You're right on the money, here. It depends almost entirely on what your
goals are or may be. ( Perhaps, like many of us, you don't know, exactly,
what your goal(s) at any given time might be.) I try to fool myself into
thinking that I do things involving computers and circuits to enhance/extemd
my ability to earn a living. Maybe that's true. However, using a task like
implementing a current strategy or algorithm on an old piece of hardware
tends to extend one's thinking, which is also of benefit to one's ability to
earn a living.
If I had the task of building a homebrew computer, I could probably fit the
whole thing on a single or maybe dual-width VME card. That's not much more
area than an S-100 board if you leave out the regulators. The way this
would have to come about, is that I'd build a processor<=>memory interface
and an interface to a bridge controller for mass storage. If the processor
didn't have I/O on board, which most of the ones in which I have interest
relating to building a system from scratch do, then a single high-speed link
to a system which had conventional resources would come next. Over time,
I'd migrate the mass storage interfaces into my computer and leave out the
bridge adapter. Likewise the external comm link(S). After that, it's
anyone's guess. It depends on what goal(s) I am chasing at the time.
Dick
-----Original Message-----
From: Allison J Parent <allisonp(a)world.std.com>
To: Discussion re-collecting of classic computers
<classiccmp(a)u.washington.edu>
Date: Sunday, April 04, 1999 5:26 PM
Subject: Re: homemade computer for fun and experience...
><I'd thought of that, but if you're not careful, your 'homebrew computer'
><contains ready-built I/O cards only. That's no so educational as building
><them yourself.
>
>Depends on what your goals are. For me building yet another serial or disk
>card is tedious and repetition. I'd rather experiment with advaced
software
>on old style but enhanced platforms.
>
><Sure. But the IIRC, this thread started with somebody wanting to make
><their first homebrew machine. IMHO, this should be a fairly simple
><machine - say a Z80 + maximum of 64K memory. Adding lots more memory,
><MMU, etc can come later. The first machine should be something that is
><likely to work first time.
>
>No arguement. As a first pass machine static is usually the least painful
>to deal with as are simple serial IO. The usual first time builder wants
>more than their skills can generally support.
>
>Allison
>
Well, I'm inclined to agree with you. I've long felt that if one needed to
spend time playing games, what was really needed was another job.
What's happened over the years, however, is that people, having seen what a
computer CAN do as shown in games, etc, have actually found ways to make
data easier to interpret, and perhaps to add meaning and emphasis to a
presentation, making it more persuasive, if not more informative, by using
the graphical capabilities of a computer.
As for the terminal vis-a-vie the color graphic display, I find the terminal
quite limiting. However, most of the stuff I do is unaffected by these
limitations. It's just the drafting work that would be limited, and perhaps
the interpretation of the results from the various simulators I use.
However, since I've got a farly hot system with which to do the drafting,
the CP/M box doesn't really need graphics, does it?
If you use a terminal, that's what you've got. There were, for a time,
attempts made at graphic terminals. These failed, however, because there
weren't standards on which they could base their usage. Consequently, if
one didn't have certain hardware, there were limitations on the software he
could use. Today, that's not the case, as EVERYONE has a PC clone with at
least 1Kx768 pixels in 256 or more colors. EVERYONE has fairly ample
resources, comparable, perhaps, to the sum of all the computer resources in
the world when I went to college.
Today, no one would normally consider a CP/M box for "useful" and essential
work. For mental masturbation, speculation about would could, would, or
should (note the subjunctive) have been, CP/M works just fine. If you have
a construction business, or a medical office, there's no reason why CP/M
can't work today every bit as well as it did back 20 or more years ago.
However, aside from the notion of operating on an "antique" there's no
reason to do so. It does serve to keep one grounded in what has changed
over the years and what hasn't.
Dick
-----Original Message-----
From: CLASSICCMP(a)trailing-edge.com <CLASSICCMP(a)trailing-edge.com>
To: Discussion re-collecting of classic computers
<classiccmp(a)u.washington.edu>
Date: Sunday, April 04, 1999 12:58 PM
Subject: Re: homemade computer for fun and experience...
>><ISA cards on a single s-100 board. This would certainly be cheap enough
i
>><most cases, to warrant such an effort. The software might get to be a
>><problem, though.
>
>>Yes, and therein lies the "problem" for the average user.
>
>This is why I prefer simple serial console. No software to write,
>no memory mapped video to take a chunk out of usable memory, no
>memory banking schemes, no dependence on finding monitors of a
>particular scan rate or interface. Just a simple I/O port!
>
>Admittedly, many folks seem to believe that a computer must have
>a video generator to be a "computer". My CP/M experience started with
>Model 33 Teletypes and blinkenlights front panels, so perhaps I'm the odd
man
>out here.
>
>--
> Tim Shoppa Email: shoppa(a)trailing-edge.com
> Trailing Edge Technology WWW: http://www.trailing-edge.com/
> 7328 Bradley Blvd Voice: 301-767-5917
> Bethesda, MD, USA 20817 Fax: 301-767-5927
Keep in mind, folks, that some of us remember when there simply weren't
computers which we could use. Back when I was a boy, even the US GOV
couldn't afford a computer capable of what my smallest simplest Windows9x
system is capable. The fact that I don't use it for pure number-crunching
is a recognition that there are other things equally deserving of the
benefits of advancing technology. If a picture is worth a thousand words,
then why type them if it's easier to draw the graphic representation.
Moreover, if you have to read my 1000 words (about a page) and then draw
yourself a graphic representation to understand and digest it anyway, why
not stick to the graphic? A simple graphic can save both ends of a
communication time, effort, and embarassment from when there's a
miscommunication.
If you tend to use your hands when you're explaining something, you probably
know what I mean.
Dick
-----Original Message-----
From: jpero(a)cgocable.net <jpero(a)cgocable.net>
To: Discussion re-collecting of classic computers
<classiccmp(a)u.washington.edu>
Date: Sunday, April 04, 1999 5:03 PM
Subject: Re: homemade computer for fun and experience...
>Date sent: Sun, 4 Apr 1999 16:23:49 -0600
>Send reply to: classiccmp(a)u.washington.edu
>From: "Richard Erlacher" <edick(a)idcomm.com>
>To: "Discussion re-collecting of classic computers"
<classiccmp(a)u.washington.edu>
>Subject: Re: homemade computer for fun and experience...
>Originally to: <classiccmp(a)u.washington.edu>
>
>> Well, I'm inclined to agree with you. I've long felt that if one needed
>> to spend time playing games, what was really needed was another job.
>>
>> What's happened over the years, however, is that people, having seen what
>> a computer CAN do as shown in games, etc, have actually found ways to
make
>> data easier to interpret, and perhaps to add meaning and emphasis to a
>> presentation, making it more persuasive, if not more informative, by
using
>> the graphical capabilities of a computer.
>
>Beauiful point and to point that out, we're still doing it even way
>back from stone and clubs days. Graphics on any media is easier
>to understand than stackful of numbers and words even spoken!
>Even when I was in elementery and HS school we generate
>numbers and words into graphics respentions or vice vesa. That
>was in CP/M to 386 days but didn't use computers.
>
>>
>> As for the terminal vis-a-vie the color graphic display, I find the
>> terminal quite limiting. However, most of the stuff I do is unaffected
by
>
>Terminal is out because one would have to code so much that you
>really made another i/o out of software. I'm at that mental stage
>now. Better start in hardware then put bit of homebrew s/w in to
>get started.
>
>> Today, no one would normally consider a CP/M box for "useful" and
>> essential work. For mental masturbation, speculation about would could,
>> would, or should (note the subjunctive) have been, CP/M works just fine.
>> If you have a construction business, or a medical office, there's no
>> reason why CP/M can't work today every bit as well as it did back 20 or
>> more years ago. However, aside from the notion of operating on an
>> "antique" there's no reason to do so. It does serve to keep one grounded
>> in what has changed over the years and what hasn't.
>
>We're still writing characters/numbers and users are using so
>powerful computers *because* it's easier for them to see in
>graphics form and graphics items to operate at or in to get basic
>functions done besides writing and putting numbers in. That eat up
>more power just to do that. Crunching numbers and doing massive
>moves in graphics, is one thing I accept. And, decent windowing
>that does quickly and low on resources requirements does get job
>done bit easier that about it as I know so far.
>
>> Dick
>>
>Wizard
<Yes, although you could consider "P-code" to be a lower level.
It is lower level.
<> But why would anyone want something that was microcoded to run Pascal?
<
<Speed mostly, that and ease of use with compilers.
P-code is not Pascal. It's a platform that all Pascal P-compilers grind
code to. Speed wise it was slow as the Pcode was interpreted on all but
the WD Pascal microengine. Why was it done? In the late 70s Pascal was
emerging as a teaching language and it was highly standardized. Most
small systems had the resources to run it as native compiled but the
authors (UCSD) decided that portable Pcode would allow more platforms
to run it as the Pengine was easier to code than a whole compiler. The
additional part was the P-system was an integrated system with all the
tools (all written in pascal) for editing, compiling and running programs.
It was a good teaching environment as it isolated the user form the machine
very well. the entire environment was menu driven and integrated.
I have a NS* that runs the z80 version.
Allison
<trouble with it also, but can peal it apart. So basically you're stuck
<having to read it using Windows, which quite bluntly strikes me as stupid,
I have winders and the problem is I have to buy the dang viewer...Not me.
Fortunately having worked with 741, 744 and 7441 power supply and regulators
this (740) is not too strange for me.
Allison
<> OH, "SROCTH" is Some Rare Old Computer To Hack.
<
<I can name a number of people here who easily have enough clues to keep
Clearly I'm not refering to them.
<I am, however, a little worried with what happens to some of the machines
<sold on E-bay. I suspect that some of them go to people who want them but
<haven't a clue how to look after them.
That is the exact audience I refer to.
<Ouch!. It's bad enough when a fault causes melted PCB traces (Some SMPSUs
<suffer from this).
This was abuser installed. I've cleaned it up a lot. The current bug
fixed in the PS was the chopper shorted and then fried the SCR used for
crowbar.
We are up again. Now back to the stuck bit.
Allison
<Admittedly, many folks seem to believe that a computer must have
<a video generator to be a "computer".
Retrorevisionist PCism. Though a VC8E is a definate computer thing.
< My CP/M experience started with
<Model 33 Teletypes and blinkenlights front panels, so perhaps I'm the odd m
<out here.
Same here. It's seems silly for me to recreate the logic and programming
of my vt100 when I have a vt100 that works better than any I could put
together. There is advantages to useing a PC keyboard and monitor as they
already take space on my desk but then again I can provide a serial port
and use the PC for a terminal.
Allison
<>Been there and done that.
<>Also, there are no less than two articles on how to go from s100 to ISA.
<>
<There's a substantial reach from an article to a solid and working circuit
<on a board capable of actually supporting the functions. This is
The articles actually are descriptions of circuits that do work.
<particularly true since, now, the user is required to understand the inner
<workings of his own machine as well as those of the card he wishes to use.
<Having the usual inkling about ISA is not always sufficient.
That was true then. Unlike PCs with plug and pray configuring a s100
system meant you generally had knowledge and documentation or were plain
crazy.
<><ISA cards on a single s-100 board. This would certainly be cheap enough
<><most cases, to warrant such an effort. The software might get to be a
<><problem, though.
<
<Yes, and therein lies the "problem" for the average user.
define average user... today that's mom and pop that buy a PC and plug
it in. Definatly not the case in 1980. The reality of older computers
like the S100, SS50, Multibus types were you had to be a knowledgeable
user. Or if that was a problem you went to Apple or TRS80 styled machine
where the hardware was generally a fixed and software was the experimental
arena.
Allison
<C-machines (and I wouldn't class Unix as being one, although Unix would
<ran rather well _on_ one) have existed as well. They're machines where
<the instruction set is optimised for running C (so things like pointer
<indirection, particular loop structures, etc are machine
<instructions/operations).
the PDP-11 instruction set was the near ideal C machine due to it's
available addressing modes.
Allison
<I'd thought of that, but if you're not careful, your 'homebrew computer'
<contains ready-built I/O cards only. That's no so educational as building
<them yourself.
Depends on what your goals are. For me building yet another serial or disk
card is tedious and repetition. I'd rather experiment with advaced software
on old style but enhanced platforms.
<Sure. But the IIRC, this thread started with somebody wanting to make
<their first homebrew machine. IMHO, this should be a fairly simple
<machine - say a Z80 + maximum of 64K memory. Adding lots more memory,
<MMU, etc can come later. The first machine should be something that is
<likely to work first time.
No arguement. As a first pass machine static is usually the least painful
to deal with as are simple serial IO. The usual first time builder wants
more than their skills can generally support.
Allison
