22 Jan
2011
22 Jan
'11
7:15 p.m.
Hi all --
Obviously no one here "collects" the ENIAC, but it's certainly old
enough to be on-topic. :)
Having said that, here is a new site devoted to ENIAC --
www.the-eniac.com -- the six orange links up top are packed solid with
useful information.
Disclosure: it's NOT my site, but I was involved in its creation.
22 Jan
22 Jan
11:26 p.m.
On 2011 Jan 22, at 4:15 PM, Evan Koblentz wrote:
Hi all --
Obviously no one here "collects" the ENIAC, but it's certainly old
enough to be on-topic. :)
Having said that, here is a new site devoted to ENIAC --
www.the-eniac.com -- the six orange links up top are packed solid with
useful information.
Disclosure: it's NOT my site, but I was involved in its creation.
Great to see more about the ENIAC. I see somebody has done a graphical
simulation. Another project I don't have to do now (I did a graphical
sim of the ABC, always wanted to do one for the ENIAC, could still be
fun though).
But holy smokes:
http://www.cjr.org/page_views/number_cruncher.php
The rancor just goes on and on (.. not taking sides, just noting ..)
23 Jan
23 Jan
12:05 a.m.
Oh boy, let's argue the "first computer" question again....
This site promote the hoary old argument that the ABC wasn't really a computer, but
ENIAC was. From what I've read of both machines, the ABC can't be seen to be less
of a computer than ENIAC - although it was certainly less successful. The argument of
this website is that the ABC was nothing more than a calculator. I would suggest that
both ABC and ENIAC were in fact computers of limited application, to wit, the calculation
of complex and repetitive arithmetic, and both were originally designed for particular
purposes. ENIAC was later programmed for other arithmetic tasks, and perhaps that would
have been possible with the ABC as well - but history suggests that the ABC just
wasn't very well built (it didn't have federal dollars behind it!) and was not as
robust as one might hope. But robust and popular or flaky and obscure, the factual
history - not the mud-slinging against documentarian Smiley for what may well have been
sloppy journalism - seems to suggest that the ABC was the first of the genre,
"electronic digital computer," as limited as those initial machines were - all
of them.
The fact that Colossus was purpose-built and served to decipher codes should not take away
from the importance of its design. It was possibly as 'programmable' as the
ENIAC, but the scope of programming development was somewhat limited by fear of
annihilation at the hands of the Nazi regime. One must wonder what would have become of
it had its destruction not been ordered and its very existence declared a state secret.
I think ENIAC's overwhelming historical importance was the conversations it started.
It was not an obscure project by a college professor of physics nor a super-secret defense
project (although it did start out that way, it later became an 'open' secret),
and its significance was more about why it wasn't all it could be than what it was.
Von Neumann's observations would likely not have occurred without the context of the
ENIAC, nor would the interest in the potential of information technology have been quite
so stoked. I wonder how history would have been written if Atanasoff had been more
interested in fame and fortune than in quantum physics, or if he and Berry had been (or
could have hired) better engineers.
The Eckstein citation on this site does not constitute scholarship, just recursively poor
journalism.
It's a shame this site diminishes an otherwise worthwhile goal of celebrating the
ENIAC by regurgitating an old, emotional and unsupported argument. ENIAC was far more
important to history, but that does not mean we must disregard history that demonstrates
it was not first of its class. (First "general purpose" machine? Clearly
later, with EDSAC/EDVAC/UNIVAC, which were deliberately built by principle to be computers
and to subsequently fulfill specific roles at the behest of their programmers.) -- Ian
________________________________________
From: cctalk-bounces at classiccmp.org [cctalk-bounces at classiccmp.org] On Behalf Of
Evan Koblentz [evan at snarc.net]
Sent: Saturday, January 22, 2011 4:15 PM
To: On-Topic and Off-Topic Posts
Subject: ENIAC ....
Hi all --
Obviously no one here "collects" the ENIAC, but it's certainly old
enough to be on-topic. :)
Having said that, here is a new site devoted to ENIAC --
www.the-eniac.com -- the six orange links up top are packed solid with
useful information.
Disclosure: it's NOT my site, but I was involved in its creation.
1:23 a.m.
On 2011 Jan 22, at 9:05 PM, Ian King wrote:
Oh boy, let's argue the "first computer"
question again....
This site promote the hoary old argument that the ABC wasn't really a
computer, but ENIAC was. From what I've read of both machines, the
ABC can't be seen to be less of a computer than ENIAC - although it
was certainly less successful. The argument of this website is that
the ABC was nothing more than a calculator. I would suggest that both
ABC and ENIAC were in fact computers of limited application, to wit,
the calculation of complex and repetitive arithmetic, and both were
originally designed for particular purposes. ENIAC was later
programmed for other arithmetic tasks, and perhaps that would have
been possible with the ABC as well - but history suggests that the ABC
just wasn't very well built (it didn't have federal dollars behind
it!) and was not as robust as one might hope. But robust and popular
or flaky and obscure, the factual history - not the mud-slinging
against documentarian Smiley for what may well have been sloppy
journalism - seems to suggest that the ABC was the first of the genre,
"electronic digital computer," as limited as those initial machines
were - all of them.
Yes, there is too much ABC bashing in there. They won't even allow the
ABC credit for what it did do, or fail to understand what it did.
They state: "It?s true that the ABC used vacuum tubes for counting, but
that concept was already established in other inventions such as ray
counters." The ABC did not use counters, one of it's most significant
accomplishments was that it broke from counter-based arithmetic to do
arithmetic with logic and gates.
The fact that Colossus was purpose-built and served to
decipher codes
should not take away from the importance of its design. It was
possibly as 'programmable' as the ENIAC, but the scope of programming
development was somewhat limited by fear of annihilation at the hands
of the Nazi regime. One must wonder what would have become of it had
its destruction not been ordered and its very existence declared a
state secret.
I looked into the Colossus claims of programmability a little, but was
not convinced, at least from the technical description I recall seeing.
There was some flexibility in there, but very limited compared to the
ENIAC.
I think ENIAC's overwhelming historical importance
was the
conversations it started. It was not an obscure project by a college
professor of physics nor a super-secret defense project (although it
did start out that way, it later became an 'open' secret), and its
significance was more about why it wasn't all it could be than what it
was. Von Neumann's observations would likely not have occurred
without the context of the ENIAC, nor would the interest in the
potential of information technology have been quite so stoked.
Are you referring to the 'stored program' 'observations'? I believe it
has been reasonably established that the stored program concept could
not be attributed to Von Neumann - that the concept either came out of
the group discussions for EDVAC and could not be attributed to
individuals, or that Eckert and Mauchly had already arrived at the
notion before Von Neumann's involvement.
I wonder how history would have been written if
Atanasoff had been
more interested in fame and fortune than in quantum physics, or if he
and Berry had been (or could have hired) better engineers.
The Eckstein citation on this site does not constitute scholarship,
just recursively poor journalism.
It's a shame this site diminishes an otherwise worthwhile goal of
celebrating the ENIAC by regurgitating an old, emotional and
unsupported argument. ENIAC was far more important to history, but
that does not mean we must disregard history that demonstrates it was
not first of its class. (First "general purpose" machine? Clearly
later, with EDSAC/EDVAC/UNIVAC, which were deliberately built by
principle to be computers and to subsequently fulfill specific roles
at the behest of their programmers.) -- Ian
9:40 a.m.
They state: "It?s true that the ABC used vacuum
tubes for counting,
but that concept was already established in other inventions such as
ray counters." The ABC did not use counters, one of it's most
significant accomplishments was that it broke from counter-based
arithmetic to do arithmetic with logic and gates.
You're right. And you're the second person to point this out to me.
"Counting" is now changed to "logic".
Are you referring to the 'stored program'
'observations'? I believe it
has been reasonably established that the stored program concept could
not be attributed to Von Neumann - that the concept either came out of
the group discussions for EDVAC and could not be attributed to
individuals, or that Eckert and Mauchly had already arrived at the
notion before Von Neumann's involvement.
Correct. Unfortunately VN got credit for it. (Explanation in my
previous reply.)
4:34 p.m.
On 2011 Jan 23, at 6:40 AM, Evan Koblentz wrote:
I think you still miss the point. Your text now reads:
"It?s true that the ABC used vacuum tubes for logic, but that concept
was already established in other inventions such as ray counters."
Just changing the one word does not account for it.
'Ray counters', if we are talking the same thing, were flip-flops
chained into a typical digital counter.
Atanasoff's development was to take the notion of performing binary
arithmetic in terms of logic gates, not in terms of counting and not
implemented in flip-flops and counters, and implement that logic in
electronics. (To what degree he actually conceptualised it in _formal_
logic terms, I don't know.)
Yes, there were electronic 'coincidence' gates before Atanasoff and the
ray counters may have used them for gating pulses into the counter, but
there is a fundamental difference between what Atanasoff did and 'ray
counters'.
An interesting point here is that the ENIAC did not use that concept,
instead taking the more-conventional-for-the-time approach of
flip-flop-based counters, and it wasn't until the DR on the EDVAC that
the logic concept re-emerged, and the EDVAC looked more like the ABC in
that they were both bit-serial separated-memory-and-logic machines.
You also continue to write further down the page:
"because it used vacuum tubes for counting"
They state: "It?s true that the ABC used
vacuum tubes for counting,
but that concept was already established in other inventions such as
ray counters." The ABC did not use counters, one of it's most
significant accomplishments was that it broke from counter-based
arithmetic to do arithmetic with logic and gates.
You're right. And you're the second person to point this out to me.
"Counting" is now changed to "logic".
4:51 p.m.
Atanasoff's development was to take the notion of
performing binary
arithmetic in terms of logic gates
I have evidence that he wasn't the first to do this. I know this will
sound lame to you, but I haven't got that evidence handy and in a
digestable form. I will get it for you.
You also continue to write further down the page:
"because it used
vacuum tubes for counting"
Thanks for catching that. Will edit it.
I'll also add -- www.the-eniac.com is meant to be mostly historical in
nature, not extremely technical. Essentially the site owner (as Al
mentioned, it's Bill Mauchly) asked my help in writing some historical
content. His hope is that people find the site as a useful alternative,
in light of Smiley's book.
Other than Intel's John Gustafson, who's an avowed ABC guy, there is *
nobody * defending her book. (Not counting clueless book reviewers, who
focus on the nice story and how she tells it, vs. what actually happened.)
Sadly, according to Amazon sales ranks, Smiley's book is on its way to
becoming to best-selling computer history book of all time. It's
riddled with major mistakes, and it not only calls Mauchly a thief, but
even implies that he had something to do with Cliff Berry's death.
(Tongue firmly in cheek, Bill M. wrote to me, "My dad wasn't only a
crook, he was a murderer ... not bad!")
I know Al can't speak for CHM so I will not put him on the spot.
However I personally feel it's terrible that CHM is giving Smiley a
forum to tell her ill-researched story and be taken seriously.
There was even one senior CHM employee -- I will absolutely not name
that person or give any hints that could reveal their identity -- who
wrote to me and AGREED with my assessment of Smiley's book.
So I hope that CHM and its audience uses the opportunity to challenge
Smiley in person, not accept her drivel as good work. (Smiley herself
is massively backpedaling in recent interviews. She claims the book is
about ALL of the people whose work LED to computers. If that's true,
then why did she allow her publisher to call it "The Man Who Invented
the Computer" vs. "The * MEN (plural)"...... ???? As a Pulitzer winner
she certainly has the leverage to not let a publisher push her around.
6:44 p.m.
On 2011 Jan 23, at 1:51 PM, Evan Koblentz wrote:
Could be interesting, but keep in mind the point is implementing those
concepts in electronics (part of my statement is cut above), not just
the theory. The concepts were out there in the 30's (e.g. Shannon's
paper on relays and logic), and being implemented in relays: Stibitz's
K-model and the CNC. Roughly, what Stibitz did for calculation and
relays, Atanasoff did for calculation and electronics.
Atanasoff's development was to take the
notion of performing binary
arithmetic in terms of logic gates
I have evidence that he wasn't the first to do this. I know this will
sound lame to you, but I haven't got that evidence handy and in a
digestable form. I will get it for you.
9:37 a.m.
This site promote the hoary old argument that the ABC
wasn't really a computer
How can any machine be a "computer" if it doesn't have any ability to
run a program?
the calculation of complex and repetitive arithmetic
Actually, the ABC required * human intervention * after every calculation.
both were originally designed for particular purposes
That's incorrect. You hit upon a common fallacy regarding ENIAC -- that
it was "designed" for Army ballistics purposes. The truth is, Mauchly
was working on a general-purpose computer, and then the Army Ballistics
Lab * funded * his work. Essentially the Army said, "We'll pay for your
research if we get to use it first when it's done." Unfortunately many
people today think that means that ENIAC was commissioned * by * the
Army or at least specifically designed for them. It's not true!
ENIAC was later programmed for other arithmetic tasks,
and perhaps that would have been possible with the ABC as well
Perhaps. But it wasn't. Perhaps Columbus would have eventually sailed
in the right direction. He didn't.
history suggests that the ABC just wasn't very
well built (it didn't have federal dollars behind it!)
Also not true. Atanasoff spent years and lots of money working in the
Navy Ordnance Lab (at times with Mauchly's assistance) yet he still
failed to make a competitive computer.
the factual history ... seems to suggest that the ABC
was the first of the genre, "electronic digital computer,"
.... Except that its electronic logic was impeded by its
electromechanical everything else; and that without the ability to run a
program, it doesn't meet the definition of "computer".
The fact that Colossus was purpose-built and served to
decipher codes should not take away from the importance of its design.
Well we agree on that. :)
Von Neumann's observations would likely not have
occurred without the context of the ENIAC
You mean his plagiarizing of other people's work and sticking his own
name on it, and then widely distributing that classified document
without permission? Yes, i agree. :)
I wonder how history would have been written if
Atanasoff had been more interested in fame and fortune than in quantum physics, or if he
and Berry had been (or could have hired) better engineers.
Nobody is saying (for reals) that Atanasoff wasn't a brilliant guy.
It's not his fault that Jane Smiley and Mollenhoff before her wrote an
abortion of a "history book" and put Atanasoff's legacy through all of
this. Atanasoff made a fine electromechanical calculator, and he was
one of many people who helped others such as Mauchly meld their own ideas.
The Eckstein citation on this site does not constitute
scholarship, just recursively poor journalism.
Which citation do you mean? Remember, everything on the bottom half of
the site is merely someone's blog posts.
It's a shame this site diminishes an otherwise
worthwhile goal of celebrating the ENIAC by regurgitating an old, emotional and
unsupported argument.
Emotional? Unsupported? Back up those claims. Otherwise you'll sound
like Jane Smiley.
ENIAC was far more important to history, but that does
not mean we must disregard history that demonstrates it was not first of its class.
Not at all. The page "Was it the first computer?" clearly shows how
Babbage had many of the original ideas, Zuse was first to implement
them, and Colossus was first to make them all-electronic.
(First "general purpose" machine? Clearly
later, with EDSAC/EDVAC/UNIVAC, which were deliberately built by principle to be computers
and to subsequently fulfill specific roles at the behest of their programmers.)
All of those others were post-ENIAC, which (as I already explained) was
also "
deliberately built by principle to be computers and to subsequently fulfill specific roles
at the behest of their programmers". Also, some of these (mainly the stored-program
concept) are addressed in the "UNIVAC generation" link.
11:25 a.m.
On 23 Jan 2011 at 9:37, Evan Koblentz wrote:
The word "computer" is a strange topic. Note that IBM didn't call
the 704 a "computer" (data processing machine, calculator, but not a
computer).
So we're essentially trying to define things by a word that has its
meaning rooted in later devices. Good luck with that.
I knew some people who refused to call a MITS 8800 a "computer"...
--Chuck
This site promote the hoary old argument that the
ABC wasn't really
a computer
How can any machine be a "computer" if it doesn't have any ability to
run a program? 
11:35 a.m.
On 1/23/11 11:25 AM, Chuck Guzis wrote:
I knew some people who refused to call a MITS 8800 a
"computer"...
I've had one guy look me dead in the eye and tell me that my
PDP-11/70 couldn't possibly be a computer, and that I had no idea of
what I was talking about. Where's the mouse? Where's the "Start"
button? Circa 1974, huh? No, I'm clearly nuts, computers only came
around in the 1990s, "when modems were invented so we could surf the
Internet".
I very nearly experienced spontaneous combustion.
I'm reminded of a T-shirt that I had in the mid-1990s. It said "NO,
the Internet is NOT cool! Go away!"
-Dave
--
Dave McGuire
Port Charlotte, FL
11:47 a.m.
On 23 Jan 2011 at 11:35, Dave McGuire wrote:
On 1/23/11 11:25 AM, Chuck Guzis wrote:
This was in 1975 and it was a (female) co-worker who wanted to see my
computer. Her response was withering: "That's not a computer!
That's a toy!".
I don't think I've ever recovered from that.
--Chuck
I knew some people who refused to call a MITS
8800 a "computer"...
I've had one guy look me dead in the eye and tell me that my
PDP-11/70 couldn't possibly be a computer, and that I had no idea of
what I was talking about. Where's the mouse? Where's the "Start"
button? Circa 1974, huh? No, I'm clearly nuts, computers only came
around in the 1990s, "when modems were invented so we could surf the
Internet". 
6:11 p.m.
On Sun, 23 Jan 2011, Dave McGuire wrote:
button? Circa 1974, huh? No, I'm clearly nuts,
computers only came
around in the 1990s, "when modems were invented so we could surf the
Internet".
The definition is quite "fluid" - it is "defined" primarily by people
who
have no expertise. The majority of people will define "computer", or at
least "first computer" based ENTIRELY on what ahd when first got their
attention. Thus, there are some media "writers" who think only in terms
of Apple][, IBM PC, Mac.
Too many people equate the history of computers with WWW.
If there had not been porn on the web, would the public have ever even
noticed computers?
--
Grumpy Ol' Fred cisin at xenosoft.com
12:46 p.m.
On Sunday, January 23, 2011 6:38 AM Evan Koblentz Wrote:
You mean his plagiarizing of other people's work
and sticking his own name
on it, and then widely distributing that classified
document without
permission? Yes, i agree. :)
I agree with most of the rest of your text, but not with this.
Von Neumann did not plagiarize anything. He documented what he learned and
passed it up the chain. The distribution was done by "military
intelligence" and the names of those involved in the top-secret project
redacted by the same "intelligence."
Von Neumann didn't mean to steal credit from the other Johns.
Erik Klein
www.vintage-computer.com
www.vintage-computer.com/vcforum - The Vintage Computer Forums
marketplace.vintage-computer.com - The Vintage Computer and Gaming
Marketplace
-----Original Message-----
From: cctalk-bounces at classiccmp.org [mailto:cctalk-bounces at classiccmp.org]
On Behalf Of Evan Koblentz
Sent: Sunday, January 23, 2011 6:38 AM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: ENIAC ....
This site promote the hoary old argument that the ABC
wasn't really a
computer
How can any machine be a "computer" if it doesn't have any ability to
run a program?
the calculation of complex and repetitive arithmetic
Actually, the ABC required * human intervention * after every calculation.
both were originally designed for particular purposes
That's incorrect. You hit upon a common fallacy regarding ENIAC -- that
it was "designed" for Army ballistics purposes. The truth is, Mauchly
was working on a general-purpose computer, and then the Army Ballistics
Lab * funded * his work. Essentially the Army said, "We'll pay for your
research if we get to use it first when it's done." Unfortunately many
people today think that means that ENIAC was commissioned * by * the
Army or at least specifically designed for them. It's not true!
ENIAC was later programmed for other arithmetic tasks,
and perhaps that
would have been possible with the ABC as well
Perhaps. But it wasn't. Perhaps Columbus would have eventually sailed
in the right direction. He didn't.
history suggests that the ABC just wasn't very
well built (it didn't have
federal dollars behind it!)
Also not true. Atanasoff spent years and lots of money working in the
Navy Ordnance Lab (at times with Mauchly's assistance) yet he still
failed to make a competitive computer.
the factual history ... seems to suggest that the ABC
was the first of the
genre, "electronic digital computer,"
.... Except that its electronic logic was impeded by its
electromechanical everything else; and that without the ability to run a
program, it doesn't meet the definition of "computer".
The fact that Colossus was purpose-built and served to
decipher codes
should not take away from the importance of its design.
Well we agree on that. :)
Von Neumann's observations would likely not have
occurred without the
context of the ENIAC
You mean his plagiarizing of other people's work and sticking his own
name on it, and then widely distributing that classified document
without permission? Yes, i agree. :)
I wonder how history would have been written if
Atanasoff had been more
interested in fame and fortune than in quantum physics, or
if he and Berry
had been (or could have hired) better engineers.
Nobody is saying (for reals) that Atanasoff wasn't a brilliant guy.
It's not his fault that Jane Smiley and Mollenhoff before her wrote an
abortion of a "history book" and put Atanasoff's legacy through all of
this. Atanasoff made a fine electromechanical calculator, and he was
one of many people who helped others such as Mauchly meld their own ideas.
The Eckstein citation on this site does not constitute
scholarship, just
recursively poor journalism.
Which citation do you mean? Remember, everything on the bottom half of
the site is merely someone's blog posts.
It's a shame this site diminishes an otherwise
worthwhile goal of
celebrating the ENIAC by regurgitating an old, emotional and
unsupported
argument.
Emotional? Unsupported? Back up those claims. Otherwise you'll sound
like Jane Smiley.
ENIAC was far more important to history, but that does
not mean we must
disregard history that demonstrates it was not first of its class.
Not at all. The page "Was it the first computer?" clearly shows how
Babbage had many of the original ideas, Zuse was first to implement
them, and Colossus was first to make them all-electronic.
(First "general purpose" machine? Clearly
later, with EDSAC/EDVAC/UNIVAC,
which were deliberately built by principle to be
computers and to
subsequently fulfill specific roles at the behest of their programmers.)
All of those others were post-ENIAC, which (as I already explained) was
also "
deliberately built by principle to be computers and to subsequently fulfill
specific roles at the behest of their programmers". Also, some of these
(mainly the stored-program concept) are addressed in the "UNIVAC generation"
link.
1:17 p.m.
Von Neumann did not plagiarize anything. ?He
documented what he learned and
passed it up the chain. ?The distribution was done by "military
intelligence" and the names of those involved in the top-secret project
redacted by the same "intelligence."
I am curious - what was the *official* classification of the ENIAC
project? Is there a primary source available?
--
Will
3:28 p.m.
You mean his
plagiarizing of other people's work and sticking his own name on it, and then widely
distributing that classified document without permission?
I agree with most of the
rest of your text, but not with this. Von Neumann did not plagiarize anything. He
documented what he learned and passed it up the chain. ... Von Neumann didn't mean to
steal credit from the other Johns.
4:02 p.m.
On 2011 Jan 23, at 6:37 AM, Evan Koblentz wrote:
Because it depends on the definition one is working with. As usual, we
are in the realm of varying definitions. If we go with the modern
understanding of the word, neither the ABC or ENIAC was a computer.
Perhaps people should stick to facts: "this did that"; rather than
declarations: "this was that".
All these machines had their contributions. ENIAC doesn't get to claim
everything, and when you boil off all the hyperbole on both sides, I
think that was the primary point or outcome of the court case.
I drew up this years ago when I was reading more about these issues:
http://www3.telus.net/~bhilpert/tmp/conceptsmachines.gif
although I don't think I was the first to arrive at such a diagram. Of
course there are various improvements that could be made. And I suppose
there may be arguments about the arrow from the ABC to ENIAC.
This is not correct.
On my site at: http://www.cs.ubc.ca/~hilpert/e/ABC/omflow.html
is a flowchart of the problem solving process using the ABC.
Note it carefully distinguishes between what the human operator did and
what the ABC did. Note the portions in blue (particularly in stages
2,3,4) include a loop structure with decisions and calculations (on
multiple 50-bit binary numbers), all performed automatically.
There were 2 or 3 mostly-hard-wired programs in the ABC, those programs
had some plug-board flexibility and were data-sensitive.
I'm not entirely disagreeing with you but this strikes me as another
grey area. Yes, Mauchly was researching automatic computation prior to
ENIAC, for other than ballistics purposes. However, I've never seen
anything to suggest he had arrived at the design that would become the
ENIAC before his time at the Ballistics lab. I'd guess it highly likely
Mauchly had his greater purposes in mind as he and Eckert designed
ENIAC, but I believe the design of ENIAC took place in the working
environment of the Ballistics Lab and was very much targeted at the
ballistics calculations it was funded for. The Army funded ENIAC for a
specific purpose, not as a general or head-in-the-clouds research
effort.
On the other hand, I also like to argue the developments would have
taken place regardless of the military and the only reason the ENIAC
was a 'military' machine was (as with the mis-attribution to Willi
Sutton) "That's where the money is".
It was a case of desire and interest (a researcher wanting greater
computing capabilities), meeting dire and timely need (war/ballistics)
which motivated funding (from society through the Army).
Yes, the ABC (like Colossus) was limited in it's flexibility relative
to ENIAC.
Yes, the ABC was limited by the mechanical drum rotation. However, in
his paper about the ASM (Add-Subtract Mechanism[1]), Atanasoff did
recognise and promote that the arithmetic or logical computation
through that mechanism took place at electronic speeds.
To mention, all the drum computers (IBM 650, Bendix G-15, LGP-30, etc)
of the 50s were electromechanical by this definition.
[1] Use of the word 'Mechanism' in the name is a bit of an anachronism
- the ASM was a fully electronic mechanism.
This site promote the hoary old argument that the
ABC wasn't really a
computer
How can any machine be a "computer" if it doesn't have any ability to
run a program? 
the
calculation of complex and repetitive arithmetic
Actually, the ABC required * human intervention * after every
calculation. both were
originally designed for particular purposes
That's incorrect. You hit upon a common fallacy regarding ENIAC --
that it was "designed" for Army ballistics purposes. The truth is,
Mauchly was working on a general-purpose computer, and then the Army
Ballistics Lab * funded * his work. Essentially the Army said, "We'll
pay for your research if we get to use it first when it's done."
Unfortunately many people today think that means that ENIAC was
commissioned * by * the Army or at least specifically designed for
them. It's not true! ENIAC was
later programmed for other arithmetic tasks, and perhaps
that would have been possible with the ABC as well
Perhaps. But it wasn't. Perhaps Columbus would have eventually sailed
in the right direction. He didn't.
> history suggests that the ABC just wasn't very
well built (it didn't
> have federal dollars behind it!)
The ABC was a small project compared to the others being spoken of.
There was one aspect of it (base-2 card printing and punching) that had
a problematic error rate. The processing and memory aspects apparently
did well.
Also not true. Atanasoff spent years and lots of
money working in the
Navy Ordnance Lab (at times with Mauchly's assistance) yet he still
failed to make a competitive computer.
That was a different effort and project than the ABC, let's not confuse
them.
the factual
history ... seems to suggest that the ABC was the first
of the genre, "electronic digital computer,"
.... Except that its electronic logic was impeded by its
electromechanical everything else; and that without the ability to run
a program, it doesn't meet the definition of "computer".
4:25 p.m.
Perhaps people should stick to facts: "this did
that"; rather than
declarations: "this was that". All these machines had their
contributions. ENIAC doesn't get to claim everything
I agree. That's why the page at http://the-eniac.com/first/ is
organized exactly as you described. The only thing claimed is that
ENIAC was the first "general-purpose, all-electronic" computer.
Babbage designed a computer. Unfortunately he wasn't able to build it.
Zuse designed a computer, but it was electromechanical. Colossus was an
all-electronic computer, but it was single-purpose. What ENIAC did was
make the advancement to being general-purpose AND all-electronic.
On my site at:
http://www.cs.ubc.ca/~hilpert/e/ABC/omflow.html is a
flowchart of the problem solving process using the ABC. Note it
carefully distinguishes between what the human operator did and what
the ABC did.
A lot to read there! I will have to get back to you.
On the other hand, I also like to argue the
developments would have
taken place regardless of the military and the only reason the ENIAC
was a 'military' machine was (as with the mis-attribution to Willi
Sutton) "That's where the money is".
Absolutely.
5:51 p.m.
On 2011 Jan 23, at 1:25 PM, Evan Koblentz wrote:
Perhaps people should stick to facts: "this
did that"; rather than
declarations: "this was that". All these machines had their
contributions. ENIAC doesn't get to claim everything
I agree. That's why the page at http://the-eniac.com/first/ is
organized exactly as you described. The only thing claimed is that
ENIAC was the first "general-purpose, all-electronic" computer.
Babbage designed a computer. Unfortunately he wasn't able to build
it. Zuse designed a computer, but it was electromechanical. Colossus was an all-electronic computer, but it was
single-purpose.
TMU, Colossus was actually like the ABC in that it did some processing
in electronics but the primary data flow still had electro-mechanics in
the path (giant paper tape loops).
The electronics was still beneficial in both because continually
rotating mechanisms (ABC:drum, Colossus:paper tape loops) could operate
at higher speeds than toggling relays (changing momentum).
What ENIAC did was make the advancement to being
general-purpose AND
all-electronic.
Yes.
But the ABC still needs credit in that list for what it did do.
On my site at:
http://www.cs.ubc.ca/~hilpert/e/ABC/omflow.html is a
flowchart of the problem solving process using the ABC. Note it
carefully distinguishes between what the human operator did and what
the ABC did.
A lot to read there! I will have to get back to you.
On the other hand, I also like to argue the
developments would have
taken place regardless of the military and the only reason the ENIAC
was a 'military' machine was (as with the mis-attribution to Willi
Sutton) "That's where the money is".
Absolutely.
7:08 p.m.
TMU, Colossus was actually like the ABC in that it did
some processing
in electronics but the primary data flow still had electro-mechanics
in the path (giant paper tape loops).
So you're saying Colossus was * not * fully electronic? I didn't know
that. Very interesting.
What ENIAC did was make the advancement to being
general-purpose AND
all-electronic.
Yes.
But the ABC still needs credit in that list for what it did do.
24 Jan
24 Jan
12:04 a.m.
On 2011 Jan 23, at 4:08 PM, Evan Koblentz wrote:
Yes, it was not fully electronic. The ciphertext was on a long paper
tape loop cycling at high speed through pulleys and an optical reader.
The wheels of the Lorenz cipher machine were simulated with electronics
and data from the tape and wheels were compared through a logical
function again in electronics. The paper tape data was not "loaded
into" the machine, it was a primary data source throughout the
processing. Although - it was running at 5000 characters per second, so
I guess that's getting into a grey area as to whether the electronics
(of the time) were being limited by the mechanical.
You might have a point there. I'll think about adding in what the ABC
* did * do. Maybe I'll write something about how it innovated in
using electronics for calculating.
I think that would be good. If I can help clarify anything, feel free
to ask. Check out the beginning page of my ABC article if you haven't,
as well as the ASM page.
TMU, Colossus
was actually like the ABC in that it did some
processing in electronics but the primary data flow still had
electro-mechanics in the path (giant paper tape loops).
So you're saying Colossus was * not * fully electronic? I didn't know
that. Very interesting. What ENIAC did was make the advancement to being
general-purpose
AND all-electronic.
Yes.
But the ABC still needs credit in that list for what it did do. But I'm also sticking with "no program = not
a computer."
As you wish for the definition of computer you choose. Not
programmable: I'd tend to agree. No program: that's not so clear, I'd
tend to disagree.
23 Jan
23 Jan
8:50 p.m.
I drew up this years ago when I was reading more about
these issues:
http://www3.telus.net/~bhilpert/tmp/conceptsmachines.gif
although I don't think I was the first to arrive at such a diagram. Of
course there are various improvements that could be made. And I
suppose there may be arguments about the arrow from the ABC to ENIAC.
I admit that your ABC page is over my head, technical-wise. So, I asked
Bill M. for his opinion. Here is what he said (with permission for me
to quote it here):
----------------------------------
The [Atanasoff] machine had several operations, which were initiated by
pressing the appropriate button. Two of them were decimal input and
output, and just one that did arithmetic. Calling the arithmetic
operation "programmable" and assigning it an adjective like
"data-sensitive" is a stretch; if you use those terms so loosely, you
will have to say a mechanical adding machine also has those properties.
There were 30 separate adders and 30 pairs of coefs. The user had to
select which pair was to be eliminated. This selection happened by the
human moving a jumper cable from one set of holes to another. Selecting
one of the 30 for final output happened in a similar way, I believe. The
I/O for cards happened in groups of 5 and had real switches.
But except for selecting an input, there are no "programmable options"
on the arithmetic. There was a switch you could set, so that it would
either start with a subtract or start with an add. (Once it started it
would flip back and forth.) The user has to know which way to set this
by examining the signs of the coefs. I wouldn't call this a
programmable option as much as a hazard. If you set it the wrong way,
the machine runs forever and never finishes. (Well, not forever, it
will definitely finish in 2 to 50th seconds. 35M years.)
As to the other part, data sensitivity, this is somewhat true. The
machine is doing a "long" division by repeated adding and subtracting.
In the end, the result is just a simple math operation, but the length
of time it takes is dependent on the actual numbers you plug in. It is
literally just testing if the sign bit flipped, and throwing a relay
each time it does. I guess you could say every multiplier is data dependent.
11:32 p.m.
On 2011 Jan 23, at 5:50 PM, Evan Koblentz wrote:
I drew up this
years ago when I was reading more about these issues:
http://www3.telus.net/~bhilpert/tmp/conceptsmachines.gif
although I don't think I was the first to arrive at such a diagram.
Of course there are various improvements that could be made. And I
suppose there may be arguments about the arrow from the ABC to ENIAC.
I admit that your ABC page is over my head, technical-wise. So, I
asked Bill M. for his opinion. Here is what he said (with permission
for me to quote it here):
---------------------------------- The [Atanasoff] machine had several operations, which
were initiated
by pressing the appropriate button. Two of them were decimal input
and output, and just one that did arithmetic.
This is not correct. The decimal input and output operations did binary
<-> decimal conversion. These involved a loop with repeated
full-register additions and subtractions (arithmetic) and sensing of
the data state (data-sensitive) along the way. This is shown in the
flowchart. The other operation was pair-elimination which was a similar
loop.
Calling the arithmetic operation
"programmable" and assigning it an
adjective like "data-sensitive" is a stretch; if you use those terms
so loosely, you will have to say a mechanical adding machine also has
those properties.
To be accurate I did not call or claim it to be programmable and I too
would be reluctant to do so. I said there was some flexibility in the
internal programs and I agree it was limited. The jumpers mentioned
below actually select which register the sign and zero detection would
be performed on in the loop.
Dismissing the points by equating them with mechanical adding machines
is flippant. No, mechanical adding machines do not have the properties
mentioned, however mechanical machines with built-in and multiplication
and division were complex machines. More below about M & D. Sometimes
one has to look at underlying concepts rather than common surface-level
perceptions.
There were 30 separate adders and 30 pairs of coefs.
The user had to
select which pair was to be eliminated. This selection happened by
the human moving a jumper cable from one set of holes to another.
Selecting one of the 30 for final output happened in a similar way, I
believe. The I/O for cards happened in groups of 5 and had real
switches.
But except for selecting an input, there are no "programmable options"
on the arithmetic. There was a switch you could set, so that it
would either start with a subtract or start with an add. (Once it
started it would flip back and forth.) The user has to know which way
to set this by examining the signs of the coefs. I wouldn't call this
a programmable option as much as a hazard. If you set it the wrong
way, the machine runs forever and never finishes. (Well, not forever,
it will definitely finish in 2 to 50th seconds. 35M years.)
--
As to the other part, data sensitivity, this is
somewhat true. The
machine is doing a "long" division by repeated adding and subtracting.
In the end, the result is just a simple math operation, but the
length of time it takes is dependent on the actual numbers you plug
in. It is literally just testing if the sign bit flipped, and
throwing a relay each time it does. I guess you could say every
multiplier is data dependent.
This is largely correct, but dismissing the ABC on the above is again
flippant. Multiplication and division have been implemented as small
programs throughout the history of computers, be it in a hardware state
machine, microcode, or (frequently) as instruction-level programs as
many machines do not provide those ops as instructions. Various 'real
computers' have gotten by with little or nothing more than sign- and/or
zero-detection for data-sensitivity/conditional operation, just as the
ABC provides.
One of the first three programs to run on the Manchester Baby (first
stored program machine) was Turing's long division routine. The other
two were factoring routines and no more complex. Routines like this
were enough to exercise the machine and show it could do practical
work.
The relative scale of the projects should also be kept in mind, ENIAC
was a large well-funded project, ABC was a small two-man project.
Despite it's small size, the concepts embodied and implemented in the
ABC were significant. ENIAC was much larger but which machine actually
embodied more original concepts is not obvious to answer.
I don't like to overstate what the ABC did, which is why I prepared
that web article about it, to try to put down concisely what it did
technically. I think some have overstated what it did it on occasion,
but it is also about time the ENIAC supporters acknowledged what the
ABC did do, rather than just trying to dismiss it. I understand there
are the familial/emotional issues of having Mauchly's name dragged
through the mud and I can sympathise on that. And I understand there
may still be differing opinions on how much influence Mauchly obtained
from Atanasoff. But the technical assessments need to
be separated from
those matters.
If one were to take the same attitude I perceive from ENIAC supporters
in their dismissal of the ABC, one could just as easily dismiss the
ENIAC: it wasn't a stored program machine, took hours to change a
program, wasn't a universal machine, could only operate on decimal
numbers, could not do general symbolic logic at least with any
facility, had next to no memory, couldn't do the things the machines
that came after it could, was an architectural dead-end, was just an
overblown calculator, it simply wasn't a 'computer'. The ABC had an
architecture closer to the stored-program proto-design EDVAC than did
ENIAC. Computing history began with the stored program machines, Turing
laid the theoretical basis for stored program machines, the Manchester
Baby was the first one to run, the EDSAC was the first one to provide
real service. And those were all British developments. So there. (..and
I'm not British.)
24 Jan
24 Jan
12:07 a.m.
On 2011 Jan 23, at 8:44 PM, William Donzelli wrote:
I do hope it was clear I was being utterly rhetorical (one never knows
how things will be interpreted).
Computing
history began with the stored
program machines,
The one thing that strikes me after reading this thread is that it
seems there is no precise point when computing history began.
2:13 a.m.
Gentlemen
I think what is missing here is he concept of evolution. There is no
'first', only steps on the path from primitive counting to to-days systems.
First implies an absolute like climbing Mount Everest. With computing there
is no reachable end goal. The mountain is of infinite height.
Identifying significant contributors is very valid however.
Regards
?
Rod Smallwood
?
-----Original Message-----
From: cctech-bounces at classiccmp.org [mailto:cctech-bounces at classiccmp.org]
On Behalf Of Eric Smith
Sent: 24 January 2011 06:03
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: ENIAC ....
William Donzelli wrote:
The one thing that strikes me after reading this
thread is that it
seems there is no precise point when computing history began.
Huh? I thought it began on January 13, 11 CE :-)
8:57 a.m.
On Monday, January 24, 2011, Rod Smallwood wrote:
Gentlemen
I think what is missing here is he concept of evolution. There
is no
'first', only steps on the path from primitive
counting to to-days
systems.
I don't believe in machine "evolution", only in machine "intelligent
design". :)
Pat
--
Purdue University Research Computing --- http://www.rcac.purdue.edu/
The Computer Refuge --- http://computer-refuge.org
25 Jan
25 Jan
2:25 p.m.
From: Eric Smith
Sent: Sunday, January 23, 2011 10:03 PM
William Donzelli wrote:
> The one thing that strikes me after reading this thread is that it
> seems there is no precise point when computing history began.
Huh? I thought it began on January 13, 11 CE :-)
OK, I give up. What happened on the Ides of January in the year 764 AUC?[1]
[1] QVOD OCCIDIT IDVM IANVARII ANNO DCCLXIV AVC?
Rich Alderson
Vintage Computing Sr. Server Engineer
Vulcan, Inc.
505 5th Avenue S, Suite 900
Seattle, WA 98104
mailto:RichA at vulcan.com
mailto:RichA at LivingComputerMuseum.org
http://www.PDPplanet.org/
http://www.LivingComputerMuseum.org/ <=== New site now live!!!!
2:32 p.m.
On Tue, 25 Jan 2011, Rich Alderson wrote:
From: Eric Smith
Sent: Sunday, January 23, 2011 10:03 PM
I think that was a reference to CHM's "Revolution: The First 2000 Years of
Computing" which opened on Jan 13 this year :-)
Alexey
William Donzelli wrote:
> The one thing that strikes me after reading this thread is that it
> seems there is no precise point when computing history began.
Huh? I thought it began on January 13, 11 CE :-)
OK, I give up. What happened on the Ides of January in the year 764 AUC?[1]
[1] QVOD OCCIDIT IDVM IANVARII ANNO DCCLXIV AVC?
3:14 p.m.
New subject: Computing epoch (Was: ENIAC ....
The one thing
that strikes me after reading this thread is that it
seems there is no precise point when computing history began.
Huh? I thought it
began on January 13, 11 CE :-) > seems there is no precise point when computing
history began.
So, like Morrow's standards ("Everyone can have a unique one
of their
own"), there are MANY precise points when computing history began.
--
Grumpy Ol' Fred cisin at xenosoft.com
3:54 p.m.
New subject: Computing epoch (Was: ENIAC ....
On Tue, Jan 25, 2011 at 3:14 PM, Fred Cisin <cisin at xenosoft.com> wrote:
Any Unix type can tell you that history began at 00:00:00 January 1,
1970. ? But, they do allow negative timestamps...
Mac history began January 1, 1904... to avoid the leap year exception.
VMS counts 100ns quanta (each call to get the system time is
guaranteed to be unique) since November 17, 1858, coinciding with the
beginning of the Modified Julian Day epoch (handy for working with
some astronomical data sets). It's internally represented as a 64-bit
signed number - absolute times are positive, delta times (used with
batch jobs, as in "run four hours from now") are represented as
negative numbers, so timestamps can't be negative, but time values
_can_ (operator overloading, of a sort).
VMS time fails on 31-JUL-31086, but there are comments in the source
that the date formatting routines will fail on 1-Jan-10000 (the code
assumes that years are always represented in four digits).
So VMS is pretty robust at handling time, but has a known Y10K problem
(and less than 7989 years left to fix it!)
-ethan
The one thing that strikes me after reading this
thread is that it
seems there is no precise point when computing history began.
Huh? I thought it
began on January 13, 11 CE ?:-)
4:23 p.m.
New subject: Computing epoch (Was: ENIAC ....
VMS time fails on 31-JUL-31086, but there are comments
in the source
that the date formatting routines will fail on 1-Jan-10000 (the code
assumes that years are always represented in four digits).
VMS - the cockroach of the operating system world.
--
Will
26 Jan
26 Jan
2:04 p.m.
New subject: Computing epoch (Was: ENIAC ....
VMS time fails on 31-JUL-31086, but there are comments
in the
source that the date formatting routines will fail on
1-Jan-10000 (the code assumes that years are always represented in
four digits).
There was a semi-official response somewhere that said (essentially)
"we know and we'll fix it before then". Sadly I can't find my copy
right now ...
Antonio
25 Jan
25 Jan
4:11 p.m.
> Huh? I
thought it began on January 13, 11 CE :-)
I think that was a reference to
CHM's "Revolution: The First 2000 Years of
Computing" which opened on Jan 13 this year :-)
> OK, I give up. What happened on the Ides of January in the year 764 AUC?[1]
7:16 p.m.
-----Original Message-----
From: cctalk-bounces at classiccmp.org [mailto:cctalk-
bounces at classiccmp.org] On Behalf Of Fred Cisin
Sent: Tuesday, January 25, 2011 1:11 PM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: RE: ENIAC ....
Years of
You mean Ken Olsen had precedent?!?
> Huh?
I thought it began on January 13, 11 CE :-)
I think that was a reference to
CHM's "Revolution: The First 2000 Computing" which opened on Jan 13 this year
:-)
> OK, I give up. What happened on the Ides of January in the year
764 AUC?[1]
Was that the date when they realized that persobnal computers would be
a very bad idea, and threw the Antikytheran device into the sea to get
rid of it? :-)
26 Jan
26 Jan
1:23 a.m.
> OK, I give up. What happened on the Ides of
January in the year
764 AUC?[1]
Was that the date when they realized that persobnal computers would be
a very bad idea, and threw the Antikytheran device into the sea to get
rid of it? :-) 
25 Jan
25 Jan
2:56 p.m.
William Donzelli wrote:
The one thing that strikes me after reading this
thread is that it
seems there is no precise point when computing history began.
I wrote:
Huh? I thought it began on January 13, 11 CE :-)
Rich Alderson wrote:
OK, I give up. What happened on the Ides of January in
the year 764
AUC?[1]
[1] QVOD OCCIDIT IDVM IANVARII ANNO DCCLXIV AVC?
It was 2000 years prior to the opening of the Computer History Museum's
"Revolution: The First 2000 Years of Computing" exhibit. Though I
suppose I am probably off by ten days due to the change from the Julian
calendar to the Gregorian calendar. Perhaps the correct date is January
23, 11 CE.
Eric
24 Jan
24 Jan
12:11 a.m.
Dismissing the points by equating them with mechanical
adding machines
is flippant. No, mechanical adding machines do not have the properties
mentioned, however mechanical machines with built-in and
multiplication and division were complex machines.
There are many complex calculating machines that aren't computers.
dismissing the ABC on the above is again flippant.
Multiplication and
division have been implemented as small programs throughout the
history of computers, be it in a hardware state machine, microcode, or
(frequently) as instruction-level programs as many machines do not
provide those ops as instructions. Various 'real computers' have
gotten by with little or nothing more than sign- and/or zero-detection
for data-sensitivity/conditional operation, just as the ABC provides.
That m/d is one point among many; nobody's saying it was a useless
technique.
the Manchester Baby (first stored program machine)
Not necessarily; see my email citing the 1947 EDVAC memo.
I don't like to overstate what the ABC did, which
is why I prepared
that web article about it, to try to put down concisely what it did
technically. I think some have overstated what it did it on occasion,
but it is also about time the ENIAC supporters acknowledged what the
ABC did do, rather than just trying to dismiss it. I understand there
are the familial/emotional issues of having Mauchly's name dragged
through the mud and I can sympathise on that. And I understand there
may still be differing opinions on how much influence Mauchly obtained
from Atanasoff. But the technical assessments need to be separated
from those matters.
You're right. I am rewriting parts of the new web site to reflect this.
ENIAC .... took hours to change a program
That was certainly one of the biggest knocks against it. A fast
computer has limited utility if setting up the program was tedious.
wasn't a universal machine
Please elaborate.
was an architectural dead-end
That's not true, but I think you and I each made our points already
about ABC and ENIAC.
was just an overblown calculator, it simply wasn't
a 'computer'
:-)
stored program machines, the Manchester Baby was the
first one to run
See above, re: Baby.
EDSAC was the first one to provide real service.
Oy! ENIAC was providing "real service", i.e. running important
applications for academia, government, and the Army, by 1945 (and with a
stored program by 1948). EDSAC wasn't even a pipe dream in 1945, and
didn't go into operation until 1949.
Truce?
1:32 a.m.
On 2011 Jan 23, at 9:11 PM, Evan Koblentz wrote:
Yes, I did see that and it sounds interesting, but the consensus for
the time being is still the Baby.
And one of the primary issues that led to the development of the stored
program concept. See below.
In the Universal Turing Machine sense, from computing theory. A UTM can
compute a certain class of functions. Any machine with a certain set of
capabilities can be shown to be a UTM equivalent and thereby able to
compute the same set of functions. With the assumption of being given
enough memory and enough time. That is, the assessment is made on
qualitative characterisation not quantitative.
All our general purpose machines today are UTM equivalents. Harvard
arch microcontrollers arguably excepted.
I think it would be difficult to construe the ENIAC into a UTM
equivalent.
Although this quote from the Wikipedia article is intriguing:
"Minsky goes on to demonstrate Turing equivalence of a counter
machine"
There was an interesting article a few years ago that showed the Zuse
Z3 (1941, harvard-arch relay machine) could be turned into a UTMeq by
providing it with a program implementing a stored-program pseudo
machine.
The point is that the EDVAC and the
stored-program-machine/UTM-equivalents that came after bore no
architectural resemblance to the ENIAC.
(Architecture here meaning the processor structure, not the
implementation (tubes, relays, etc.))
ENIAC's accomplishment was to achieve electronic speed computation. At
the same time it achieved that, it exposed it's own shortcoming: the
programming bottleneck. Enormous amounts of computing time were lost
while the program was changed. All those cables and plugs and switches
needed to be changed into electronic switches controlled by an equally
fast memory. Along with issues related to memory assignment, a radical
change in architecture was necessary and the stored program concept was
created. And there we step into the fog of what happened in those
meetings and how much Von Neumann with some knowledge of Turing's
Turing machine paper from the 30s influenced that creation.
Some interesting comments under stored program concept on:
http://en.wikipedia.org/wiki/Universal_Turing_machine
Keeping in mind the rhetoric.. if the ENIAC wasn't a computer then the
EDSAC gets that credit.
(I'd disagree about the 1948 stored program ENIAC, as I mentioned in a
previous message.)
Dismissing the
points by equating them with mechanical adding
machines is flippant. No, mechanical adding machines do not have the
properties mentioned, however mechanical machines with built-in and
multiplication and division were complex machines.
There are many complex calculating machines that aren't computers.
dismissing the ABC on the above is again
flippant. Multiplication and
division have been implemented as small programs throughout the
history of computers, be it in a hardware state machine, microcode,
or (frequently) as instruction-level programs as many machines do not
provide those ops as instructions. Various 'real computers' have
gotten by with little or nothing more than sign- and/or
zero-detection for data-sensitivity/conditional operation, just as
the ABC provides.
That m/d is one point among many; nobody's saying it was a useless
technique.
the Manchester Baby (first stored program
machine)
Not necessarily; see my email citing the 1947 EDVAC memo. I don't
like to overstate what the ABC did, which is why I prepared
that web article about it, to try to put down concisely what it did
technically. I think some have overstated what it did it on occasion,
but it is also about time the ENIAC supporters acknowledged what the
ABC did do, rather than just trying to dismiss it. I understand there
are the familial/emotional issues of having Mauchly's name dragged
through the mud and I can sympathise on that. And I understand there
may still be differing opinions on how much influence Mauchly
obtained from Atanasoff. But the technical assessments need to be
separated from those matters.
You're right. I am rewriting parts of the new web site to reflect this.
ENIAC .... took hours to change a program
That was certainly one of the biggest knocks against it. A fast
computer has limited utility if setting up the program was tedious. wasn't a
universal machine
Please elaborate. was an
architectural dead-end
That's not true, but I think you and I each made our points already
about ABC and ENIAC. was just an
overblown calculator, it simply wasn't a 'computer'
:-)
stored program machines, the Manchester Baby was
the first one to run
See above, re: Baby.
EDSAC was the first one to provide real service.
Oy! ENIAC was providing "real service", i.e. running important
applications for academia, government, and the Army, by 1945 (and with
a stored program by 1948). EDSAC wasn't even a pipe dream in 1945,
and didn't go into operation until 1949. Truce?
What? This list finally has a good thread about computing history and
you want to stop!? (Yes, I do have to get other things done, too.)
1:38 a.m.
Truce?
What? This list finally has a good thread about computing history and
you want to stop!? (Yes, I do have to get other things done, too.) ENIAC's accomplishment was to achieve electronic
speed computation. At
the same time it achieved that, it exposed it's own shortcoming: the
programming bottleneck
I agree, not like the ABC .... oh darn it, I still keep forgetting, the
ABC * didn't have * a program to slow it down. ;)
Goodnight.
25 Jan
25 Jan
6:33 p.m.
wasn't a universal machine
Please elaborate.
8:58 p.m.
On 2011 Jan 25, at 3:33 PM, Evan Koblentz wrote:
In the Universal Turing Machine sense, from computing theory. A UTM
can compute a certain class of functions. Any machine with a certain
set of capabilities can be shown to be a UTM equivalent .... I think
it would be difficult to construe the ENIAC into a UTM equivalent.
Oh. That's what I thought you meant, but I figured maybe you actually
meant something else, because ENIAC * was * Turing-complete.
I wouldn't normally use Wikipedia as a credible source, but since you
already did .... says here in the first paragraph:
http://en.wikipedia.org/wiki/ENIAC.
The chart (scroll down the Wiki page) shows how, except for the
limited Mark 1, * all * general-purpose programmable computers are
Turing-complete. (not counting theoretical impossibilities like
infinite memory.)
Yes, one does has to be careful about Wikipedia. To be clear, I wasn't
using wikipedia as a proof or evidence source.
I had never seen ENIAC characterised as Turing-complete, I thought
there was some limitation to it's programmibility and addressability
that would inhibit it, there are certain requirements for a machine to
be Turing-complete. Although as I alluded with the Minsky ref, I'm not
entirely surprised that it _might_ be possible. I will nonetheless wait
for the proof or explanation to give it credence, as well as see how
tortured the characterisation is (what it takes to accomplish it).
I do have to wonder if whoever made the characterisation is confusing
Turing-complete with a notion of programmibility or 'general-purpose'
(another ambiguous phrase).
The ENIAC page links to the "Turing completeness" page. There, Conway's
"The Game of Life" is stated to be Turing-complete. The point being, it
is interesting to look at these things from a strict or theoretical
point of view, and there is a wide range of things that can be
characterised as Turing-complete. So not to take away from the
Turing-complete matter (I introduced it), but let me put it another
way: there is a very large distinction between the architecture and
practical capabilities of the ENIAC (or an ENIAC-like machine) and the
stored-program machines that came after.
Or let me express it a little more practically: I can reasonably
imagine writing an Algol compiler or a word processor to run on the
EDVAC, even without knowing the instruction set of the EDVAC. Even the
lack of character handling could be worked around without much
difficulty. Doing so on the ENIAC is another matter, and I'm allowing
for having a large number of accumulators (data memory), as one would
in a Turing-completeness assessment. If it's Turing-complete, yes, one
can do it, but how tortured is it?
--
There are some curious things going on in the wikipedia "Turing
completeness" article. Looking at the revision history, the article
starts out with no ref to ENIAC, then this is added:
"The first machine known to be Turing-complete was ENIAC"
Later, to account for the paper on the earlier Z3 it becomes:
"Previously, the first machine known to be Turing-complete was ENIAC"
Somebody changes (reverts) that to:
"Globally however, the first machine known to be Turing-complete
continues to be ENIAC (1946)"
Whatever that means.
In Feb 2010 (and currently) it becomes:
"but the first machine explicitly designed to be Turing complete and
widely
appreciated as being universal was the 1946 ENIAC. This machine was
able to
solve a wide range of effective problems in the 1940s, many related
to
atomic bomb design."
That first bit is a rather strong statement of intent. Strikes me as a
little bit of revisionist history. I'd be rather surprised if Mauchly
and Eckert knew what Turing-complete meant when they designed the
ENIAC. The second sentence is a nice promotional statement about the
ENIAC, its presence in the Turing-completeness article is questionable.
wasn't a universal machine
Please elaborate.
9:09 p.m.
I'd be rather surprised if Mauchly and Eckert knew
what
Turing-complete meant when they designed the ENIAC.
I don't have any knowledge of who wrote what and when, re: Wikipedia.
As for "did M&E know about Turing" .... good question! The answer is
"almost certainly yes" because von Neumann knew Turing.
As for * when * von Neumann taught M&E about Turing -- if they didn't
already know -- that's something I will try to find out.
26 Jan
26 Jan
3:18 a.m.
On 2011 Jan 25, at 6:09 PM, Evan Koblentz wrote:
We're talking about the ENIAC, not the EDVAC. TMU, ENIAC was already
under design and construction when von Neumann became aware of it.
The question is not did they know about Turing, rather did they know
about Turing-completeness.
Turing's paper "On Computable Numbers" (it's on the web) was highly
abstract and in the realm of theory and the philosophy of logic at the
time, a long ways from people building calculating machines. The paper
was directed at other issues (note title) not characterising machine
architectures. I don't know the exact history of the theoretical
developments, but I suspect the notion of Turing-completeness came
later, a derivation from the paper. I really don't think people were
going around trying to characterise their implementations as
Turing-complete at the time of the ENIAC. I really don't see E & M
thinking "We must ensure the ENIAC has the functionality to make it
Turing-complete." Frankly, it sounds ludicrous.
One way or the other if people are going to make such claims as in
w'pedia it's for them to show them, not others to disprove them. I did
a search for "ENIAC Turing complete". All I've seen so far is limpid
statements of claim, such as:
"ENIAC also differed from earlier calculating devices in that it was
designed and used to be Turing-complete - that is, a truly universal
computing device"
Which just from the phrasing suggests the person writing it doesn't
know what they are talking about.
It's been sometime since I went looking for stuff on the 'first
computer'/ENIAC/ABC topic on the web.
Man, there is a lot of crap out there (and it's not limited to ENIAC
supporters).
I'd be rather surprised if Mauchly and Eckert
knew what
Turing-complete meant when they designed the ENIAC.
I don't have any knowledge of who wrote what and when, re: Wikipedia.
As for "did M&E know about Turing" .... good question! The answer is
"almost certainly yes" because von Neumann knew Turing.
As for * when * von Neumann taught M&E about Turing -- if they didn't
already know -- that's something I will try to find out. 
24 Jan
24 Jan
3:48 p.m.
On 2011 Jan 23, at 6:37 AM, Evan Koblentz wrote:
Because it depends on the definition one is working with. As usual, we
are in the realm of varying definitions. If we go with the modern
understanding of the word, neither the ABC or ENIAC was a computer.
Ditto Colossus (I believe).
I have no problem with that. They are significant machines, they are
certainy staps in the devlopment of the modern computer, but that doesn't
make thmm computers themselves. So?
And of course defitions change over time. 15 years ago (say), if you said
that yuour PC had 1GByte of memory, you would most likely have been
regarded as clueless because you were confuising hard disk size and RAM
size. But go back 35 years and companies like HP refered to 'disk memory'
and 'cassette memory' in their manuals. It's hard to claim that HP were
clueless.
And the defintiion of 'computer' has certain;y changed. I think the
original definiton was 'somebody who does calculations'. Then it bacame
'a person who operates a calculating machine'. And then the
(programamble) machine itself. I think one of the papers on what we would
now call a 'computer' (probably EDSAC) uses the word 'computer' to mean
what we would now call an ALU (or maybe the data path -- ALU + registers).
This site promote the hoary old argument that the
ABC wasn't really a
computer
How can any machine be a "computer" if it doesn't have any ability to
run a program? Perhaps people should stick to facts: "this did
that"; rather than
declarations: "this was that".
Sure. And I don't think anybody doubts that ENIAC made a contribution to
computing.
However, the term 'eelctronic digital computer' in its modern meaning
would seem to imply a stroed program. Byt that definition, ENIAC in its
origianl form was not a 'computer'. Whether ENIAC was the first machine
(after modifications) to run a stored program is something I do not know.
-tony
23 Jan
23 Jan
5:26 p.m.
On Sun, 23 Jan 2011, Evan Koblentz wrote:
electromechanical everything else; and that without
the ability to run a
program, it doesn't meet the definition of "computer".
Perhaps you could enlighten us with the definitive definistion of
"computer"?
You mean his plagiarizing of other people's work
and sticking his own
name on it, and then widely distributing that classified document
without permission? Yes, i agree. :)
like Vannevar Bush?
--
Grumpy Ol' Fred cisin at xenosoft.com
5:35 p.m.
electromechanical everything else; and that without the ability to run a program, it
doesn't meet the definition of "computer".
Perhaps you could
enlighten us with the definitive definistion of "computer"?
6:09 p.m.
On 2011 Jan 23, at 2:35 PM, Evan Koblentz wrote:
Contrary to what you're implying, I really do 100% agree that it's
subjective. BUT .... wouldn't everyone also agree that, at the very
least, to be a "computer" a machine must be programmable?
I know you want to distinguish between a computer and a calculator, and
of course there is a pertinent distinction to be made between
programmable devices and those that are not, but I think historically
'computer' is pretty much anything that computes. The Antikythera
mechanism is referred to as a computer. Such things as bombsights and
artillery aiming devices are application-specific analog computers.
They are configurable with different parameters, but they always solve
the same equation. They are different than programmable analog
computers, but nonetheless are still called computers.
If I want to distinguish something as programmable I call it
programmable.
electromechanical everything else; and that without the ability to
run a program, it doesn't meet the definition of "computer".
Perhaps
you could enlighten us with the definitive definistion of
"computer"?
6:16 p.m.
Contrary to
what you're implying, I really do 100% agree that it's
subjective. BUT .... wouldn't everyone also agree that, at the very
least, to be a "computer" a machine must be programmable?
I know you want to distinguish between a computer and a calculator,
and of course there is a pertinent distinction to be made between
programmable devices and those that are not, but I think historically
'computer' is pretty much anything that computes. The Antikythera
mechanism is referred to as a computer. Such things as bombsights and artillery aiming devices
are
application-specific analog computers. They are configurable with
different parameters, but they always solve the same equation. They
are different than programmable analog computers, but nonetheless are
still called computers.
Analog is a whole different can of worms. If it makes you happy then I
consent to adding the prefix "digital" to everything I've said so far.
24 Jan
24 Jan
4:09 p.m.
'computer' is pretty much anything that computes. The Antikythera
mechanism is referred to as a computer.
Actually I've been arguing for years that he Ant. is not a computer. Analog is a whole different can of worms. If it makes
you happy then I
consent to adding the prefix "digital" to everything I've said so far.
Isn't the Ant. device analogue?
-tony
25 Jan
25 Jan
2:53 p.m.
On 2011 Jan 24, at 1:09 PM, Tony Duell wrote:
I wasn't aware anyone conisdered it to be a computer. The earliest
known
mechncail calculating device, sure.
I'm not fully acquainted with the Antikythera device but, it was
composed of >30 gears, with some sort of differential mechanism in
there.
It solves a couple of equations in parallel. I'm not completely clear
on how complex those equations are, i.e. whether they are more than a
single simple relationship each, but it wasn't a simple device, it was
more than a rotary slide rule.
For the sake of discussion, here are some incremental definitions of
'computer' (not to say that these are the only possible definitions):
1 - something that performs a computation/calculation
2 - something that executes a program automatically (solves an
equation
where the equation is more than one simple arithmetic operation)
3 - something that is programmable (can solve a wide variety of
equations)
4 - a stored-program (/universal) machine
By some standards, something that fits *only* in category 1 is a
calculator, e.g. a slide rule is an analog calculator, an adding
machine is a digital calculator.
Parallellism adds another level of complexity in the characterisations.
And if one has some appreciation for the history of mathematics and
logic, one know it becomes another argument as to what constitutes a
computation or calculation.
The following categorisations could be made:
1: EDVAC, ENIAC, Harvard Mark 1, ABC, Antikythera, adding machine,
slide rule
2: EDVAC, ENIAC, Harvard Mark 1, ABC, Antikythera
3: EDVAC, ENIAC, Harvard Mark 1
4: EDVAC
It may be one could argue the Antikythera is in a grey area between 1
and 2.
Depending on your definition, one can call the Antikythera Device a
fixed-program analog computer.
-
One can pick or choose a definition and declare their device a computer
to suit one's purposes if one is so inclined. Things become a little
more controversial when one declares somebody else's device not a
computer, unless you're being completely clear about what definition
you're working with.
'computer' is pretty much anything that
computes. The Antikythera
mechanism is referred to as a computer.
Actually I've been arguing for years that he Ant. is not a computer.
3:32 p.m.
I'm not fully acquainted with the Antikythera
device
I highly recommend a book called "Decoding the Heavens". It came out a
year or two ago. It covers the latest research on this device. The
famous (but old) article "Gears of the Greeks" has been proven wrong by
researchers using modern optic tools.
3:35 p.m.
On 2011 Jan 24, at 1:09 PM, Tony Duell wrote:
I wasn't aware anyone conisdered it to be a computer. The earliest
known
mechncail calculating device, sure.
I'm not fully acquainted with the Antikythera device but, it was
composed of >30 gears, with some sort of differential mechanism in
there.
I am not very familiar with it either, I should probably find out more
about it...
'computer' is pretty much anything that
computes. The Antikythera
mechanism is referred to as a computer.
Actually I've been arguing for years that he Ant. is not a computer. It solves a couple of equations in parallel. I'm
not completely clear
on how complex those equations are, i.e. whether they are more than a
single simple relationship each, but it wasn't a simple device, it was
more than a rotary slide rule.
Sure. But to me that's stil a calculator. It's a special-purpose machine,
and I prefer to reserve the term 'computer' for more general devices. But
I understand that others use the term somewhat differnently
Actusally, like some others here, I find this 'Is it a computer or not'
argument somewhat pointless. There are things that everybody would class
as a computer (I don't think anyone here would claim the PDP8/e on my
desk was anyhting other than a computer). Other things that nobody would
claim to be a computer (the 746 telephone currently sitting next to it,
for example). Othre things thate may or may not be computers. How about
the HP9810 sitting alongside my desk? The makers called it a
'calcualtor', It its basic form it handles digits only. But it is
user-programmable, and can be expanded to have a simple form of character
I/O. Is that a computer? How abotu the servo system of an RK07 hard
drive. That's a dedicated analogue computer by some definitions.
When things are close to being computers, even if they are not computers
by some definitions -- things like the ABC, ENIAC, the Ant, mechnaism,
etc then I doubt anyone who has a serious interest in classic computing
could be uninterested in them. Comptuers or not, they are certainly
importat, interesting deviecs that are closely related to things that
everybody calls computers.
For the sake of discussion, here are some incremental definitions of
'computer' (not to say that these are the only possible definitions):
1 - something that performs a computation/calculation
2 - something that executes a program automatically (solves an
equation
where the equation is more than one simple arithmetic operation)
3 - something that is programmable (can solve a wide variety of
equations)
4 - a stored-program (/universal) machine
Waht do you call machins that contain a stored program, even a
use-alterable stored progrma with loops, conditionals and subroutines
(say) but which doesn't actually perform any numerical computations? I
have deisgned a few such things over there years -- special-purpose
programamble control systems that repond to inputs and generate outputs.
Computers? Sequencers? Programamble controllers? Whatever you call them
they must be related to at least part of a 'computer'
-tony
3:58 p.m.
For the sake
of discussion, here are some incremental definitions of
'computer' (not to say that these are the only possible definitions):
1 - something that performs a computation/calculation
2 - something that executes a program automatically (solves an
equation
where the equation is more than one simple arithmetic operation)
3 - something that is programmable (can solve a wide variety of
equations)
4 - a stored-program (/universal) machine
Waht do you call machins that
contain a stored program, even a use-alterable stored progrma with loops, conditionals and
subroutines (say) but which doesn't actually perform any numerical computations? I
have deisgned a few such things over there years -- special-purpose programamble control
systems that repond to inputs and generate outputs. Computers? Sequencers? Programamble
controllers? Whatever you call them they must be related to at least part of a
'computer'
26 Jan
26 Jan
1:39 p.m.
Wow .... Tony and I agree!!! It's a miracle.
It does happen from time to time...
It all goes back to my simplest possible definition .... computer =
automated calculating device that runs a program. :)
What do you mean by 'calculating'? Does this implky numerical computions
only?
Everything else (analog/digital;
mechanical/electronic; binary/decimal;
special-purpose/general-purpose; factory-tuned/user-programmable) --
they're all just subsets.
How do you define the 'program' of an analugue computer? The
interconnetions betwen the suming amplifiers, integrators, etc together
with the gains/functiuos of those units? If so, I guess patching together
digital function blocks (registers, adders, etc) is also a program, in
which case ENIAC was a programmable machine I think.
-tony
27 Jan
27 Jan
6:54 p.m.
On 2011 Jan 26, at 4:14 PM, Evan Koblentz wrote:
Suppose we wish to solve the equation: z=xy+5
On a programmable analog computer, we program it by:
- plugging a cable from A to B
- plugging a cable from C to D
- setting switch K to 5
On the ENIAC, we program it by:
- plugging a cable from A to B
- plugging a cable from C to D
- setting switch K to 5
The difference is how the machine *executes* the program (.. just an
implementation detail).
So the analog computer solves it simultaneously rather than
sequentially - is it any less a program?
Would be interesting to look into when the word began being used with
computing machinery. Was it used for the differential analysers?, the
Analytical engine (in the period)?, what about early mechanical
automatons?
What do you mean by 'calculating'? Does
this implky numerical
computions only?
My definition gets a tad awkward re: Colossus ...
How do you define the 'program' of an
analugue computer?
I was wondering the same thing. But I'm not losing sleep over it. :) 
8:55 p.m.
From: hilpert at cs.ubc.ca
Subject: Re: ENIAC ....
Date: Thu, 27 Jan 2011 15:54:57 -0800
To: cctalk at classiccmp.org
On 2011 Jan 26, at 4:14 PM, Evan Koblentz wrote:
Suppose we wish to solve the equation: z=xy+5
On a programmable analog computer, we program it by:
- plugging a cable from A to B
- plugging a cable from C to D
- setting switch K to 5
On the ENIAC, we program it by:
- plugging a cable from A to B
- plugging a cable from C to D
- setting switch K to 5
The difference is how the machine *executes* the program (.. just an
implementation detail).
So the analog computer solves it simultaneously rather than
sequentially - is it any less a program?
Would be interesting to look into when the word began being used with
computing machinery. Was it used for the differential analysers?, the
Analytical engine (in the period)?, what about early mechanical
automatons?
The Antikithera astronomical calculator 2K years ago, but there are probably calculating
and computing machines even before that.
A control system to regulate water for irrigation comes to mind, with a flow/level
measurement and balance to regulate water diversion into ditches. That probably goes back
500 - 1K years earlier.
What do you mean by 'calculating'? Does
this implky numerical
computions only?
My definition gets a tad awkward re: Colossus ...
How do you define the 'program' of an
analugue computer?
I was wondering the same thing. But I'm not losing sleep over it. :)
9:30 p.m.
A control system to regulate water for irrigation
comes to mind, with a flow/level measurement and balance to regulate water diversion into
ditches. ?That probably goes back 500 - 1K years earlier.
Well, you could say a caveman with a fire is also a feedback control
system - Og is cold, so he adds more wood. Og is warm, so he relaxes.
History is just one big grey area.
--
Will
6:33 p.m.
Well, you
could say a caveman with a fire is also a feedback control
system - Og is cold, so he adds more wood. Og is warm, so he relaxes.
Surely there wre feedback control systems in biological organisms long
before homo sapiens evolved.
27 Jan
27 Jan
9:56 p.m.
On 2011 Jan 27, at 5:55 PM, Randy Dawson wrote:
I guess my phrasing was a little ambiguous, I was trying to ask
(poorly) when the word "program" was first applied to computing
machinery, e.g. Did Charles and Ada use the word when talking about the
Analytical engine? Did Vannevar Bush use the word when talking about
setting up the Differential Analyser? ...
From: hilpert
at cs.ubc.ca
On 2011 Jan 26, at 4:14 PM, Evan Koblentz wrote:
Suppose we wish to solve the equation: z=xy+5
On a programmable analog computer, we program it by:
- plugging a cable from A to B
- plugging a cable from C to D
- setting switch K to 5
On the ENIAC, we program it by:
- plugging a cable from A to B
- plugging a cable from C to D
- setting switch K to 5
The difference is how the machine *executes* the program (.. just an
implementation detail).
So the analog computer solves it simultaneously rather than
sequentially - is it any less a program?
Would be interesting to look into when the word began being used with
computing machinery. Was it used for the differential analysers?, the
Analytical engine (in the period)?, what about early mechanical
automatons?
The Antikithera astronomical calculator 2K years ago, but there are
probably calculating and computing machines even before that.
A control system to regulate water for irrigation comes to mind, with
a flow/level measurement and balance to regulate water diversion into
ditches. That probably goes back 500 - 1K years earlier.
...
How do you define the 'program' of an analugue computer?
I was wondering the same thing. But I'm not losing sleep over it.
:)
10:03 p.m.
> I guess my phrasing was a little ambiguous, I was
trying to ask
> (poorly) when the word "program" was first applied to computing machinery
http://ieeexplore.ieee.org/search/freesrchabstract.jsp?navigation=no&ar…
10:28 p.m.
On 2011 Jan 27, at 7:03 PM, Evan Koblentz wrote:
That's nice, unfortunately the meat of it is behind the ieee access
walls. As the abstract says: "in the context of the ENIAC". That does
not preclude its use prior. There were programmable calculating
machines before the ENIAC.
> I guess my phrasing was a little ambiguous, I
was trying to ask
> (poorly) when the word "program" was first applied to computing
> machinery
http://ieeexplore.ieee.org/search/freesrchabstract.jsp?
navigation=no&arnumber=476561 
11:10 p.m.
New subject: English definitions (was ENIAC..)
I've been observing the discussion on the definition of "computer" and other
computer related words over the last few days/weeks. I've been involved in
forums for a number of years now and the most heated arguments seem to take
place regarding the definitions of particular words. Often they go along
the lines of who invented the first (widget/technique/concept) then evolve
into vigourous debates as to the precise definition of the said
widget/technique/concept etc.
I'm all for precise definitions. You need them for clarity of
communication. But the English language itself and the way it evolves is a
problem here. It can lead to an unavoidable fuzziness and drifting of
definitions over time. In my opinion it's hard (and in some cases
pointless) to be too pedantic about these things. The meaning of terms do
drift. Unlike French, there is no "authorised" version of English. The
various dictionaries we use do not actually define words but rather just
reflect their common usage and meanings at the time of publication. This is
both the strength of English (it's adaptable and dynamic) and it's weakness
(it can be ambiguous).
Anyway, just my two cents worth. I never lose much sleep over the percieved
sloppiness of a definition unless it's clearly out of that grey area that
seems to surround them all.
Terry (Tez)
28 Jan
28 Jan
3:48 p.m.
New subject: English definitions (was ENIAC..)
I've been observing the discussion on the definition of "computer" and
other
computer related words over the last few days/weeks. I've been involved in
forums for a number of years now and the most heated arguments seem to take
place regarding the definitions of particular words. Often they go along
the lines of who invented the first (widget/technique/concept) then evolve
into vigourous debates as to the precise definition of the said
widget/technique/concept etc.
I'm all for precise definitions. You need them for clarity of
communication. But the English language itself and the way it evolves is a
problem here. It can lead to an unavoidable fuzziness and drifting of
I think it was Martin Gardner who pointed out that many terms lie at
opposite ends ofa contimuum, but that does not make them worthless.
For example, nobody can give a precise wavelenght that separates
'blue' from 'green'. But this deosn't mean it's meaningless to say
that
the phospohr on the IBM5151 monitor I am using knw emits green light. Or
that the Intellec MCS8i is blue.
When we come ot 'first computer', how you define 'computer' clearly makes
a differnece to the answer. But I think we all know that whatever you
class as the 'first computer'. it was developed from earlier machiens
that you don't class as computers, but others might. What is important,
really, is to describe the machines accurately giving their mathod of
operation, etc. Then others can decider whehter or not _they_ want to
class them as computers, and more importantly can see how they fit into
the overall history.
-tony
6:06 p.m.
New subject: English definitions (was ENIAC..)
On 28 Jan 2011 at 20:48, Tony Duell wrote:
> I think it was Martin Gardner who pointed out that
many terms lie
at
opposite ends ofa contimuum, but that does not make
them worthless.
And the same words can mean something and its opposite. The verb "to
dust" for example. If I dust the furniture, I (hopefully) am
removing something from the surface of the coffee table. On the
other hand, if I dust my toast with powdered cinnamon, I'm putting
something on the surface of my Wonder Bread.
Yet, we don't seem to get the two meanings confused...
--Chuck
29 Jan
29 Jan
9:07 a.m.
New subject: English definitions (was ENIAC..)
On 28 Jan 2011 at 20:48, Tony Duell wrote:
Actually, at least how 'dustiong' is normally down over here, it means
getting the dust from the coffee table into the air, so that it lands
somewhere else. Personally, I'd rather have dust on the coffee table than
inside one of my comptuers :-)
> I think it was Martin Gardner who pointed out
that many terms lie
at
opposite ends ofa contimuum, but that does not
make them worthless.
And the same words can mean something and its opposite. The verb "to
dust" for example. If I dust the furniture, I (hopefully) am
removing something from the surface of the coffee table. On the other hand, if I dust my toast with powdered cinnamon,
I'm putting
something on the surface of my Wonder Bread.
Yet, we don't seem to get the two meanings confused...
I remember reading some chains of words where each word and the next were
almost synonyms, but the ends of the chain had completely opposite
meanings.
Of coruse the meanings of words also vaires with context. A 'bus driver'
might be the man who was sitting just in front and to the right of me on
a Routemaster about 4 hours ago. Or it might be a 74LS244 on a PCB in the
comptuer I was fixing. Again, we don't seem to get confused by this.,
-tony
25 Jan
25 Jan
4:08 p.m.
New subject: PDP is not a computer (was ENIAC)
Tony wrote:
(I don't think anyone here would claim the PDP8/e
on my
desk was anyhting other than a computer).
With tongue in cheek, I suggest that the PDP8/e is *not* a computer.
PDP stood for "Programmed Data Processor". DEC specifically did not
want to market their machines as computers, because they feared that the
bean counters that process purchasing requests for potential customers
for this machine might see the word "computer" and freak out envisioning
huge glass-walled rooms with gigantic mainframes and huge air
conditioning systems. The term "Programmed Data Processor" was an
attempt to avoid the stigma (for lack of a better word) that executive
types of the day had attached to the term computer.
Before flaming, please remember...tongue is firmly pressing on cheek.
Rick Bensene
4:41 p.m.
New subject: PDP is not a computer (was ENIAC)
On 25 Jan 2011 at 13:08, Rick Bensene wrote:
Before flaming, please remember...tongue is firmly
pressing on cheek.
Neither was the IBM 704 called a "computer" in IBM's literature of
the time. "Calculator" yes, and "Data Processing Machine", but not
"computer".
--Chuck
4:48 p.m.
New subject: PDP is not a computer (was ENIAC)
The term computer predates all this digital stuff it is a human who calculates
http://en.wikipedia.org/wiki/Human_computer
So the reasons you are all mentioning for not using the term should
include the common use at the time
Dave Caroline
On Tue, Jan 25, 2011 at 9:08 PM, Rick Bensene <rickb at bensene.com> wrote:
Tony wrote:
(I don't think anyone here would claim the
PDP8/e on my
desk was anyhting other than a computer).
With tongue in cheek, I suggest that the PDP8/e is *not* a computer.
PDP stood for "Programmed Data Processor". ?DEC specifically did not
want to market their machines as computers, because they feared that the
bean counters that process purchasing requests for potential customers
for this machine might see the word "computer" and freak out envisioning
huge glass-walled rooms with gigantic mainframes and huge air
conditioning systems. ?The term "Programmed Data Processor" was an
attempt to avoid the stigma (for lack of a better word) that executive
types of the day had attached to the term computer.
Before flaming, please remember...tongue is firmly pressing on cheek.
Rick Bensene
26 Jan
26 Jan
1:50 p.m.
New subject: PDP is not a computer (was ENIAC)
Tony wrote:
Actually, I think this is the very heart of the matter...
Abraham lincoln (I think) said (when arguing about a rather more serious
matter) that 'calling a tail a leg does not make it one'. I would add
that also calling a tail something else does not stop it from being a tail.
In this case, calling something a 'computer' does not necessarily make it
a computer by all definitions. It may be a computer by some definitions
and not others. It may not be a computer by any sensible defition
Similarly, the fact that the PDP8 was called a 'Programemd Data Processor'
and that the HP9830 [1] was called a 'Model 30 Calculator' by HP, both for
perfectly good marketing reasons, does not stop either of them being
classed as computers by a reasonable definition of the word.
[1] A desktoip machine with an alphanumeric display and BASIC in ROM.
It's hard to make a sensible definition of 'computer' that excludes it.
So, yes, you can, very reasoanbly, call ENIAC, Colossus, the ABC, etc,
'computers'. Equally, yuo may have a definition, such as requiring a
stored program, that excludes them. Neither is 'right' or 'wrong'.
-tony
(I don't think anyone here would claim the
PDP8/e on my
desk was anyhting other than a computer).=20
With tongue in cheek, I suggest that the PDP8/e is *not* a computer.
8:45 p.m.
New subject: PDP is not a computer (was ENIAC)
From: Rick Bensene
Sent: Tuesday, January 25, 2011 1:09 PM
Tony wrote:
> (I don't think anyone here would claim the
PDP8/e on my
> desk was anyhting other than a computer).
With tongue in cheek, I suggest that the PDP8/e is
*not* a computer.
PDP stood for "Programmed Data Processor".
DEC specifically did not
want to market their machines as computers, because they feared that the
bean counters that process purchasing requests for potential customers
for this machine might see the word "computer" and freak out envisioning
huge glass-walled rooms with gigantic mainframes and huge air
conditioning systems. The term "Programmed Data Processor" was an
attempt to avoid the stigma (for lack of a better word) that executive
types of the day had attached to the term computer.
Before flaming, please remember...tongue is firmly
pressing on cheek.
Certainly not flaming here. However...
The decision by DEC to name their first entry into the computer industry
the Programmed Data Processor 1, or PDP-1, was based on just such a view
of the market. However, by the time of the PDP-8/e, the naming was such
simply because it was DEC's trademarked designation for all of their
computers, used across 3 architectures (and soon after across 4).
Rich Alderson
Vintage Computing Sr. Server Engineer
Vulcan, Inc.
505 5th Avenue S, Suite 900
Seattle, WA 98104
mailto:RichA at vulcan.com
mailto:RichA at LivingComputerMuseum.org
http://www.PDPplanet.org/
http://www.LivingComputerMuseum.org/
4:02 a.m.
On 2011 Jan 25, at 12:35 PM, Tony Duell wrote:
You can substitute "equation" with "a problem in symbolic
logic".
If I understand correctly what you describe I would say it falls in
category 2.
I just used equation and arithmetic as this stuff is often discussed in
relation
to the historic developments.
Actusally, like some others here, I find this 'Is
it a computer or not'
argument somewhat pointless. There are things that everybody would
class
Yes, which is why I suggested at the very beginning of the thread
"characterise, don't declare".
For the sake
of discussion, here are some incremental definitions of
'computer' (not to say that these are the only possible definitions):
1 - something that performs a computation/calculation
2 - something that executes a program automatically (solves an
equation
where the equation is more than one simple arithmetic
operation)
3 - something that is programmable (can solve a wide variety of
equations)
4 - a stored-program (/universal) machine
Waht do you call machins that contain a stored program, even a
use-alterable stored progrma with loops, conditionals and subroutines
(say) but which doesn't actually perform any numerical computations? I
have deisgned a few such things over there years -- special-purpose
programamble control systems that repond to inputs and generate
outputs.
Computers? Sequencers? Programamble controllers? Whatever you call them
they must be related to at least part of a 'computer'
25 Jan
25 Jan
3:33 p.m.
> Perhaps you could enlighten us with the
definitive definistion of
> "computer"?
On Sun, 23 Jan 2011, Evan Koblentz wrote:
Contrary to what you're implying, I really do 100%
agree that it's
subjective. BUT .... wouldn't everyone also agree that, at the very
least, to be a "computer" a machine must be programmable?
Surprisingly, NO.
I was not implying that your definition was WRONG - your definition and
mine are pretty close. I was merely pointing out that unless we define
the term, then our definitions are no more defensible than, "a device that
connects me to eBay and Facebook"!
Are non-programmable "netbook"s available yet? The general public would
still call those computers, even more so than a 360 ("It doesn't even
have a screen or internet!")
It would not surprise me if Steve Jobs were to release a machine that can
not be programmed by anybody but HIS staff.
--
Grumpy Ol' Fred cisin at xenosoft.com
3:41 p.m.
Are non-programmable "netbook"s available
yet? The general public would still call those computers, even more so than a 360
("It doesn't even have a screen or internet!")
It would not surprise me if Steve Jobs were to release a machine that can not be
programmed by anybody but HIS staff.
I knew someone would eventually bring up the "appliance" variable. Sure
there are always appliance-like devices that * users * can't program,
but there is still * someone * would programs these, otherwise they
would be silicon doorstops.
4:14 p.m.
-----Original Message-----
From: cctalk-bounces at classiccmp.org [mailto:cctalk-
bounces at classiccmp.org] On Behalf Of Fred Cisin
Sent: Tuesday, January 25, 2011 12:33 PM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: ENIAC ....
It would not surprise me if Steve Jobs were to release
a machine that
can
not be programmed by anybody but HIS staff.
He did: the iPhone and iPad. IMHO, from Jobs' perspective, the people who write apps
*do* work for him.... -- Ian (who owns neither device)
4:14 p.m.
On 25 Jan 2011 at 12:33, Fred Cisin wrote:
On Sun, 23 Jan 2011, Evan Koblentz wrote:
A Jacquard loom is programmable and even runs loops...
*I* wouldn't call it a computer.
--Chuck
Contrary to what you're implying, I really do
100% agree that it's
subjective. BUT .... wouldn't everyone also agree that, at the very
least, to be a "computer" a machine must be programmable?
Surprisingly, NO. 
4:53 p.m.
On 25/01/2011 21:14, Chuck Guzis wrote:
On 25 Jan 2011 at 12:33, Fred Cisin wrote:
Ah, but he said "at least" -- so he means it's necessary, he didn't say
sufficient :-)
--
Pete Peter Turnbull
Network Manager
University of York
On Sun, 23 Jan 2011, Evan Koblentz wrote:
A Jacquard loom is programmable and even runs loops...
*I* wouldn't call it a computer. Contrary to what you're implying, I really do
100% agree that it's
subjective. BUT .... wouldn't everyone also agree that, at the very
least, to be a "computer" a machine must be programmable?
Surprisingly,
NO.
5:03 p.m.
Contrary to
what you're implying, I really do 100% agree that it's
subjective. BUT .... wouldn't everyone also agree that, at the
very least, to be a "computer" a machine must be programmable?
Surprisingly, NO.
5:11 p.m.
On 25 Jan 2011 at 21:53, Pete Turnbull wrote:
Ah, but he said "at least" -- so he means
it's necessary, he didn't
say sufficient :-)
Yes, but all the wiggle room is in that word, isn't it?
"When I use a word," Humpty Dumpty said, in rather a scornful tone,
"it means just what I choose it to mean--neither more nor less."
Otherwise known as a sucker's game.
--Chuck
26 Jan
26 Jan
1:56 p.m.
A Jacquard loom is programmable and even runs
loops...
*I* wouldn't call it a computer.
'A computer must be programamble' is not the same thing as 'Any
programmable device is a computer'. I would class a Jacquard loom as a
programamble non-computational device.
Another example : Some years ago I designed and built a device that
outputted various carefully-timed pulses tocontrol another device. The
unit was programamble, there werecounters that you could load under
program control and which then counted down, conditioanl jumps on a
counter getting to 0-, or on external trigger inputs, a 4 level return
stack so you could have subroutines in the control program, etc. But I
wouldn't call that a computer. A programamble control system, sure...
-tony
5:43 p.m.
On 26 Jan 2011 at 18:56, Tony Duell wrote:
'A computer must be programamble' is not the
same thing as 'Any
programmable device is a computer'. I would class a Jacquard loom as a
programamble non-computational device.
Would you consider the MC14500 to be a "computer"? Motorola didn't
term it as such (they called it an "Industrial Control Unit"), but I
submit that it can be made to calculate, is programmable, and so is a
computer.
--Chuck
27 Jan
27 Jan
6:27 a.m.
Chuck Guzis wrote:
Would you consider the MC14500 to be a
"computer"? Motorola didn't
term it as such (they called it an "Industrial Control Unit"), but I
submit that it can be made to calculate, is programmable, and so is a
computer.
In fact, the MC14500B architecture is very similar to that of the
PDP-14. I suspect that is not a coincidence, as they were intended for
essentially the same market.
Eric
11:42 a.m.
On 27 Jan 2011 at 3:27, Eric Smith wrote:
Chuck Guzis wrote:
Okay, let's drop down one step.
Is a PLC a computer? (basically not much more than a latch and an
EPROM)? It's programmable, and it can, after a fashion, be made to
calculate, given the right feedback terms. Basically a state
machine.
--Chuck
Would you consider the MC14500 to be a
"computer"? Motorola didn't
term it as such (they called it an "Industrial Control Unit"), but
I
> submit that it can be made to calculate, is programmable, and so
is a > computer.
In fact, the MC14500B architecture is very similar to that of the
PDP-14. I suspect that is not a coincidence, as they were intended
for essentially the same market.
2:52 p.m.
On 1/27/11 6:27 AM, Eric Smith wrote:
How tough are those MC14500B chips to find these days? I'd love to
mess with one.
-Dave
--
Dave McGuire
Port Charlotte, FL
Would you
consider the MC14500 to be a "computer"? Motorola didn't
term it as such (they called it an "Industrial Control Unit"), but I
submit that it can be made to calculate, is programmable, and so is a
computer.
In fact, the MC14500B architecture is very similar to that of the
PDP-14. I suspect that is not a coincidence, as they were intended for
essentially the same market.
5:34 p.m.
Dave McGuire wrote:
How tough are those MC14500B chips to find these days?
I'd love to
mess with one.
The only way I was able to get my hands on any was to buy some of the
DS14500A demonstration systems (consisting of the DS14500B demonstration
unit and DS14500C I/O simulator) on eBay.
I think there are only eight flip-flops in the chip, with only a small
amount of gating logic. I believe it might fit in a 22V10; there should
be no problem fitting it into a larger CPLD.
There an untested VHDL model on my web page:
http://www.brouhaha.com/~eric/retrocomputing/motorola/mc14500b/
Eric
3:30 p.m.
On 26 Jan 2011 at 18:56, Tony Duell wrote:
I can weasel my way out of that one by claiming that a chip -- on its own
-- is never a computer. It needs power, possibly a clock signal, etc
before it does anything.
More seriously, from what you say, I think I would consider it a computer
in some applications but maybe not in all. To give you another example,
is a Z80-based machine a compouter? Well, jasut behind me is a TRS-80
Model 4 which uses the Z80 as its processor,. I think most, if not all,
of us would call that a computer/ But next to it is an HP terminal that
uses a Z80 as its controller. It is not user programamble, it can only
ever runs the firmware. So I'd probably call that a terminal, not a
computer, at least when I'm using it (when I am repairing it, of course,
I have to know how the Z80 executes code, what the Z80 opcodes are, etc).
-tony
'A computer must be programamble' is not
the same thing as 'Any
programmable device is a computer'. I would class a Jacquard loom as a
programamble non-computational device.
Would you consider the MC14500 to be a "computer"? Motorola didn't
term it as such (they called it an "Industrial Control Unit"), but I
submit that it can be made to calculate, is programmable, and so is a
computer.
5:26 p.m.
On 27 Jan 2011 at 20:30, Tony Duell wrote:
More seriously, from what you say, I think I would
consider it a
computer in some applications but maybe not in all. To give you
another example, is a Z80-based machine a compouter? Well, jasut
behind me is a TRS-80 Model 4 which uses the Z80 as its processor,. I
think most, if not all, of us would call that a computer/ But next to
it is an HP terminal that uses a Z80 as its controller. It is not user
programamble, it can only ever runs the firmware. So I'd probably call
that a terminal, not a computer, at least when I'm using it (when I am
repairing it, of course, I have to know how the Z80 executes code,
what the Z80 opcodes are, etc).
In fact, quite a number of "computers" (particularly
microcontrollers) are used for non-computational (and often non-
programmable) applications. (e.g. a simple timer).
Does that mean that a thing being a "computer" must also be used in
an application that requires it to act as such?
Or is a computer independent of the application and must merely be
"capable" (whatever that means) of being programmed?
--Chuck
28 Jan
28 Jan
3:15 p.m.
More
seriously, from what you say, I think I would consider it a
computer in some applications but maybe not in all. To give you
another example, is a Z80-based machine a compouter? Well, jasut
behind me is a TRS-80 Model 4 which uses the Z80 as its processor,. I
think most, if not all, of us would call that a computer/ But next to
it is an HP terminal that uses a Z80 as its controller. It is not user
programamble, it can only ever runs the firmware. So I'd probably call
that a terminal, not a computer, at least when I'm using it (when I am
repairing it, of course, I have to know how the Z80 executes code,
what the Z80 opcodes are, etc).
In fact, quite a number of "computers" (particularly
microcontrollers) are used for non-computational (and often non-
programmable) applications. (e.g. a simple timer).
27 Jan
27 Jan
5:29 p.m.
Chuck wrote:
Would you consider the MC14500 to be a
"computer"?
Tony wrote:
More seriously, from what you say, I think I would
consider it a computer
in some applications but maybe not in all.
Note that the MC14500B did not contain a program counter. It was really
just a one-bit ALU with a small amount of logic to interface to an
external sequencer. That sequencer could simply be a binary counter, if
you didn't need branching, or it could be something elaborate with a
subroutine stack. The MC14500B Industrial Control Unit Handbook gives
an example using three Fairchild Macrologic CMOS 34706 control sequencer
bit slices, which are conceptually similar to the Am2909/2911.
(Motorola doesn't mention Fairchild by name, and it took me a while to
figure out what they were, because at some point Fairchild changed the
part number from 34706 to just 4706.)
Eric
28 Jan
28 Jan
3:18 p.m.
Chuck wrote:
I don;t think I would consider the 14500 on its own as a 'computer'. But
I don't see why a computer can't be built using one or more of them.
FWIW, I don't think the Z80 (replace with your vavourit MPU) is a
computer. But it can certaiuly be used as the processor in something I
would call a computer.
Would you consider the MC14500 to be a
"computer"?
Tony wrote:
More seriously, from what you say, I think I
would consider it a computer
in some applications but maybe not in all.
Note that the MC14500B did not contain a program counter. It was really
just a one-bit ALU with a small amount of logic to interface to an
external sequencer. That sequencer could simply be a binary counter, if you didn't need branching, or it could be
something elaborate with a
subroutine stack. The MC14500B Industrial Control Unit Handbook gives
Is this book on-line anywhere? I've not checked the obvious places, so if
it's on bitsavers or datasheetarchive or similar, don't flame too hard ;-)
-tony
4:41 p.m.
I wrote:
The MC14500B Industrial Control Unit Handbook gives
Tony Duell wrote:
Is this book on-line anywhere?
http://www.brouhaha.com/~eric/retrocomputing/motorola/mc14500b/
5:27 p.m.
New subject: MC14500 / was Re: ENIAC ....
On 2011 Jan 28, at 1:41 PM, Eric Smith wrote:
I wrote:
Are there any known examples of real-world products or equipment that
used the 14500 ?
(Leaving aside evaluation boards & such.)
I had the impression of it as a bit of a stillborn product, like the
Fairchild CMOS Macrologic series (34700/4700 ICs), of limited
cost-effectiveness or benefit by the time it was introduced.
Anyone know when was it introduced? Earliest ref I have is 1979 in a
Motorola full-line catalog.
The MC14500B Industrial Control Unit Handbook
gives
Tony Duell wrote:
Is this book on-line anywhere?
http://www.brouhaha.com/~eric/retrocomputing/motorola/mc14500b/
9:39 p.m.
New subject: MC14500 / was Re: ENIAC ....
Brent Hilpert wrote about the MC14500B:
Anyone know when was it introduced? Earliest ref I
have is 1979 in a
Motorola full-line catalog.
The date of the second printing of the MC14500B Industrial Control Unit
Handbook is December 1977. I don't have the first printing.
10:50 p.m.
New subject: MC14500 / was Re: ENIAC ....
On 28 Jan 2011 at 18:39, Eric Smith wrote:
Brent Hilpert wrote about the MC14500B:
It's also worth noting that the thing does not appear in the 1981
Motorola Microprocessor Databook. So it's obvious that Motorola did
not consider it to be a microprocessor. (duh!)
--Chuck
Anyone know when was it introduced? Earliest ref
I have is 1979 in
a
Motorola full-line catalog.
The date of the second printing of the MC14500B Industrial Control
Unit Handbook is December 1977. I don't have the first printing.
29 Jan
29 Jan
9:13 a.m.
New subject: MC14500 / was Re: ENIAC ....
It's also worth noting that the thing does not
appear in the 1981
Motorola Microprocessor Databook. So it's obvious that Motorola did
not consider it to be a microprocessor. (duh!)
My experience, however, is that manufacturers didn't always put chips in
the 'obvious' databooks, possibly make customers buy several books to get
all the data sheets they needed. IIRC, my Motoroal 8-bit microiprocessor
databook (some years later than that) doesn't include things like the
6803 or 6805, there was a sepatate 'microcontrolelr' databook I think.
Perhaps the most annoyine one is that the Intel microprocessor and
peripherals databook in the early 1990s didn't include the 8250 or 8251 (it
did include the 8253, 8255, etc). There was a 'communciatons' databook
that inlcuded thye seiral chips.
-tony
3:30 p.m.
New subject: databooks (was Re: MC14500 / was Re: ENIAC ....)
Tony Duell wrote:
My experience, however, is that manufacturers
didn't always put chips in
the 'obvious' databooks, possibly make customers buy several books to
get
all the data sheets they needed.
Despite there being prices printed on the covers of some databooks,
customers didn't buy them. The sales reps and distributors all but
forced them onto customers.
The prices were what other people such as hobbyists paid.
Eric
3:49 p.m.
New subject: databooks (was Re: MC14500 / was Re: ENIAC ....)
On 2011 Jan 29, at 12:30 PM, Eric Smith wrote:
Tony Duell wrote:
I must have been lucky as a kid, I don't think I was ever charged for a
databook at the industrial/commercial component sellers. I think I did
pay for a couple at Radio Shack where they sold Nat Semi reprints.
-
In the 1979 Motorola catalog, the 14500 is listed both with "CMOS
Processors" (6805,141000,141200..) and in the standard CMOS 4xxx
SSI/MSI section.
My experience, however, is that manufacturers
didn't always put
chips in
the 'obvious' databooks, possibly make
customers buy several books
to get
all the data sheets they needed.
Despite there being prices printed on the covers of some databooks,
customers didn't buy them. The sales reps and distributors all but
forced them onto customers.
The prices were what other people such as hobbyists paid. 
4:20 p.m.
New subject: databooks (was Re: MC14500 / was Re: ENIAC ....)
It is a requirement of the publishing and book business that if you are
going to put a product on the shelf you need a price and other bits of
info. Even if a book was never sold, some manufacturers always put the
extra on the book just in case. and sometimes I'm sure it was the
contractor who did the book for the chip people who routinely did that
on their product as well.
Reps were always the best ones to get to, followed by a distributor from
who you purchased a reasonable amount of product. Hamilton was always
nice as was Avnet before they merged. They would let you go into their
library and just load up here in Irvine. Sometimes they would ask
questions because they wanted to track the projects at your company, and
if you were loading up on weird stuff you'd tell them it was for your
own research.
Jim
4:57 p.m.
New subject: databooks (was Re: MC14500 / was Re: ENIAC ....)
Despite there being prices printed on the covers of
some databooks,
customers didn't buy them. ?The sales reps and distributors all but forced
them onto customers.
The prices were what other people such as hobbyists paid.
I doubt there were many hobbyists that actually purchased those
databooks. Everyone I knew back in the databook era would make up a
phony company name, and then get flooded with databooks, industrial
periodicals, and best of all, samples! Yes, all those tech reps knew
that they were phony companies, and most of what they were handing out
for free would yield nothing, but they also knew that a few of those
freebies might pay back in a big way.
I still get junkmail with the company name I came up with 28 years ago!
Basically, only fools paid the cover price for databooks or the industrial rags.
--
Will
30 Jan
30 Jan
12:32 a.m.
New subject: databooks (was Re: MC14500 / was Re: ENIAC ....)
On Sat, Jan 29, 2011 at 4:57 PM, William Donzelli <wdonzelli at gmail.com> wrote:
... Everyone I knew back in the databook era would
make up a
phony company name, and then get flooded with databooks, industrial
periodicals, and best of all, samples!
I figured that one out when I was about 16. I needed a fake company
name or I couldn't ever get datasheets, etc. After I invented one,
the stuff all flooded in (mostly, I was interested in RCA COSMAC stuff
and Synertek/Rockwell 6502-family stuff, though the preliminary 68000
datasheet I latched onto held specs beyond my wildest 8-bit dreams).
I still get junkmail with the company name I came up
with 28 years ago!
Me too.
-ethan
10:08 a.m.
New subject: databooks (was Re: MC14500 / was Re: ENIAC ....)
Despite there being prices printed on the covers
of some databooks,
customers didn't buy them. =A0The sales reps and distributors all but for=
ced
them onto customers.
The prices were what other people such as hobbyists paid.
I doubt there were many hobbyists that actually purchased those
databooks. Everyone I knew back in the databook era would make up a
phony company name, and then get flooded with databooks, industrial
periodicals, and best of all, samples! Yes, all those tech reps knew
that they were phony companies, and most of what they were handing out
for free would yield nothing, but they also knew that a few of those
freebies might pay back in a big way.
I still get junkmail with the company name I came up with 28 years ago!
Basically, only fools paid the cover price for databooks or the industrial =
rags.
29 Jan
29 Jan
9:02 a.m.
I wrote:
THnaks... I've had a quickj look and will look at it more carefully when
I hae some spare time. From what I've seen, I certainly wouldn't call this
chip on its own a 'computer', but I don't see why it could't be part of
one.
It's a curious device though. Even at the time it was made it didn't
replace that many IC packages, and if you're not careful, using something
like this means the solution to the real problem becomes a kludge so as
to use said chip. I am not at all suprsied it wasn't that popular.
-tony
The MC14500B Industrial Control Unit Handbook
gives
Tony Duell wrote:
Is this book on-line anywhere?
http://www.brouhaha.com/~eric/retrocomputing/motorola/mc14500b/ 
28 Jan
28 Jan
11:54 p.m.
From: ard at p850ug1.demon.co.uk
I don;t think I would consider the 14500 on its own as a 'computer'. But
I don't see why a computer can't be built using one or more of them.
Hi Tony
I always thought is would be fun to make a variable
bit serial computer. Many operations only need a single bit
while the single bit ALU could operate on data, like
adding, one bit at a time and use any length one wanted.
Dwight
29 Jan
29 Jan
1:46 a.m.
On 28 Jan 2011 at 20:54, dwight elvey wrote:
I always thought is would be fun to make a variable
bit serial computer. Many operations only need a single bit
while the single bit ALU could operate on data, like
adding, one bit at a time and use any length one wanted.
How about instead of bits, 4-bit digits (perhaps with a programmable
switch between BCD and binary math)? I suspect that it wouldn't be
much more complicated than 1-bit serial design. Add an extra bit to
each digit as a flag and you've got yourself a 1620. :)
--Chuck
3:32 a.m.
New subject: bit-serial processors (was Re: ENIAC ....)
Dwight wrote:
I always thought is would be fun to make a variable
bit serial computer. Many operations only need a single bit
while the single bit ALU could operate on data, like
adding, one bit at a time and use any length one wanted.
Chuck Guzis wrote:
How about instead of bits, 4-bit digits (perhaps with
a programmable
switch between BCD and binary math)? I suspect that it wouldn't be
much more complicated than 1-bit serial design.
You've just described the architecture of the proprietary processors
used in Hewlett-Packard calculator models released between 1972 and 1999.
There were four versions of the architecture that were bit-serial but
did arithmetic one decimal digit at a time, or four binary bits. The
word size was 48, 56, or 64 bits, and arithmetic operations could work
on any range of digits (or aligned four-bit binary groups) of a word.
Classic: 56-bit, BCD only, strict Harvard architecture (10-bit ROM
words in separate address space)
Woodstock/Spice/Topcat: 56-bit, selectable BCD/binary, strict Harvard
architecture, 10-bit ROM
Cricket: 48-bit, selectable BCD/binary, strict Harvard architecture,
10-bit ROM
Nut: 56-bit, BCD/binary, 10-bit ROM, but with instructions to read ROM
words as data (modified Harvard Architecture)
Saturn: 64-bit, nibble serial (rather than bit-serial), von Neuman
architecture (unified address space, instructions encoded in variable
length of consecutive 4-bit nibbles)
Starting in the 1990s, HP replaced the custom architecture(s) with
commercial microcontrollers and microprocessors, including the Hitachi
H8, SunPlus/GeneralPlus 6502 derivatives, Samsung S3C2410 ARM920T
(graphing calculators), and most recently the Atmel AT91SAM7L128 ARM7.
The graphing calculators still run a simulation of the Saturn
architecture for legacy code.
The use of the Atmel part in recent models (HP 12c, HP 20b, HP 30b) is
nice because it is a flash-based part, so you can write your own
replacement firmware for it, something that was not previously possible
in entry to midrange HP calculators.
Eric
9:32 a.m.
New subject: bit-serial processors (was Re: ENIAC ....)
Dwight wrote:
My first thoguht was of a slightly larger HP calculator, the 9800 series.
[...]
I always thought is would be fun to make a
variable
bit serial computer. Many operations only need a single bit
while the single bit ALU could operate on data, like
adding, one bit at a time and use any length one wanted.
Chuck Guzis wrote:
How about instead of bits, 4-bit digits (perhaps
with a programmable
switch between BCD and binary math)? I suspect that it wouldn't be
much more complicated than 1-bit serial design.
You've just described the architecture of the proprietary processors
used in Hewlett-Packard calculator models released between 1972 and 1999. Classic: 56-bit, BCD only, strict Harvard
architecture (10-bit ROM
words in separate address space)
Woodstock/Spice/Topcat: 56-bit, selectable BCD/binary, strict Harvard
architecture, 10-bit ROM
The HP919C must be in this group too (same ACT processor), what about the
other Sting model, the 10A? And of course the 67 uses an ACT too, even
though it's sometimes clased as a 'classic series' by people who think
the case is nor interesting than the procesor :-)
-tony
9:22 a.m.
=20
Are you implying that the data registers should be variable length, or
that a variable number of bits should be used in each instruction?
Perhaps I can be clearer. The operation of reading a bit from a register
on a bit-serial machine normally inplies a rotation of that register, so
that repeating the operation reads the next bit, and so on.
Now if the register is, say, 16 bits long, do you want a '7 bit'
operation to roatate the whole register by 7 bits (so that it's different
at the end of the operation), or a rotation of the bottom 7 bits only
(which would leave the register unchanged at the end of the operation)?
Incidentally, I have the schematics for a bit-serial processor in front
of me at the moment... The HP9800. That's mostly [1] bit serial, and the
microcode clock circuitry alows a given bit operation to be repeated 1
to 16 times. So that an add of 2 16 bit reigsters is one
microinstruciton, and an increment takes 2 microinstruction (add 1 once,
and 0 15 times).
[1] Yes, mostly. There are a couple of BCD operations between the A
accumator and the T memory data register that oeprate a nybble at a time.
-tony
From: ard at p850ug1.demon.co.uk
I don=3Bt think I would consider the 14500 on its own as a 'computer'. Bu=
t=20
I don't see why a computer can't be built
using one or more of them.
=20
Hi Tony
I always thought is would be fun to make a variable
bit serial computer. Many operations only need a single bit
12:47 p.m.
On 29 Jan 2011 at 14:22, Tony Duell wrote:
Are you implying that the data registers should be
variable length, or
that a variable number of bits should be used in each instruction?
Why any (visible) registers at all? Memory-to-memory was the basis
for several IBM computers. The aforementioned 1620 was a 2-address
machine with variable word length with no visible registers (the P-
counter was not readable by a program). Word length was essentially
limited by the size of a bank of memory (about 20K decimal digits).
Word length was at least 2 digits, with an extra bit serving the dual
purpose of a sign (if set on the low-order digit of a word) or a
demarcator (if set on the high-order digit).
While it's true that the 1620 used 5 digit address registers for
memory access and the P-counter, none was readable by a program.
It sounds like it was clumsy, but the architecture was very friendly
to hordes of early 1960s students. Being decimal, it was a system
whose learning curve was not steep. The usual unit record
interfaces, in addition to a console typewriter and an optional disk
interface (to the 1311) were present.
Arithmetic was performed by table lookup, with the addition and
multiplication tables located in low memory and alterable by user
programs.
Dijkstra hated the machine, but it was popular and as the 1710, was
one of IBM's first industrial process control computers.
--Chuck
3:31 p.m.
New subject: Variable length (in bits) instructions (was Re: ENIAC ....)
Tony Duell wrote:
Are you implying that the data registers should be
variable length, or
that a variable number of bits should be used in each instruction?
The only processor I've ever seen for which instructions are a variable
number of bits (not bytes or words) is the Intel iAPX 432. For the
release 1 GDP, instructions ranged from 6 to 344 bits long.
Eric
4:37 p.m.
New subject: Variable length (in bits) instructions (was Re: ENIAC ....)
Tony Duell wrote:
IIRC, the Transputer is just plain weird here. You have to build up the
instruction in an internal register before you exexute it (there are
opcodes that modify this instruction register in various ways). I think
the most common instructions can be loaded in 1 byte.
It's rather different to the multi-byte instructions on otehr processor
anyway.
-tony
Are you implying that the data registers should
be variable length, or
that a variable number of bits should be used in each instruction?
The only processor I've ever seen for which instructions are a variable
number of bits (not bytes or words) is the Intel iAPX 432. For the
release 1 GDP, instructions ranged from 6 to 344 bits long. 
30 Jan
30 Jan
6:43 a.m.
New subject: Variable length (in bits) instructions (was Re: ENIAC ....)
On 29/01/2011 21:37, Tony Duell wrote:
IIRC, the Transputer is just plain weird here. You
have to build up the
instruction in an internal register before you exexute it (there are
opcodes that modify this instruction register in various ways). I think
the most common instructions can be loaded in 1 byte.
It's rather different to the multi-byte instructions on otehr processor
anyway.
-tony
A long time ago I wrote a Java JIT compiler for the Transputer and IIRC
the instructions are variable length in mutiples of 4 bits, with as Tony
points out the smallest being 8 bits. It also has 3 general purpose
registers (A, B & C) organised as a stack. A colleague did look at
creating the code generator for a recent GCC and left the company
shortly after ;)
A kind soul has uploaded the ISR here if anyone is _really_ interested :
http://www.transputer.net/iset/pdf/tis-acwg.pdf
James
31 Jan
31 Jan
4:09 p.m.
New subject: Variable length (in bits) instructions (was Re: ENIAC ....)
On 29/01/2011 21:37, Tony Duell wrote:
The odd thing is that provide you have the right value in the (internal)
intruction register when you come to execte it, it doens't matter how you
got it there. So in genral there is no unique set of external
instrucitons for a given machine instruciton. There are various ways of
getting bit patterns in the instruciton regiaster.
The Transputer data sheets do not give the instruciton set IIRC. You
weren't supposed to need it, but rather to use the Occam compiler from
Inmos. There is a book something like 'The Transputer Instruction Set --
a Compilker Writer's Guide' which gives the instruciton set, etc. It's on
my bookshelf, I assume it's on a web site somwehrr...
-tony
IIRC, the Transputer is just plain weird here.
You have to build up the
instruction in an internal register before you exexute it (there are
opcodes that modify this instruction register in various ways). I think
the most common instructions can be loaded in 1 byte.
It's rather different to the multi-byte instructions on otehr processor
anyway.
-tony
A long time ago I wrote a Java JIT compiler for the Transputer and IIRC
the instructions are variable length in mutiples of 4 bits, with as Tony
points out the smallest being 8 bits. It also has 3 general purpose
5:13 p.m.
New subject: Variable length (in bits) instructions (was Re: ENIAC ....)
When it comes to variable-length instructions, I'm surprised that no
one mentioned the IBM 1401, particularly in its ability to "chain"
instructions.
--Chuck
5:30 p.m.
New subject: Variable length (in bits) instructions (was Re: ENIAC ....)
From: Chuck Guzis
Sent: Monday, January 31, 2011 2:13 PM
When it comes to variable-length instructions, I'm
surprised that no
one mentioned the IBM 1401, particularly in its ability to "chain"
instructions.
The instructions on the 1401 are 1 character long, that is, 6 bits.
Or have I misunderstood the discussion so far. I thought we were talking
about different operations being effected by different lengths of bits as
the op codes.[1]
[1] Note that "e", as opposed to "a".
Rich Alderson
Vintage Computing Sr. Server Engineer
Vulcan, Inc.
505 5th Avenue S, Suite 900
Seattle, WA 98104
mailto:RichA at vulcan.com
mailto:RichA at LivingComputerMuseum.org
http://www.LivingComputerMuseum.org/
6:29 p.m.
New subject: Variable length (in bits) instructions (was Re: ENIAC ....)
On 31 Jan 2011 at 14:30, Rich Alderson wrote:
From: Chuck Guzis
Sent: Monday, January 31, 2011 2:13 PM
Are we talking about instructions themselves, or their operands? In
either case, the 1401 qualifies--it's a variable word-length machine
and instruction length is dependent on the presence or absence of
operands.
So, there is only one opcode for the "ADD" instruction, but I can
write:
A 123 456
Causing the (variable-length) number whose low-order digit is
addressed by 123 to be added to the (variable-length) number whose
low-order digit is addressed by 456. I can also write:
A 123
Same opcode--but this indicates that the number addressed by 123 is
added to itself. Or I can write:
A
Which adds two numbers preceding (in memory) the pair last addressed
by an A or other instruction.
Similarly, the "unconditional branch" and the "conditional branch"
instructions use a common opcode, but the conditional branch appends
a digit that specifies the condition.
The "trick" behind all of this is that the instruction execution
logic locates the opcode by scanning until it finds a character with
the word mark set. While it's scanning, operands are shifted through
the operand regsters.
So, I submit that the 1401 is a variable length machine (it's
decimal, by the way) both in the sense of data that's being
manipulated and the instructions themselves.
Have I missed a third category of "variable length"?
--Chuck
When it comes to variable-length instructions,
I'm surprised that no
one mentioned the IBM 1401, particularly in its ability to "chain"
instructions.
The instructions on the 1401 are 1 character long, that is, 6 bits.
Or have I misunderstood the discussion so far. I thought we were
talking about different operations being effected by different lengths
of bits as the op codes.[1]
7:19 p.m.
New subject: Variable length (in bits) instructions (was Re: ENIAC ....)
From: Chuck Guzis
Sent: Monday, January 31, 2011 3:30 PM
On 31 Jan 2011 at 14:30, Rich Alderson wrote:
> From: Chuck Guzis
> Sent: Monday, January 31, 2011 2:13 PM
>> When it comes to variable-length instructions,
I'm surprised that no
>> one mentioned the IBM 1401, particularly in its ability to "chain"
>> instructions.
> The instructions on the 1401 are 1 character long,
that is, 6 bits.
> Or have I misunderstood the discussion so far. I
thought we were
> talking about different operations being effected by different lengths
> of bits as the op codes.[1]
Are we talking about instructions themselves, or their
operands?
[snip long discussion of 1401, which was my first computer 42 years ago]
Have I missed a third category of "variable
length"?
I don't know. Have you? Eric Smith wrote the following:
From: Eric Smith
Sent: Saturday, January 29, 2011 12:32 PM
Tony Duell wrote:
> Are you implying that the data registers should be variable length, or
> that a variable number of bits should be used in each instruction?
The only processor I've ever seen for which
instructions are a variable
number of bits (not bytes or words) is the Intel iAPX 432. For the
release 1 GDP, instructions ranged from 6 to 344 bits long.
commenting on Tony Duell's followup to Dwight Elvey's message:
> From: dwight elvey
> Sent: Friday, January 28, 2011 8:54 PM
> I always thought is would be fun to make a
variable
> bit serial computer. Many operations only need a single bit
> while the single bit ALU could operate on data, like
> adding, one bit at a time and use any length one wanted.
Notice that Dwight is talking about variable numbers of *bits* in the
instruction (by which I think he means "op code"), and Tony and Eric
both appear to at elast wonder if that's what he means.
Rich Alderson
Vintage Computing Sr. Server Engineer
Vulcan, Inc.
505 5th Avenue S, Suite 900
Seattle, WA 98104
mailto:RichA at vulcan.com
mailto:RichA at LivingComputerMuseum.org
http://www.LivingComputerMuseum.org/
7:31 p.m.
New subject: Variable length (in bits) instructions (was Re: ENIAC ....)
On 31 Jan 2011 at 16:19, Rich Alderson wrote:
I interpreted Dwight's post as a 1-bit serial ALU operating on
variable length fields.
As far as variable-length opcodes, again, that's a matter of opinion.
Many microcontrollers have variable length opcodes. with the
instruction packed into a fixed-length word.
--Chuck
> I always
thought is would be fun to make a variable
> bit serial computer. Many operations only need a single bit
> while the single bit ALU could operate on data, like
> adding, one bit at a time and use any length one wanted.
Notice that Dwight is talking about variable numbers of *bits* in the
instruction (by which I think he means "op code"), and Tony and Eric
both appear to at elast wonder if that's what he means.
1 Feb
1 Feb
1:56 p.m.
New subject: Variable length (in bits) instructions (was Re: ENIAC ....)
From: cclist at sydex.com
On 31 Jan 2011 at 16:19, Rich Alderson wrote:
I interpreted Dwight's post as a 1-bit serial ALU operating on
variable length fields.
As far as variable-length opcodes, again, that's a matter of opinion.
Many microcontrollers have variable length opcodes. with the
instruction packed into a fixed-length word.
--Chuck
Hi
I'll have to admit, I was mostly talking about ALU operations.
One thing that I'd often though might be cool to play with
is context base instruction decoding. This could make a difference
in instructions based on operations just performed. This might
be in packed instructions of fixed length or variable length with
a serial memory.
Dwight
> I
always thought is would be fun to make a variable
> bit serial computer. Many operations only need a single bit
> while the single bit ALU could operate on data, like
> adding, one bit at a time and use any length one wanted.
Notice that Dwight is talking about variable numbers of *bits* in the
instruction (by which I think he means "op code"), and Tony and Eric
both appear to at elast wonder if that's what he means. 
2 Feb
2 Feb
12:29 p.m.
New subject: Variable length (in bits) instructions (was Re: ENIAC ....)
On 1/31/2011 4:13 PM, Chuck Guzis wrote:
When it comes to variable-length instructions, I'm
surprised that no
one mentioned the IBM 1401, particularly in its ability to "chain"
instructions.
--Chuck
I'd completely forgotten the 1401 could do instruction chaining. Now
I'm going to have to go find my instruction card or something to jog my
memory as to how and where that was done. :-)
IIRC, the 1401 instruction lengths varied only in increments of 6 (or 7
or 8) bits, depending on whether one counts the word mark and/or parity
bit as part of each instruction character.
Later,
Charlie C.
1 Feb
1 Feb
3:39 a.m.
New subject: Variable length (in bits) instructions (was Re: ENIAC ....)
The odd thing is that provide you have the right value
in the (internal)
intruction register when you come to execte it, it doens't matter how you
got it there. So in genral there is no unique set of external
instrucitons for a given machine instruciton. There are various ways of
getting bit patterns in the instruciton regiaster.
The Transputer data sheets do not give the instruciton set IIRC. You
weren't supposed to need it, but rather to use the Occam compiler from
Inmos. There is a book something like 'The Transputer Instruction Set --
a Compilker Writer's Guide' which gives the instruciton set, etc. It's on
my bookshelf, I assume it's on a web site somwehrr...
-tony
You are correct - the instruction set isn't in the data sheet. The
compiler writers guide (appendix E in
http://www.transputer.net/iset/pdf/tis-acwg.pdf) does indeed have the
opcodes listed. IIRC the processor reads one nibble of the instruction
that containins the opcode, where 0..14 are frequently used instructions
(pop, add, etc) and 15 is a jump out to an extended set. The instruction
may be followed by immediate data. There's also the 'pfix' opcode that's
used to extend the instruction or the immediate data fields by an
arbitary amount. I'm tempted to dig out that JIT work now - this was all
very clear at the time ;)
James
23 Jan
23 Jan
6:26 p.m.
On 23 Jan 2011 at 14:26, Fred Cisin wrote:
On Sun, 23 Jan 2011, Evan Koblentz wrote:
Well, if you collect such documents, pull out the ANSI standards
document for COBOL or FORTRAN and read the preface commentary. A
"computer" is designated by anything, and particularly humans, that
computes. Period.
--Chuck
electromechanical everything else; and that
without the ability to
run a program, it doesn't meet the definition of "computer".
Perhaps you could enlighten us with the definitive definistion of
"computer"?
12:59 p.m.
[...]
The fact that Colossus was purpose-built and served to
decipher codes shoul=
d not take away from the importance of its design. It was possibly as 'pro=
Whether or not a device is important (and I will happily agree that both
Colossus and ENIAC are historically and socially importnat) has nothing
whatsoever to do with whether it's a computer.
grammable' as the ENIAC, but the scope of
programming development was somew=
hat limited by fear of annihilation at the hands of the Nazi regime. One m=
ust wonder what would have become of it had its destruction not been ordere=
d and its very existence declared a state secret. =20
I have yet to see any proper technical information on either ENIAC or
Colossus. Alas most of the technical-looking links on the
currently-discussed ENIAC site go pages whre you have to be an IEEE
member to go further. But I was under the impression that neither, in
their original form, were stored-program machines, and that to me would
seem to be one criterion for defining 'computer'
-tony
3:09 p.m.
I have yet to see any proper technical information on
either ENIAC or Colossus. Alas most of the technical-looking links on the
currently-discussed ENIAC site go pages whre you have to be an IEEE member to go further.
You might like this page:
http://ftp.arl.army.mil/~mike/comphist/46eniac-report/index.html
But I was under the impression that neither, in their
original form, were stored-program machines
True.
and that to me would seem to be one criterion for
defining 'computer'
One could even get into the definition of stored program. ENIAC's
program was in the form of patch cables until a later revision allowed
for stored-in-memory programs.
3:44 p.m.
I have yet to see any proper technical
information on either ENIAC
or Colossus. Alas most of the technical-looking links
on the
currently-discussed ENIAC site go pages whre you have to be an IEEE
member to go further.
You might like this page:
http://ftp.arl.army.mil/~mike/comphist/46eniac-report/index.html But I was under the impression that neither, in
their original form,
were stored-program machines
True.
and that to me would seem to be one criterion for
defining 'computer'
One could even get into the definition of stored program. ENIAC's
program was in the form of patch cables until a later revision allowed
for stored-in-memory programs.
4:16 p.m.
When was this stored-program (in the normal sense)
system added? Before or after other stored-program machines such as EDSAC Manchester Mk1,
etc?
Good question. Conventional wisdom says it happened in 1948, which is
the same year as the Manchester computer. Wikipedia (FWIW) says the
Manchester computer first operated on June 21. I don't know when in
1948 ENIAC got its own stored-program ability.
But .... here's something interesting to ruffle the Manchester fans'
feathers: http://www.columbia.edu/acis/history/acm-meeting.jpg. It's a
memo stating that, at the first meeting of the ACM in 1947, the
unfinished EDVAC computer was working and demonstrated. EDVAC was a
stored-program machine too. So if the memo's claims are true, then it's
possible this function was show pre-Manchester.
(Personally I think that argument is pointless. It's not like the
Manchester team didn't do any testing; I have to assume that the June 21
day of running its first program was not actually the first day it was
turned on and tested. Still, the memo linked above is talking about
March 1947, which is a long time before June 1948. So it just might
hold water.)
there were several projects in electronics magazines
over here in the 1970s to make a pocket calculator 'programamble'. Typically they
ahad a small RAM, the otuptus of which were used to simulate keypresses on the calcualtor.
The better ones had some kind of conditional (often dtecting the -ve sign or the position
of the decimal point in the display. My view is that the calcuaotr with this kuldge is a
stored-program comptuer, without it, it isn;'t. So going ban to ENIAC, it may well
have been modified int oa computer at some point, but this doesn't mean it was a
computer at the start.
I disagree about the need for a program to be of the stored variety for
the machine to be a computer. What's important is that the program can
be accessed at electronic speeds vs. electromechanical speeds.
To the ABC advocates, I ask, "How did your machine access its program
... oh wait, I almost forgot, it didn't have one." :)
Ergo a calculator -- and not even an all-electronic one at that.
5:30 p.m.
On 2011 Jan 23, at 1:16 PM, Evan Koblentz wrote:
The 1948 modifications were to use the function tables (hundreds of
manually-set switches) as an instruction ROM. This didn't really make
it a 'stored-program' machine in terms of what the concept meant even
back then (data and modifiable instructions in one R/W memory), it was
a half-way step. In modern terms it would probably be described as a
ROM-based Harvard-architecture machine at that point.
Some core was later added, but that of course would have to be after
core was developed (>1952). Exactly how the architecture would be
described at that point I don't know.
When was this stored-program (in the normal
sense) system added?
Before or after other stored-program machines such as EDSAC
Manchester Mk1, etc?
Good question. Conventional wisdom says it happened in 1948, which is
the same year as the Manchester computer. Wikipedia (FWIW) says the
Manchester computer first operated on June 21. I don't know when in
1948 ENIAC got its own stored-program ability. But .... here's something interesting to ruffle
the Manchester fans'
feathers: http://www.columbia.edu/acis/history/acm-meeting.jpg. It's
a memo stating that, at the first meeting of the ACM in 1947, the
unfinished EDVAC computer was working and demonstrated. EDVAC was a
stored-program machine too. So if the memo's claims are true, then
it's possible this function was show pre-Manchester.
(Personally I think that argument is pointless. It's not like the
Manchester team didn't do any testing; I have to assume that the June
21 day of running its first program was not actually the first day it
was turned on and tested. Still, the memo linked above is talking
about March 1947, which is a long time before June 1948. So it just
might hold water.)
10:56 a.m.
On 1/22/11 4:15 PM, Evan Koblentz wrote:
Having said that, here is a new site devoted to ENIAC
You neglected to mention the fact that this is John Machly's son's site.
View the contents with that in mind.
Jane Smiley will be giving a lecture at CHM this Thursday
http://www.computerhistory.org/events/index.php?id=1288639215
The new CHM "Revolution" exhibit includes several sections of ENIAC on loan
from the Smithsonian, and the reconstruction of ABC on
loan from the University
of Iowa.
4919
days inactive
4929
days old
132 comments
27 participants
participants (27)
-
aek＠bitsavers.org -
alexeyt＠freeshell.org -
arcarlini＠iee.org -
ard＠p850ug1.demon.co.uk -
cclist＠sydex.com -
cisin＠xenosoft.com -
classiccmp＠vintage-computer.com -
csquared3＠tx.rr.com -
dave.thearchivist＠gmail.com -
dkelvey＠hotmail.com -
eric＠brouhaha.com -
ethan.dicks＠gmail.com -
evan＠snarc.net -
hilpert＠cs.ubc.ca -
IanK＠vulcan.com -
james＠machineroom.info -
jws＠jwsss.com -
mcguire＠neurotica.com -
pat＠computer-refuge.org -
pete＠dunnington.plus.com -
rdawson16＠hotmail.com -
RichA＠vulcan.com -
rickb＠bensene.com -
rodsmallwood＠btconnect.com -
spectre＠floodgap.com -
terry＠webweavers.co.nz -
wdonzelli＠gmail.com