a laptop at work
once. I gave it
back in disgust after two months because I couldn't get _work_ done with
the darn thing. Returned to my Thinkpad running Linux and work got done
again ;-)
So long as you have something that works for you, good.
I must say, though, that most of the old-time Linux users that I know
-- the determined hardcore who were using it in the 1990s, who've been
on it for a decade or more -- have almost *all* switched to Mac OS X
now.
But Android solves a different problem: a mobile
phone/tablet OS, not
a workstation OS.
Sure.
But this shows that the approach of getting rid of X.11 is not limited
to a single niche.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
12 Jun
12 Jun
7:14 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Wed, 12 Jun 2013, Peter Corlett wrote:
me to write Unix software for you, insisting that I
must only use a ratty old
Pentium 4 with 512MB of RAM running a locked-down Windows XP that won't even
let me install PuTTY or an editor is not acceptable.
Absolutely.
BUT, until you DO, you don't have any idea how your code performs for
those who are forced to USE it in such an environment.
A developer who "requires" better than a ratty old Pentium 4 with 512M of
RAM running a locked-down copy of Windows XP" writes software that does
not perform acceptably for those conditions.
If I am in a position of having to use such a machine, then how satisfied
am I likely to be with software written on a better machine, and NEVER
REALLY TESTED on a machine like mine?
Microsoft Windoze 3.0/3.1 worked nicely in 800x600 with ATI video cards.
Most other video display configurations were marginal at best.
In those days, you could tell what hardware their programmers used by
comparing how well systems worked right!
You have gotta have more than one machine. There must be testing done on
all levels of the target platform, and even enough not-explicitly-testing
work done using them for the developer to have a feel for what it is like
to be using one.
By all means, do your work on a state of the art machine. But, if that's
ALL that you will ever use, then the official "testing" will be
inadequate. Have a complete assortment, and find SOME components of your
work that can be done on the crap.
(The machine used by your bookkeeper does NOT need to be statew of the
art)
Elitist assholes who refuse to EVER use machines like what their customers
have, need to be canned. In well whale oil.
--
Grumpy Ol' Fred cisin at xenosoft.com
7:30 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
I think many developers quit trying to have their apps run on old average
hardware a long time ago and say so in the specs for it.
-----Original Message-----
From: Fred Cisin
Sent: Wednesday, June 12, 2013 7:14 PM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: Devs should be testing on slow machines - Re: What versions of
Linux (or UNIX) are good on old 486 boxes?
Elitist assholes who refuse to EVER use machines like what their customers
have, need to be canned. In well whale oil.
--
Grumpy Ol' Fred cisin at xenosoft.com
9:15 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Wed, 12 Jun 2013, TeoZ wrote:
I think many developers quit trying to have their apps
run on old average
hardware a long time ago and say so in the specs for it.
A well defined spec clearly identifies what range is suitable. It is then
the developer's responsibility to work right for everything anywhere
within that range.
A tightly defined spec makes for an acceptable range.
If they define their target platform as "ALL PCs" then that is what they
need to work with.
Setting a target platform spec that is too all-encompassing creates an
obligation that they can't meet.
If the software supposedly runs on a ratty old Pentium 4 with 512MB of RAM
running a locked-down Windows XP that won't even let me install PuTTY or
an editor, then that is what the developer needs to deal with. Not all of
the time, but at least enough to have a feel for what the hell is like.
Again, a proper development environment heeds to have the entire spectrum
of the target platform. High end machines for developer productivity AND
POS to do REAL, not just token, testing. MICROS~1 never understood that
seeing if it booted, and running a single game of Solitaire (their only
thoroughly tested product at the time) doe NOT really constitute
"testing". Making the staff USE 286's for their word processing and
other "lesser" tasks would have improved their awareness of the platform.
Microsoft stopped testing 3.0 with 8088 long before they stopped selling
it for that, and stopped testing 3.10 with 286 long before they stopped
selling it for that.
Properly planned testing might significantly reduce the amount of testing
needed. Badly designed testing does nothing but waste everybody's time.
Number of games of Solitaire played is not the right metric unless that is
the primary intended use (maybe it was?)
--
Grumpy Ol' Fred cisin at xenosoft.com
13 Jun
13 Jun
12:27 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 6/12/13 6:15 PM, "Fred Cisin" <cisin at xenosoft.com> wrote:
Properly planned testing might significantly reduce the amount of testing
needed. Badly designed testing does nothing but waste everybody's time.
Number of games of Solitaire played is not the right metric unless that is
the primary intended use (maybe it was?)
Yes, you're right, but there are assumptions you're ignoring (presumably
for simplicity's sake). Inadequately staffed testing does the best it can
within the constraints (calendar) prescribed by the corporate masters.
One can plan testing until one is bleary-eyed, but if the product changes
weekly because the marketing weenies say it must (with no change in ship
dates), testing will not be able to plan meaningfully. I'm not referring
to changes in implementation of feature X: I'm talking about, "We GOTTA
have this new feature! EVERYONE wants it!" I'm not joking, and having
dredged up these painful memories from my days at [insert humongous
software company here] I probably won't sleep well tonight....
I don't mind if dev works on the fast machines, as long as when test
brings them a bug from a real-world, dog-slow machine, they don't respond
(wait for it): "It works fine on my machine." IMHO, mayhem is then
justified. -- Ian
12:43 a.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 6/12/2013 10:27 PM, Ian King wrote:
On 6/12/13 6:15 PM, "Fred Cisin" <cisin
at xenosoft.com> wrote:
Two things I remember from back then.
A) We need it yesterday.
B) And this feature, the client just paid for it.
Properly planned testing might significantly reduce the amount of testing
needed. Badly designed testing does nothing but waste everybody's time.
Number of games of Solitaire played is not the right metric unless that is
the primary intended use (maybe it was?)
Yes, you're right, but there are assumptions you're ignoring (presumably
for simplicity's sake). Inadequately staffed testing does the best it can
within the constraints (calendar) prescribed by the corporate masters.
One can plan testing until one is bleary-eyed, but if the product changes
weekly because the marketing weenies say it must (with no change in ship
dates), testing will not be able to plan meaningfully. I'm not referring
to changes in implementation of feature X: I'm talking about, "We GOTTA
have this new feature! EVERYONE wants it!" I'm not joking, and having
dredged up these painful memories from my days at [insert humongous
software company here] I probably won't sleep well tonight.... I don't mind if dev works on the fast machines, as
long as when test
brings them a bug from a real-world, dog-slow machine, they don't respond
(wait for it): "It works fine on my machine." IMHO, mayhem is then
justified. -- Ian
Ben.
14 Jun
14 Jun
7:28 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Wed, Jun 12, 2013 at 06:15:30PM -0700, Fred Cisin wrote:
[...]
A well defined spec clearly identifies what range is
suitable. It is then the
developer's responsibility to work right for everything anywhere within that
range.
A magic code want that I can wave at a computer to make it do what I want would
be nice too. But just like a well-defined spec, I'm unlikely to ever see one in
my lifetime.
8:59 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
> A well defined spec clearly identifies what range
is suitable. It is then the
> developer's responsibility to work right for everything anywhere within that
> range.
On Sat, 15 Jun 2013, Peter Corlett wrote:
A magic code wanD that I can wave at a computer to
make it do what I
want would be nice too. But just like a well-defined spec, I'm unlikely
to ever see one in my lifetime.
But probably more likely than a well designed/written spec.
12 Jun
12 Jun
10:42 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 12/06/13 7:30 PM, TeoZ wrote:
I think many developers quit trying to have their apps
run on old
average hardware a long time ago and say so in the specs for it.
Which year's average hardware?
If they want to give users of "average hardware" a poor experience,
well, that's good business or bad business? For whom?
--T
-----Original Message----- From: Fred Cisin
Sent: Wednesday, June 12, 2013 7:14 PM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: Devs should be testing on slow machines - Re: What versions
of Linux (or UNIX) are good on old 486 boxes?
Elitist assholes who refuse to EVER use machines like what their customers
have, need to be canned. In well whale oil.
--
Grumpy Ol' Fred cisin at xenosoft.com
8:20 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
> [...] me to write Unix software for you, insisting
that I must only
> use a ratty old Pentium 4 with 512MB of RAM running a locked-down
> Windows XP that won't even let me install [...] is not acceptable.
BUT, until you DO, you don't have any idea how
your code performs for
those who are forced to USE it in such an environment.
While I think your general point is valid, I have trouble imagining
caring how "Unix software" performs on Windows of any stripe, on any
hardware.
A developer who "requires" better than a
ratty old Pentium 4 with
512M of RAM running a locked-down copy of Windows XP" writes software
that does not perform acceptably for those conditions.
May or may not. There are lots of cases where the development
environment is expected to be substantially beefier than the use
environment, and in many of them, there's nothing wrong with that.
I've written code to run on an 8051; I didn't actually do the
development on an 8051, and would consider anyone who did to be
completely bonkers. I once did some Android development; doing a
rebuild of the system called for a much bigger-and-faster system than
the phone which was the target platform. (Yes, we were rebuilding all
of Android on occasion; this was not just all-above-the-shiny app
development.)
If I am in a position of having to use such a machine,
then how
satisfied am I likely to be with software written on a better
machine, and NEVER REALLY TESTED on a machine like mine?
Now now. Just because it was written on a beefier machine doesn't mean
it wasn't tested on its target environment. That 8051 code was tested
on 8051s, not on the machine I wrote it on. The Android code was
tested on the target hardware.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
8:59 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
> If I am in a position of having to use such a
machine, then how
> satisfied am I likely to be with software written on a better
> machine, and NEVER REALLY TESTED on a machine like mine?
On Wed, 12 Jun 2013,
Mouse wrote:
Now now. Just because it was written on a beefier
machine doesn't mean
it wasn't tested on its target environment. That 8051 code was tested
on 8051s, not on the machine I wrote it on. The Android code was
tested on the target hardware.
true
If the testing is properly thought out, not just handing it over to
company cheerleaders.
I'm not really building a strawman so much as continuing my gripes about
Microsoft in the 1980s and 1990s, when they were peddling a lot of
saftware that was bordering on unusable on any configuaration other than
what they had used in development, and not tested at all on lesser
configurations. "yes, the CGA and EGA drivers will work. But we've never
tried them." The Windows 3.10 "Beta Test" consisted entirely of
soliciting praise from selected users. They did NOT want to hear about
any problems or issues (such as SETUP locking up the machine if it ever
hit a disk error)
Alternatively, and not suitable for your example, IFF development CAN be
done on the target platform, then doing at least SOME of it there will get
the programmers to have greater familiarity with the target platform
(again, MICROS~1 from 1985 to 1995)
--
Grumpy Ol' Fred cisin at xenosoft.com
13 Jun
13 Jun
2:56 p.m.
New subject: Devs should be testing on slow machines - Re: What versions of
If I am in a
position of having to use such a machine, then how
satisfied am I likely to be with software written on a better
machine, and NEVER REALLY TESTED on a machine like mine?
Now now. Just because it was written on a beefier machine doesn't mean
it wasn't tested on its target environment. That 8051 code was tested
on 8051s, not on the machine I wrote it on. The Android code was
tested on the target hardware.
8 Jun
8 Jun
12:44 p.m.
On Sat, Jun 8, 2013 at 7:19 AM, Tothwolf <tothwolf at concentric.net> wrote:
It does look like a nice board, but I can understand why you'd upgrade it.
On Sat, 8 Jun 2013, Ethan Dicks wrote:
> Going through stuff on the shelf and I've run across a couple of 486-based
> fully-integrated LCD/touchscreen machines... a Planar Systems box...
> 4x 30-pin SIMM sockets... with "double-sided SIMM" support... 32MB max
This was part of my gripe in the other thread. We just
don't have a full
fledged or embedded Linux solution that is currently maintained that will
work well with these type of embedded boards.
Agreed.
RedHat 5.2 was the last RedHat Linux distribution I
used on 486 based boards
and IIRC, 5.2 was the last of the pre-GNOME RedHat releases (this was around
the time the Pentium II came out and the ATX form factor came to market). I
later moved all the 486 based stuff I maintained that used Linux to Debian
2.1 (slink) and 2.2 (potato). I would think that most any Debian version
from 3.0 (woody) up through 5.0 (lenny) would work moderately well with a
486 based application, as long as you aren't trying to use something like
GNOME or KDE under X11.
Thanks for refreshing those bits.
A current Debian distribution such as 6.0 (squeeze) or
7.0 (wheezy) should
work on a 486, at least in console mode (I've not even considered trying to
run current X11 on one)...
I can't imagine even semi-modern X11 on a machine with less than 64MB. It
was bad enough 10 years ago with RedHat 9 at 256MB.
Your board with 4 30-pin SIMM sockets likely tops out
at 16MB. AFAIK, no one
ever made 8MB 30-pin SIMMs,,,
They did exist, but were quite rare and are still quite expensive.
http://shopper.cnet.com/ram-random-access-memory/simple-technology-8mb-dram…
The board has 4x 2MB SIMMs in it now. Yes. 2MB. Same idea
as an 8MB SIMM but with 4Mbit chips (1Mx4) instead of 16Mbit
chips.
I also have some all-in-one "keyboard Net-Stations" (sold with an
eye towards being DOS/Novell NetWare clients, AFAIK) with the
same CPU board as this Planar unit, so I occasionally consider
fabricating a run of 8MB SIMMs, but any time I get close to
starting the project, the amount of work that is intrudes and I
pick up a different task. I should at least reverse-engineer the
2MB SIMMs I have to see how they differ from 1MB SIMMs.
I've dug
up full specs on the Scantouch 3000 innards - PCM-4890 integrated
CPU board...
The PCM-4890 was made by Advantech and is a pretty good board. I'm familiar
with them and at one time used lots of them. It is a 5.25" sized "Biscuit
PC" form factor. Mostly due to the lack of a decent Linux distribution to
use with them, I swapped out all the remaining 5.25" Biscuit PC 486 boards I
was maintaining with Socket 7 boards about 8-10 years ago. A quick Google
search even turns up a pdf file of the user guide:
ftp://ftp.emacinc.com/Archive/m-4890.pdf
I'd previously found that, but thanks.
I would suggest avoiding "EDO" memory with
these and stick with a pair of
32MB 72-pin FPM Parity type SIMMs. EDO support with many of these types of
boards is problematic and was more of an afterthought and marketing ploy.
Samsung FPM SIMMs in particular work very, very well with these types of
boards.
Very good to know. There's one SIMM in it now - a 32MB EDO SIMM (based
on the part numbers on the chips). I know I have some 32MB FPM SIMMs
from when I was running maxed-out SPARC LX boxes.
despite the "Just go buy a new PC" responses
from many of the current
Linux distribution maintainers, it is cases such as this where you can't
just replace something with a cheap off-the-shelf consumer "PC".
Agreed. This is, essentially, an embedded platform that just happens to
architecturally resemble an early-to-mid-1990s desktop machine (so that
it can run DOS and Windows without too much of a stretch).
I did find a Microsoft license key sticker under a mounting bracket on the
back - Win98 SE Embedded, so it was sold for running that. They must
have upgraded the touch screen hardware or else the forum posts I was
reading about this device were inaccurate with respect to pixels/mickeys
and Win98 touchscreen drivers.
Thanks for the info!
-ethan
5:23 p.m.
At 12:44 Uhr -0400 8.6.2013, Ethan Dicks wrote:
My Quadra 700 boxes have 16 MByte 30 pin simms. I got them a few years ago
Your board
with 4 30-pin SIMM sockets likely tops out at 16MB. AFAIK, no one
ever made 8MB 30-pin SIMMs,,,
They did exist, but were quite rare and are still quite expensive. from the US for (IIRC) ~ 10 USD / piece.
hauke
--
"It's never straight up and down" (DEVO)
8:46 a.m.
On 8 June 2013 06:38, Ethan Dicks <ethan.dicks at gmail.com> wrote:
Going through stuff on the shelf and I've run
across a couple of
486-based fully-integrated LCD/touchscreen machines. One is a SCAN
Corporation SCANtouch Model 3000, the other is a Planar Systems box.
Similarities include limited memory (2x 72-pin SIMM sockets on the
SCANtouch or 4x 30-pin SIMM sockets on the Planar, both with
"double-sided SIMM" support, so 64MB or 32MB max respectively), one
ISA socket (no PCI), onboard NE2000, serial, parallel, 2.5" PATA
disk... these systems will run MS-DOS, of course, and Win95 (the touch
screen on the SCANtouch is old enough that it apparently doesn't work
with Win98 mouse drivers, according to forum posts I dug up). I
really don't care about running DOS or Win95 on a touchscreen/LCD
machine
Just a thought, but it would be genuinely interesting, IMHO, to try to
get either one of the 1980s pen-based OSes - Go, Penpoint, Magic Cap,
anything that might work on the hardware - working on such a box, or
failing that, original PenWindows.
--
Liam Proven ? Profile: http://lproven.livejournal.com/profile
Email: lproven at cix.co.uk ? GMail/G+/Twitter/Flickr/Facebook: lproven
MSN: lproven at hotmail.com ? Skype/AIM/Yahoo/LinkedIn: liamproven
Tel: +44 20-8685-0498 ? Cell: +44 7939-087884
10 Jun
10 Jun
2:29 a.m.
On Sat, 8 Jun 2013 23:19:40 +0200
Hauke Fath <hauke at Espresso.Rhein-Neckar.DE> wrote:
NetBSD 1.x would most certainly be comfortable on the
machine.
Yes. Thats my recommendation too. Even newer versions would be mostly
OK. I run the then current release (4.0?) on a pair of 486 laptops with
4 MB RAM a few years ago. Though I had to build custom, striped down
kernels and 4 MB was _very_ tight.
--
\end{Jochen}
\ref{http://www.unixag-kl.fh-kl.de/~jkunz/}
7:41 a.m.
New subject: NetBSD 1.x - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On 10/06/13 2:29 AM, Jochen Kunz wrote:
On Sat, 8 Jun 2013 23:19:40 +0200
Hauke Fath<hauke at Espresso.Rhein-Neckar.DE> wrote:
4.0? But we are talking about NetBSD 1.x.
I am running 1.4.1, netbooted, on a MicroVAX II (9 MB RAM and not
terribly speedy) so it should be quite okay on a '486. It would run OK
in less RAM, too.
--T
NetBSD 1.x would most certainly be comfortable on
the machine.
Yes. Thats my recommendation too. Even newer versions would be mostly
OK. I run the then current release (4.0?) on a pair of 486 laptops with 4 MB RAM a few years ago. Though I had to build
custom, striped down
kernels and 4 MB was _very_ tight.
5:49 p.m.
New subject: NetBSD 1.x - Re: What versions of Linux (or UNIX) are good on old 486 boxes?
On Mon, 10 Jun 2013 07:41:53 -0400
Toby Thain <toby at telegraphics.com.au> wrote:
4.0? But
we are talking about NetBSD 1.x. I intended to say: NetBSD 1.x is a good fit for a
486DX33.
But something newer like 4.0 should work OK too, albeit a bit slower
and it will need more RAM.
NetBSD 1.x would most certainly be comfortable on the
machine.
Yes. Thats my recommendation too. Even newer versions would be mostly
OK. I run the then current release (4.0?) on a pair of 486 laptops with I am running 1.4.1, netbooted, on a MicroVAX II (9 MB
RAM and not
terribly speedy) so it should be quite okay on a '486. It would run OK
in less RAM, too.
For sure. Around 1999 / 2000 a friend used a VAXstation 2000
running
NetBSD and thttpd as webserver for his personal homepage.
--
\end{Jochen}
\ref{http://www.unixag-kl.fh-kl.de/~jkunz/}
8:09 a.m.
How would early Linux or WhateverBSD compare to, say, some of the
commercial Unix implementations that were available for the PC in the
early-mid 90s? Assuming a copy could be tracked down at a reasonable
price, it might be fun to try out UnixWare 1.x or 2.x or possibly even
Xenix or SCO UNIX.
Mike
11:17 a.m.
On 6/10/13, Michael Kerpan <mjkerpan at kerpan.com> wrote:
How would early Linux or WhateverBSD compare to, say,
some of the
commercial Unix implementations that were available for the PC in the
early-mid 90s? Assuming a copy could be tracked down at a reasonable
price, it might be fun to try out UnixWare 1.x or 2.x or possibly even
Xenix or SCO UNIX.
You could also have a look at Coherent or Minix, both of which should
run well on a modest machine of that era. (although arguably they
aren't quite "true" Unices)
11:27 a.m.
Microsoft's first Xenix for PCs ran on 5160 with 640k of RAM and 15M hard
drive. Performance was . . . as expected
It was capable of running with 256K? and 10M drive if you didn't want to
modify (solder jumpers) on the stock [Xebec] controller.
Linux is bulkier.
6:44 p.m.
On 10/06/13 8:09 AM, Michael Kerpan wrote:
How would early Linux or WhateverBSD compare to, say,
some of the
commercial Unix implementations that were available for the PC in the
The open source distributions are likely to have more modern, better
supported toolchains.
--T
early-mid 90s? Assuming a copy could be tracked down
at a reasonable
price, it might be fun to try out UnixWare 1.x or 2.x or possibly even
Xenix or SCO UNIX.
Mike
11 Jun
11 Jun
12:31 a.m.
How would
early Linux or WhateverBSD compare to, say, some of the
commercial Unix implementations that were available for the PC in
The open source
distributions are likely to have more modern, better supported toolchains.
Right. Like the way I once tried to report a bug in NetBSD's gcc and
NetBSD wasn't interested because it was an externally-maintained piece
of the system and gcc wasn't interested because I wasn't reporting
against bleeding-edge code. I've had better support from commercial
vendors, which is really saying something.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
8:03 a.m.
On 11/06/13 12:31 AM, Mouse wrote:
Actually, for many purposes they're better despite that. (For a few
purposes, even, they're better because of it.)
It's great to have more than one option, isn't it?
That's good news, because it means they have lifted their game since the
1980s when gcc was born out of the opposite situation.
--T
How would early Linux or WhateverBSD compare to, say,
some of the
commercial Unix implementations that were available for the PC in
The open source
distributions are likely to have more modern, better supported toolchains.
Right. Like the way I once tried to report a bug in NetBSD's gcc and
NetBSD wasn't interested because it was an externally-maintained piece
of the system and gcc wasn't interested because I wasn't reporting
against bleeding-edge code. I've had better support from commercial
vendors, which is really saying something.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
10 Jun
10 Jun
8:03 a.m.
On Sat, Jun 08, 2013 at 11:19:40PM +0200, Hauke Fath wrote:
At 1:38 Uhr -0400 8.6.2013, Ethan Dicks wrote:
[Unix on 486 ISA board with touchscreen]
About two decades ago, I was given a copy of NetBSD 1.1 on CD. I installed it
on my Amiga 4000/040, and it more or less worked. Except for one detail: no
matter what I did to carefully shut it down, when I booted it back up its
filesystem was utterly trashed. It was certainly not a hardware or disk fault,
as the same system ran (in fact still does run) AmigaOS and Debian m68k just
fine.
That's about the last time I bothered with NetBSD. If I want something small
for an appliance, I usually reach for either a custom Debian install or
OpenBSD.
Helpful suggestions wanted.
NetBSD 1.x
would most certainly be comfortable on the machine.
4087
days inactive
4102
days old
139 comments
34 participants
participants (34)
-
abuse＠cabal.org.uk -
aek＠bitsavers.org -
ajp166＠verizon.net -
alexandre.tabajara＠gmail.com -
als＠thangorodrim.de -
amh＠POBOX.COM -
ard＠p850ug1.demon.co.uk -
b4＠gewt.net -
bfranchuk＠jetnet.ab.ca -
blstuart＠bellsouth.net -
cclist＠sydex.com -
cisin＠xenosoft.com -
dave.g4ugm＠gmail.com -
derschjo＠gmail.com -
ethan.dicks＠gmail.com -
fraveydank＠gmail.com -
geneb＠deltasoft.com -
hauke＠Espresso.Rhein-Neckar.DE -
ian_primus＠yahoo.com -
IanK＠vulcan.com -
jason＠smbfc.net -
jbmcb1＠gmail.com -
jkunz＠unixag-kl.fh-kl.de -
joachim.thiemann＠gmail.com -
lproven＠gmail.com -
mcguire＠neurotica.com -
mjkerpan＠kerpan.com -
mouse＠Rodents-Montreal.ORG -
other＠oryx.cc -
snhirsch＠gmail.com -
spc＠conman.org -
teoz＠neo.rr.com -
toby＠telegraphics.com.au -
tothwolf＠concentric.net