IBM, sonic delay lines, and the history of the 80×24 display (2019) | Hacker News Reader

IBM, sonic delay lines, and the history of the 80×24 display (2019)

righto.com

97 pointsrbanffy45 comments

Tangentially related: is there a history covering IBM's development of microcomputers? It is clear that the traditional story of the development of the IBM PC leaves out many important details. There the 5100/5110/5120, which goes back to the mid-1970's and reflects the stereotype of IBM. There is also the System/23 DataMaster, where the hardware seems to be the basis of the IBM PC. This seems to go against the traditional story that the IBM PC was some sort of renegade project. (If anything, they appear to be companion projects. The main difference being the DataMaster's focus upon IBM firmware/software.)

Like I need another big project :-)

The IBM Datamaster is an interesting system, but it was doomed. It had an 8-bit Intel 8085 processor, cost $9000, and came out in July 1981. The IBM PC had a 16-bit 8088 processor, cost $1565, and came out a month later. So there was no reason to buy a Datamaster

There's a good description of Datamaster in "A Personal History of the IBM PC" by Dave Bradley (one of the PC's designers). Unfortunately, it's paywalled.[1]

[1] https://doi.ieeecomputersociety.org/10.1109/MC.2011.163

That sounds like a great starting point. There's a university library 500 meters from where I work, so bypassing the paywall ought to be easy.

And thank you for all of your articles over the years. They border so close to applied physics they are fascinating reads.

https://annas-archive.gl/scidb/10.1109/mc.2011.163/

> Like I need another big project :-)

You know you want it.

> So there was no reason to buy a Datamaster

It did use a screen font closer to the 3270 series. That is a big pro ;-)

IBM also was studying home computers, and was talking to Atari. There are other gorgeous design studies in “ Delete - a Design History of Computer Vaporware”, a lovely book. I want to 3D print half of those designs.

No idea if this was a factor, but 80x25 on the IBM PC allows for showing 80x24 plus that extra line of function key labels:

https://en.wikipedia.org/wiki/IBM_BASIC#/media/File%3AIBM_Ca... (IBM BASIC screenshot)

Imagine when edit.com came out and QBASIC used it for the editor. You lost two more lines of valuable code space!

https://en.wikipedia.org/wiki/MS-DOS_Editor#/media/File%3AMS...

You know, this is funny because QBasic did not use EDIT.COM. Instead, QBasic was the editor and EDIT.COM was a simple program that called "QBASIC /EDIT" :-)

It was basically the same thing. That's my point.

I recently went back to my 1993 Turbo Pascal code (mostly 2D VGA and Sound Blaster game engine experiments) on period correct hardware.

I was surprised by how claustrophobic it felt to only see 21 lines of code in e.g. Turbo Pascal 7.0. Still didn’t like the squashed 80x43 mode.

https://winworldpc.com/screenshot/c38a28c3-84c3-ba28-1011-c3...

Then I remembered how larger displays and xterm felt like such a liberation a few years later.

Power users had superEGA with 132 column 40/43/44/60/66 row modes.

I think mostly just Lotus 1-2-3 (the standard textmode spreadsheet application of the time) supported that?

Shorter displays end up making programs more modular with shorter functions.

I always encourage my fellow Python developers to stick to 80 columns for readability.

The PARC crowd thought displays should have the form factor of a sheet of paper. Hence the Alto display.[1] That never caught on.

[1] https://www.righto.com/2018/01/xerox-alto-zero-day-cracking-...

I see people doing that today.

They are correct of course.

I always assumed the reason the PC had an 80x25 display was to allow for emulating the 3270 (80x24 for content and then an extra line at the bottom for the status information) so it's interesting to see that wasn't the case.

> Also, 3270 emulation was not a requirement or goal for the IBM PC, and didn't lead to the 25th line. I would much prefer to have a tidy story where the 3270 led to the PC's display, but unfortunately that's not the case. I talked to two of the original IBM PC engineers to check on this. The said the IBM PC team felt absolutely no need to be compatible with other IBM products. In particular, several features of the PC made 3270 compatibility harder: the use of ASCII instead of EBCDIC, little-endian words, and 10 function keys instead of 12.

Man. I love the design of old terminals, computers, and such.

I am, also, extremely glad that these form factors were abandoned. Having an old terminal, it is possibly the least ergonomic machine I have ever used.

Same. And you may like this one:

https://design.tel/olivetti-tvc-250/

I’d make it my daily driver ergonomics be damned.

Though I’d probably go with an IBM 3278 or 3290 for practicality.

From a linked article on shift registers:

> To avoid these astronomical prices, some computers used the cheaper alternative of shift register memory.

Might be a direction for 2026 too?

One theory I saw argued the punch card size was the reason for 80x24. But why were punch cards that size? They were designed off of the cards used for the census. Why were the census cards that size? Because they were modeled after the dollar bill size.

I do love thought experiments like this but do believe they’re insatiably unresolvable.

In the end, all reasons resolve to either "it's what we had at the time" or "someone thought it looked good."

Not always, for example original CD disks had capacity of 74 minutes to accommodate Beethoven's Ninth Symphony.

That one also turned out a myth :) CD size was determined by Cassette tape dimensions (diagonal, human can still hold one in one hand) and that combined with conservative pits/lands/track pitch choice drove the play time.

thus CD runtime was derived from something "what we had at the time".

The story as told may be inaccurate but it wasn’t simply ‘what we had at the time’ either.

The 74 minute length resulted from Sony rejecting the Philips 60 minute 11.5cm diameter “Pinkeltje” disc size in favor of a 12cm diameter.

It’s quite possible that Sony’s Norio Ohga simply argued that the 9th symphony or various operas fitting would be enough of an advantage for the slight size increase without meaningfully decreasing portability.

Meanwhile the LP crowd was flipping sides like it was Ultima VIII (slight exaggeration). Why would it be critical for a new format to do away with multi-disc releases if the customer base has already grown accustomed to them?

Because it was annoying to flip the disk in the middle of the thing.

You got any source for this?

The symphony story might be a legend, but it's pretty well known that the original design was somewhere in 10-11cm range, but this was eventually increased to 12 cm.

The "diagonal cassette size" seems extremely far-fetched - first, who cares about this? If you are worried about boxes, shelves etc.. you want horizontal size, which 10 cm. And you are worried about holding in hand, 12cm is not very convenient for the smaller hands, a smaller size would be better.

When respectfully handling them out of the box, I always stuck my index finger in the central hole and the thumb on the border. I have large hands but I rarely held them by the borders.

If anything, I'd have guessed that the size of an 8-track in a car stereo was probably a larger influence on potential form factors for CD audio. Since car stereos were at least somewhat normalized in the early 80's. Not speaking to an adapter, just in terms of what would "fit" in a typical stereo hole in a car.

That said, I doubt that's the reality either... it's probably a number of factors. I am slightly surprised that a USB based read-only media format standard for players hasn't materialized, though it seems that online/rental models are what the industry really wants.

Source is Dr. Kees A. Schouhamer Immink. Philips’ principal engineer in the joint efforts of Sony and Philips to develop the Compact Disc (CD) https://www.turing-machines.com/pdf/cdstory.htm:

>The disk diameter is a very basic parameter, because it relates to playing time. All parameters then have to be traded off to optimise playing time and reliability. The decision was made by the top brass of Philips. 'Compact Cassette was a great success', they said, 'we don't think CD should be much larger'.

>As it was, we made CD 0.5 cm larger yielding 12 cm. (There were all sorts of stories about it having something to do with the length of Beethoven's 9th Symphony and so on, but you should not believe them.)

> The 8-14 bit channel code was agreed and all the specifications between them led to a playing time of 75 minutes.

"Everybody just liked it that way and it costs too much to change it now": https://www.exocomics.com/743/

And the reason they were modeled after the dollar bill size is because there were already many types of systems for storing and organizing them. That came in handy for the census.

The old BBC Connections series has a segment with James Burke using the old census tabulators.

https://www.youtube.com/watch?v=z6yL0_sDnX0&t=2640s

> But why were punch cards that size?

Something related to the width of a two-horse Roman carriage… Not sure ;-)

Fascinating article, I really like knowing where the old standards came from.

But I am extremely curious the first picture in the "The IBM 2260 video display terminal" section. All the other pictures show the typical extremely round CRT of the era, but that one is the characteristic cylindrical tube of Trinitrons, a technology released several years later. I am trying to find some information about it to no avail.

The manual for the IBM 2260 describes the CRT in detail but I don't think it has the information you want. My guess is that if you're IBM, you can get the CRT in whatever shape you want.

[1] https://www.bitsavers.org/pdf/ibm/2260/Y27-2046-3_2260_2848_...

Wow, that is a very detailed manual.

I am trying to find more information to no avail. Actually the only picture of an IBM terminal with a cylindrical tube is the one in your article, there is nothing else on the entire Internet. I will keep investigating.

Oh, I see the difference now. In the photo of the IBM 2260 terminal that I used, the bezel is rectangular and flat. In every other photo of the 226, the bezel is sunken and the screen is oval-shaped. I'll ask around and see if I can find out why. Maybe the photo that I used is a later version with a better CRT?

I wonder what the knob on the side was for. Vertical position seems absurd.

That’s also a very large screen compared to how much text it’s displaying. The 2260 was a very odd contraption.

The keyboard was particularly insane.

No mentioned are the PC terminal apps from BBS era dialup connections that tended to use one of the 80x25 lines as a status bar, leaving 24 lines for the remote display... not to mention avoiding or embracing wrap issues that sometimes left you from using column 80 reliably either.

It's kind of nice to see a lot of resurgence in terminal technologies and remote terminal apps these days... I think AI is letting a lot of people that have been nostalgic pick up projects that would have been too daunting otherwise. Myself included.

Deeply fascinated by these historical threads. It is precisely the various design choices made throughout history that have shaped the computer systems we use today.

The linage can be traced back to Basile Bouchon's paper tape invention in 1725. The article doesn't mention the role of punched cards in The Holocaust, though, which my blog post goes into:

https://dave.autonoma.ca/blog/2019/06/06/web-of-knowledge/

Reply on news.ycombinator.com