Now, please take a look at my additional comments embedded below. I know
they make this message long, but I dislike posting my remarks separately
from the context to which they refer.
Dick
----- Original Message -----
From: <jpero(a)cgocable.net
To:
<classiccmp(a)classiccmp.org
Sent: Saturday, May
06, 2000 3:37 PM
Subject: Re: I wrote 'Nuke Redmond'
snip
>
> > Well that works out nicely for you, but for many of the rest of us, M$
> > products are designed to solve problems we don't have, and totally
fail
> > to address the problems we *do* have.
It's designed for the drooling
> > masses *only*, it's absolutely not good for normal utility use by
nerds.
That's because use by nerds is seldom normal.
It's the fact that the
drooling masses have money that makes these computers
cheap and ubiquitous.
Let me switch this comments around:
Nerds usually know what they doing correctly with common sense. The
people out in the wild usually breaks M$ stuff because it has too
much ways to happen.
My experience with nerds would suggest that they believe that, but I don't.
Like Allison's comments about user space kept away from OSes and this
limits the blowups if a typical user make an miss is spot on.
I've never experienced this before, but I'm having difficulty parsing this
sentence.
Best devices are those very limited functions that does their
intended functions and no more. Examples: electronic reminder and
scheduler devices with personal data like birthdays, appts etc.
Another one that would do very well are web surfer, i-opener, email
devices such as Vtech w/modem built in and like for common users.
I'm inclined to agree, but that only applies to devices wherein NO provision
for future growth is built in. I'm not familiar with the iopener, nor am I
a particular fan of the way in which the "little" electronic notepads are
made or used.
What is missing is real innvoative user-interface (ala i-opener
style) for office applications and hide the OS and be Untouchable.
Again, I'm more inclined to wait until a product has been proven entirely
satisfactory before I'm inclined to buy another from the same vendor. Most
of y MS stuff is free, so I can't complain.
win9x is more suitable for users who should know bit more clueful
with common sense to get help and not like "mmm, try this one?"
like Mr Simpson would do and push a button. Boom!
Frankly, I don't like the GUI. It has advantages, primarily in that it
purportedly provides a monolithic structure for device drivers, thereby
making a device accessible to all properly written applications. I liked
things better when graphics happened only when you needed graphics, and the
rest of the time the machine was strictly used from the command line.
win9x, Unices, Vaxens, Linux is more of nerd's dormain and clueful
users who know better to call for help first if some kind of problems
becomes out of their depth understanding how to deal with it.
I can't agree with that! It's essentially a contradiction of what
you've
previously stated. Nevertheless, I'd say the the UNIX and others of that
ilk were designed for use by and for nerds, from the standpoint that
producing software is useful work. That's only true if you're a software
vendor. If you're in the business of selling tires, or of making them,
generating software is overhead that you'd like to avoid. IIRC, there were
a couple of UNIX-like systems for the VAX, not to mention VMS, which was
quite respectable, particularly for production tasks for which UNIX was
considered most unsuitable. From the console, or farther away, it didn't
look much different than UNIX. I've never liked things that said DIGITAL on
them, hence never got too friendly with VMS.
I find NT more of configuring and using without enough understanding
and things go bump in the night often, tons of BODs and kernel dumps.
I find this takes full understanding how hardware interacts with this
NT and use software drivers correctly to make sure it's reliable
enough. Also doing configurations in software especially in
networking stuff takes lot of understanding because I find reading
the wood ludicious especially without looking at whole picture why
OS or software barfed. Probs I find are: with hardware
mis-configuration and shoddy installation, 4/10 of that in software
or sloppily written drivers and/or software, 1/10th of that in true
hardware failures flaking out and low quality hardware parts or wrong
type used for given OS.
I think this goes back to the VERY ambitious attempt to establish a standard
for Plug-n-Play, among other things. Unfortunately, when a committee tries
to build a standard, it normally has lots of clauses and no teeth.
Consequently, the Plug-n-Play became Plug-N-Pray. That's why the interrupt
steering is different each time you go and fiddle with drivers, hardware,
etc. It's almost by accident, and only by accident that I get through the
installation of some video drivers. They always seem to "reach out and
touch" something when least expected to do so.
> > I recently installed Win98SE in the hope that it would be more
reliable
> > for web browsing than Win98 was (on my
machine, W98 was very flakey
about
> > connecting to the net, and the video driver
ignored my settings and
always
> > ran in 640x480x16 mode), but the stupid
thing has already locked up
the
> machine
several times while sitting totally idle.
Check that "power management" and cooling, insure memory is good.
Even top of 1 grand dollars Diagnostic stuff aka the best of world
will not detect this. I know, because I fixed few times like this
and had to rely on softwares and good working hardware to cause DUT
to act up and alteratively turn off and on one settings in bios one
at a time to flush out the problem. Had to do that on a motherboard
using win9x and bunch of games, sound card. This problem finally
solved it by swapping out pair of soldered 512K pipelined cache
chips. When I turned off that faulty cache part in bios, everything
went plum and sweet, got even better when CORRECT part got replaced.
I've yet to see a respectable set of diagnostic software for the PC. With
as much memory as most people are expected to have, I imagine a thorough
diagnostic of read/write memory would take about 50 years. I've yet to see
a diagnostic that's capable of detecting faulty external cache. Most of the
diagnostic programs I have are incapable of testing current generation
hardware. No matter what steps I take, the failures are always undetected
until after the system is delivered.
Software, likewise, seems to go largely untested throughout the industry.
For several years, it was almost impossible to buy a backup utility for
Win9x, because none of the available utilities would work. I tested
several, as part of a contracted task, and found that one (then) major
vendor's software wouldn't finish its task. Another would back-up, but
wouldn't restore. There were lots of other problems, but I find it criminal
that software vendors are allowed to sell products that don't work at all.
Even the unmanageable IRQs got shifted when I toggle certain parts in
bios menu. Many didn't realize the power of managing this by this
route.
I make it a practice to disable or even physically remove sound hardware.
It uses three interrupts and two DMA's under worst-case conditions and I
prefer my computers to be quiet. I keep one sound-equipped machine
available in case I need to process sound (.dat) files, but normally avoid
having it installed.
> This is because you're not installing your video driver correctly. It
often
> means you must delete the driver that appears in
"Device Manager" before
> proceeding. This is often a problem when W9x knows about several
versions
> of the same hardware. That's also why they
refer to the common
installation
> mode as "lug-n-pray." It defaults to
the lowest common denominator. If
you
> then install the driver precisely written for
your version of the
hardware
in question,
it will work fine. If you're off, even by one bit, your're
stuck with the lowest common denominator, however.
Any hardware drivers not just for video no matter what. Once in
awhile drivers breaks or win 9x actually lost a soundcard, happened
to me few times before and two programs didn't give any errors other
than both refused to start which deepens the mystery till I check DM.
> I feel forced to point out that although the W9x isn't even near to
being
> perfect, it's one heck of a lot better for
the "average" home user than
any
> version of *NIX or any other commercially offered
OS. The half-day
install
> for OS/2 (admittedly last attempted when OS/2 was
new) would certainly
> discourage these "average" users. Moreover, Win9x wouldn't cost
<$100
if it
> weren't for the fact that the
"average" user buys a computer and the
> software he wants. Not only would Win9x not exist, but if it did exist
in
> spite of the lack of home users, it would cost
like VMS. The hardware
would
cost quite a
bit more, too.
@100 bux is upgrade version for any 9x always. Ones that is true
version without buying a computer or certain items is over $300.
And another way is to buy whole new system which is too much for me.
"List" price for the first-install version is $249, though I doubt the
discount chains are charging that much. I recently saw the SE update for
$85. Next time I'm at Best Buy, I'll check on both versions.
That is why I'm angry at M$. And still running on 95a upgrade version.
Well, it's no wonder that you're irritated with M$. That 95A version was
a
sizeable step back from the betas. I'd upgrade to '98SE if possible. I
have had really much better results with /98 than with 95, even with OSR2.
What's more, I'd avoid loading Netscape. I've found that when somebody with
Netscape loaded on their portable drives up into my driveway, the systems in
the house start to act squirrely. It not even necessary that the foreign
notebook be attached to the net.
Maybe Windows isn't for you. I use it
because it's hard not to. I have
half a dozen LINUX versions none of which has been left installed for more
than a day or two, and they wouldn't meet my needs. Likewise, I've not
gotten a comfortable feeling with SCO, UnixWare, etc. for the '386 and up
types.
> Wizard