I'm trying to find a way to get a modern file-transfer protocol
program into my Nova 4/X.
I'm close to abandoning the Kermit source I have for Data General
RDOS... I got it from U Columbia's Kermit site; it's the FORTRAN5
output of a RATFOR preprocessor, there's no RATFOR source, and
though it compiles, it seems to not talk to the console. Since
it's algorithmic output it's very hard to read.
On simulogics.com is a KERMIT in dump format (similar to LBR
files, functionally) but it's binary.
I've got a decent FORTRAN IV compiler, what I can't find is a
FORTRAN IV kermit. There's a bunch of mainframey fortran kermits,
but no one seems to think about portability (eg. putting all the
console/serial I/O in stub functions) and none are careful to use
portable libraries, and all that.
Worse, Kermit seems to have abandoned the idea of a bootstrap
system (except for one unix version) with utterly base-minimal
compliance and features.
Anyone have code?
I am considering writing a proprietary system that transfers
fixed-length line-oriented blocks consisting only of ASCII numbers
and line control; real FORTRAN IV doesn't do text (some
proprietary libraries do).
The only thing I have now is minicom as terminal emulator.
sending/receiving paced ASCII67 text (thanks Jules for the tip on
ascii-xfer).
Dear Fellow ClassicCmp'ers,
I want to apologize for my recent behavior here. I'm sorry that I was
pushed over the edge by John's innocent suggestions of using some more
modern technology. I was clearly in the wrong state of mind when I wrote
that the purpose of my life was in using old technology and nothing new
and that I would shed blood (whether my own or anyone else's) over it.
Of course no piece of technology, old or new, is worth shedding blood over.
I don't know what came over me when I wrote that. Those who know me more
closely know that I've been leaning more and more over to alternative
spirituality and metaphysics and that technology of any kind, old or now,
currently plays a very small part in my life. So certainly I'm not going to
kill over it.
But I promise I won't bring up any more off-topic stuff here. Technology
can be fun, and even I am open to some modernisation. Like I'm trying to
transition from a VT320 to a VXT X11 terminal. To bring this back on topic,
I'll soon be working on a MOP server for 4.3BSD-Quasijarus for netbooting
these VXTs, and getting X11R4 to run on my 4.3BSD-Quasijarus hosts. Fun
fun fun.
Blessed Be everyone,
MS
>From: "Barry Watzman" <Watzman at neo.rr.com>
>
>
>I remain convinced that USB, not parallel, is the "right" way to go for
>interfacing a Universal Floppy Disk controller.
>
Hi Barry
I tend to agree with you. We should be thinking this way.
Still, I believe that development work should be done in
an environment that is handy and convenient. The USB is
just a machine interconnect. One just has to keep in mind
what the final product will be like. In other words, don't
lock the design into one specific format.
The only issue I have with USB is that it requires drivers
for each machine it is connected to. These have to be
specific to the USB device we use to interface with.
RS-232 is generic enough that we could run things from
text files using simple terminal modes on almost any machine.
I just wish that USB was that simple.
Dwight
>From: "Ashley Carder" <wacarder at usit.net>
>
>I do not own a variac, so I will need to get one of these to
>properly reform the TU56 capacitors. I see several different varieties
>on eBay. What would I need exactly?
>
>Ashley
>
>
Hi
You can always use a voltage source and a limiting resistor.
You don't need a variac ( although they are real handy ).
Dwight
Hi
One other thought is that one should have a way to
transfer the raw data from what ever media you have.
This allows one to have redundent information stored
on the same media. If you depend on normal file systems,
you run the risk of the file system not allowing you
to access the data simply because a small part is
damaged. I think most have been using normal file tools
that are not dealing with things in this sense. Most
of these types of systems have a weak link problem.
Dwight
> ...and you never have to take your hands off of the keyboard.
in most environments I work in the cord is not long enough to reach the mens room...so I don't think "never" really applies....
On Thu, 10 Mar 2005 20:13:19 -0800, Zane H. Healy <healyzh at aracnet.com> wrote:
> It would have been nice to see it continue, but after Apple bought
> NeXT rather than Be, Inc. they sort of started to run out of steam.
> I might be typing this on a dual 2Ghz G5 running Mac OS X, but I
> can't help but thinking how much better the system would be if it had
> been BeOS based, rather than OPENSTEP based. The good thing about
> them buying NeXT was getting Steve Jobs back, but BeOS was a far more
> efficient OS. And yes, I've run OPENSTEP and BeOS on the same system.
I know what you mean, but after a few years, and actually building up
an old PowerMac 7300/166 I got given for free with a load of bits off
eBay - paying money for hardware, most unlike me! - and actually
getting to *use* Mac OS X, I came to the conclusion that Apple did the
right think.
BeOS was gorgeous. My favourite PC OS ever. Everything I used to
really like about systems like Acorn RISC OS, Psion EPOC and so on, or
AmigaOS.
Like those early proprietary 16-bit and 32-bit OSes, BeOS was small,
blindingly fast, had a clean, simple GUI and came with an assortment
of useful apps. It had a small but competent portfolio of useful
software.
Unlike most of them, it had networking and Internet capabilities built
in, supported memory protection, virtual memory and multiprocessor
hardware.
And unlike any of them, it ran on cheap generic hardware,
What's not to like?
But whereas it would have made for a small and blisteringly fast Mac
OS X, which with a bundled or integrated copy of SheepSaver would have
happily run Classic MacOS & all its apps under the new OS, that's
*all* it would have made.
BeOS was unique. POSIX-like, but not Unix. Proprietary GUI but one
with no unique advantages. Traditional C/C++ development model.
Proprietary environment with all that that entails with programmers
having to learn it.
NextStep/OpenStep represented something different. Yes, all the
clutter and kludge of Unix, but:
- a new, simple development system using Objective C, Interface
Modeller & a vast & comprehensive class library (arguably the best
development system and tools for GUI apps there has ever been)
- deep thorough object orientation
- based on Unix, so familiar to millions of existing programmers,
users and techies
- based on /open source/ Unix, freeing Apple from the burden of
developing & maintaining the entire OS on their own
- a solid open networking system build deep into the OS
- designed, like BeOS, by former Apple engineers
- unlike BeOS, commercially successful (in a small way) for a decade
in the open market on various hardware and software platforms
- a new and revolutionary GUI based on Display Postscript (now
Display PDF, but it's the same difference) - one that offers
compelling advantages for designers, users and developers, with
powerful new capabilities, but that's already based on known,
widespread technology.
Up front, BeOS would have made a better successor Mac OS. NextStep had
to go through a lot of painful development to make it look and work
like MacOS and the result was initially slow and clunky.
However, /every/ successive release has made it sleeker and faster,
bucking the trend of every other OS in the industry.
But that's not what's important. What is important is that along with
bringing over all the Mac developers with Carbon, it brought in lots
of Unix developers with the Mach/BSD underpinnings, and tempted PC C
coders with a different dialect of C, and Java coders with good Java
support. Then to all of them it offered, /for free in the box/, the
most remarkably polished and effective GUI development environment
there is.
Getting programmers behind a new OS is essential. It's paramount.
That, I think, was the basis of Apple's decision.
The result? OS X is by almost any standard the single most polished,
friendly, usable, attractive, flexible desktop OS on the planet, and
it's attracting lots of new and interesting software development. But
by the same token, it's Unix, so if you want to drive it with EMACS
and X tools from a /bash/ prompt, you can.
BeOS could never have been that.
--
Liam Proven
Home: http://welcome.to/liamsweb * Blog: http://lproven.livejournal.com
AOL, Yahoo UK: liamproven * ICQ: 73187508 * MSN: lproven at hotmail.com
> I first programmed in high school in FORTRAN on an IBM
> 1620 and shortly thereafter....
I read this as "...on an IBM **IN** 1620..." and thought we really had a winner on the oldest programmer....
...oh, the coffee is ready....BRB....
I'm 56.4166666... or thereabouts.
I first programmed in high school in FORTRAN on an IBM
1620 and shortly thereafter in Autocoder on a 1401,
around 1965-66. I still remmeber reading in a column
in Popular Science that "scientists predict that in
the future computers would be the size of shoe
boxes..." Since I was a teenager and was already an
expert on computers (and most everything else), I
laughed at this silly idea... until 10 years later.
Bob Stek
Saver of Lost Sols
