Mural (and poster) showing mini, personal, and home computers

[voidstar]

I was debating this topic with my daughter (age 10), and I found that there wasn’t quite a single concise picture on the “transition” of minis/personal/home computers. So, we made one!

A Mural Dedicated to the History of Personal to Home Computers — voidstar

Years ago, I was walking thru the Vatican Museums, and they have this long hall of tapestries - centuries old, and each one told a story. So, I thought “what if a tapestry on the origins of the personal computer were made? what would it look like?” I picked the systems, but my daughter did the arrangement and background. There were numerous other systems I wanted to include, since there is so much more to the story. But to anyone who thinks the PC started in 1981 with the 5150 – well, it’s not that clear cut.

I used to think the “PET” as first (because it was the first “complete system” under $1000). Now the DP2200 from 1970 is a good contender - it’s a “sleeper” because few realize it was programmable. Another “factoid” forgotten is that in 1975 IBM was under some antitrust legal issues (Sherman Act stuff) on bundling capability - the 5100 shows they were perfectly capable of making a personal computer, but they had to be legally careful on how they went about it (their charges ended up being dropped around 1982). The legal history is neat - Datapoint was wrecked since their profit was used to prop up other companies, rather than investing in R&D (edit: a direction chosen after a new owner acquired them; one of the lead founders had died in a car accident); (edited) Jack Tramiel (Chuck Peddle was still an engineer at Commodore) enacted revenge in TI and wrecked TI’s entry into PCs as payback to how TI strong-armed the calculator industry in the early 1970s; or how IMASI bought and ruined the Altair because they were selfish and didn’t want other products being sold next to theirs in the newly emerging computer stores.

“officially” the KINBAK-1 is deemed the first personal computer. And on technical grounds, that’s not wrong – it was under $1000, and could “compute stuff.” But 256-byes isn’t very limited, think only 50 were sold.

The Altair 8800 did have that S-100 bus (the earlier Micral N also had a bus – the Micral could be “expanded”/“upgraded” into a more “complete” system, just like the Altair). But I came across the price sheet of the Altair 8800 - adding the keyboard, over 4K memory, tape system, and the early BASIC, and an async IO card (serial video cards were a thing for awhile) – and you’re quickly at near $5000. At disk drives, and you’re quickly up to $7000 - and it’s not like the stuff you ordered arrived in a week or even a month, there was substantial backorder on all components (most especially memory cards).

Then come the Trinity 1977 - complete systems, closer to $1000 and built-in ROM with enough capability to start doing stuff. But, not very extensive bus systems. (yes, yes, the Apple2 did have a multislot bus - but it had some issues that resulted in some cards having to be in specific slots, so it had some limitations; the PET had IEEE-488 expansion). I’ve seen the 1977-1980 systems best described as “trainer systems” - 64KB, could load BASIC. You could get a taste of things, but they were essentially programmable calculators with CRTs.
EDIT: That is to say, for all design faults, the IBM PC ISA bus did enable that system to remain useful (via expansions) for about 10 years, something the earlier “trainer systems” couldn’t do (being 32/48/64K limited, etc).

I looked at the sales numbers - each of the Trinity (including the Apple2) are only about 50,000 units sold, maybe 100,000 tops. (the early DP2200, Wang, and HP systems were sales around 3000, or certainly under 10k; the IBM 5100/5110 combined were roughly around 30K – these are estimates, but the point is, they were each barely 1-football stadium worth of customers – worldwide).
NOTE: to clarify - the Apple II initially wasn’t that popular or successful, it wasn’t until the later Apple IIe that the appeal started to take off.

So with all that perspective - maybe the 5150 PC “was” the first PC. Except, in 1981 it still had some wrinkles to be ironed out (like the weird 160KB disk format) – which they did, but not until 1983. Both the Apple IIe (enhanced/extended) and IBM PC quickly moved well beyond 100k units. By 1983, the 5150 was over 1m sold, and the Apple IIe soon followed. So that’s another subjective line: first to reach 10k sold? 100k sold? 1m sold? That why I now also agree the Trinity systems were “kind of a first”, but were still a kind of “learner/trainer” for everyone (both the system designers and the retailers).

EDIT: HOWEVER - all the hallmarks of being a PC were there in the 1972 DP II, 1973 Wang, 1975 IBM 5100 - as far as having an instruction set, CRT, storage system, and full size keyboard. Write letters, store some cooking of ancestry info, play (simple) games. The difference was cost, and that didn’t get to below $1000 until 1977 (the Altair kit doesn’t count – the Sol-20 represents an Altair kitted out as a PC, and that was about $2500). But if networking and high resolution graphics are your threshold, then we’re talking not until the 386 and around 1988 (PS/2 era).

The best quote I’ve come across is from Project Xanadu, something like “the only reason Microsoft exist is because DEC charged $50k for Unix”. And that’s a very interesting perspective - the vision of the 1970s was the “citizen programmer”, users interacting with the system in a way that suited their needs, not pre-packaged other-peoples-design forced upon all. Except, some kind of workflow consistency was essential to mass adoption.

Anyway, I hope this “mural” can help others to at least start a conversation on how we got to the 1980s PCs.

[jecel]

That is a great poster! It would be interesting to have that in computing museums.

Did you confuse Chuck Peddle and Jack Tramiel in your comment about the fight with TI?

It is not possible to include all interesting machines, but the TI99/4A had a key role in the home computing market in the US with its fight to the death with the Commodore VIC20 and C64.

Since you included the BBC micro, its great rival the Sinclair ZX Spectrum might deserve a mention. In fact, the ZX81 had an effect in the US market that is often forgotten today given its poor direct sales there and the odd Timex partnership. But both Commodore and Texas Instruments spent a lot of money developing “ZX81 killers”: C16 and TI99/2.

[voidstar]

My daughter and I have decided to release this to the public domain.

voidstar78/mural1970: Mural dedicated to domesticating the computer (1970s) (github.com)

We have a variety of formats suitable for large poster sized printing. Indeed, if a museum does want to do a print, we’re trying to make it easy to do so.

Happy Holidays 2022

Github currently has a single file limitation of 25MB, so at the github description is a link to her original KRITA file at another file share location. The KRITA format has the original layers she used, which may make it easier for others to make custom versions.

The “discussion link” includes the rationale of what was included (and some rationale on why certain things were not included).

For reference, here is the small preview:

[voidstar]

Yes Jack Tramiel is more correct. The price war enacted by the C64 was brutal to TI

[voidstar]

@jecel

It is not possible to include all interesting machines, but the TI99/4A had a key role in the home computing market in the US with its fight to the death with the Commodore VIC20 and C64.

Since you included the BBC micro, its great rival the Sinclair ZX Spectrum might deserve a mention. In fact, the ZX81 had an effect in the US market that is often forgotten today given its poor direct sales there and the odd Timex partnership. But both Commodore and Texas Instruments spent a lot of money developing “ZX81 killers”: C16 and TI99/2.

Thanks for these suggestions. I’ve mulled these over some more…

The TI-99/A could go right where the Graphic Tablet currently is - putting it nicely on its own line, as a unique series of TI-based processors (TMS9900). I’d make the Graphic Tablet a bit smaller and have the TI-99/A next to it, as close within the 1979 column as possible. [ including the Graphic Tablet was important, to emphasize the idea of added-peripherals and the concept of extending these systems via the bus to have capabilities beyond what was originally imagined – the asyc. card and modem being the first example of that capability ]

I remember using a ZX Spectrum at some point (that black case and small keyboard). And agreed, the ZX81 probably deserves an entry. The ZX Spectrum could replace the ColecoVision (the Coleco was sold in the millions initially, but didn’t last till the 1990s like the Sinclair’s). But in trying to stick with “first releases”, I’d go with the ZX80 and have it replace the TRS-80 Model 3 (largely because the Model 3 wasn’t widely sold as a home computer, it was far more practical for Radio Shack stores to offer the Color Computer rather than the bulky Model 3’s). Even though the ZX80 had some NEC-based CPU (and wasn’t particularly popular), it ultimately led to the Z80-based ZX81.

One reason I did want to include the Model 3 was because it showed the inclusion of 5.25" disk drives didn’t start with the IBM PC and DOS. That’s not to say the Model 3 pioneered the 5.25" format (that was really due to one of the 1975 Wang system updates), but the Model 3 was one of the first micros to have them “built into” the case, and not a large external box (and the idea of something “more than boot up BASIC” was needed to organize files and processes, was several years in the making before MS-DOS – DEC wanted $20k for Unix, and Kildall wanted a fair price for CP/M - but Microsoft’s little “QDOS” cost-cut all of that).

I’ve added the TI-99/4 and ZX-series as Honorable Mentions

Another option is to remove the “software row” (VisiCalc etc) and place those systems around there (but I think it’s important to show how it was “software titles” that ultimately helped make microcomputers successful, as a whole new form of media arose {box software on shelves}, in both productivity and entertainment software {I wanted to have one edutainment title, but I couldn’t find any boxed titles pre-1985 - like Carmen San Diego was 1985; there are some non-boxed titles like typing tutors and such}).

[jecel]

In Brazil a clone of the TRS-80 Model III called CP500 was the most popular TRS-80 of all. Before the arrival of the MSX machines in 1986 it was the only real rival to the Apple II clones. There were the Sinclair clones below, but these were toys. And there were the PC clones above, but they were too expensive for the home.

[voidstar]

We finally got time to make a Version2 of the mural my daughter and I prepared a few months back.

The changes are:

• replaced TRS-80 Model 3 with Sinclair ZX80. I liked the TRS-80 Model 3 because it represented one of the earliest systems with “integrated 5.25 disk drives”. But, I also wanted to avoid having “follow on” models of systems (e.g. it’s why there is no PET 40XX or Apple IIe or IBM 5110 etc). Which is also why I wanted the ZX80 (“first run”) instead of the more popular ZX81 (and that the ZX80 fits nicely in the 1980 column).

• Added TI-99/4, squeezed in the open space between the graphic tablet and IBM 5150. Again, while the TI-99/4 itself wasn’t particular popular, it was the “first 16-bit micro” and it was a decent system (at least the follow on TI-99/4A was) - TI did at least field and support the product, just that competition ultimately got the best of them.
  NOTE: with the expansion modules, the TI-99/4 now shows that concept of disk drives becoming more “mainstream” parts of the system (prior to the IBM PC release in '81).

• Tried to fit the TMS9900 CPU that the TI-99/4 was based on at the bottom. It looks like it basically mid-1976 just like the Z80, and one thing to notice is how it had many more pins than other contemporary processors. It’s a tight squeeze, should I just remove the TMS9900 processor from the bottom? NOTE: and I realize this gives some confusion. Generally there is some lag between when a processor is developed and when it appears in a fielded product - so it’s not that those processors were used in the year-column of its release, but that’s when it was first available. If one every made an “interactive” version of this image, I suppose mouse-over of the CPU would halo or highlight the corresponding systems that used them. [ interesting-- there is no image format that supports that kind of “interactive expression”, we have animated GIFs but not “interactive images”, we have do a lot of web-stuff to make something like that happen ]

• Added the only “screen shot” of Electric Pencil that I could find, above the SOL-20. With the limited resources back then (e.g. 4K RAM), programs then rarely had much of “intro splash screen.” Technically the “electric pencil” icon was more from 1981, but the software was started around 1976. Aside from the SOL-20, it seems Electric Pencil was “first” used on a TRS-80 Model 1. (so I thought about putting the green pencil icon closer to the TRS-80, above the Microchess to blend those “green things” closer together – thoughts on that?)

• Added “EASY WRITER WORD PROCESSOR” next to the Apple II. I couldn’t find evidence that Electric Pencil had been ported to the Apple2. And on the PET line, Commodore had their own suite of producivity software for the 6502. So in the Apple domain, I do think EasyWriter was effectively that systems first word processing tool. [ thoughts on that? is “EASY WRITER” too distracting there? it definately establishes that word processor was the producivity tool killer-app before VisiCalc ]

• some changes at the bottom for WRCBBS to just CBBS. WRC was “Ward Christensen” and I thought I came across a computer club newsletter that referred to it as such, but it seems to have been more commonly just called “Computerized BBS.” Added image of the IMASI that Ward used to write and run that BBS (which tugs at the idea that even then, people (who knew what they were doing) could “build their own PC” somewhat like we do today – but the “home computer” concept appealed more to consumers that could rely on pre-packaged known-good configurations).

• also moved 5.25" disk image from '78 to '76. It was at '78 because that’s when drives and the media started to become affordable and not-unusual to be available to all the main microcomputer systems. But '76 is technically when the form factor was developed (initially as part of some early expensive Wang systems).

We can’t make major system additions here - I’d still love to add the Mark-8 somehow. But mostly wanted to make sure we’re not grossly misrepresenting anything?

[oldben]

HOW about a small clip of a typical AD atound 1984, showing the mail order parts of clone
and home brew computers kits. I picked up a z80 s100 bus kit, but my power supply shorted killing the whole kit.

[voidstar]

Well, I did end up doing something that involved some old ads from 1984. BYTE has some great stuff archived, even to 75/76.

THE FOLLOWING IS A PREVIEW video (unlisted), that I’ll probably remove before the end of the week. No audio yet.

This is just a preview to show a concept of some ideas we’re kicking around - but would appreciate to message me any thoughts about it.

Again, this is unlisted, it’s still a work in progress:

Domesticating the Computer (1080 @ 30fps) [ tentative preview ] - YouTube

[oldben]

A minor error is Intel was making other chips than just the processors, in the early years.
MOS LSI was the big thing then, 50x more gates than TTL .https://www.cpushack.com/wp-content/uploads/2018/06/VintageIntelMicrochipsRev4.pdf Even the Datapoint used intel memory.

[voidstar]

Quite true, the 1101 and the 1702 (EEPROM) were game changers.

The CPU is important, but it takes a fleet of diverse chips (and components) to make a system. I think including the “single board computer” aspect speaks to some of that – early computers (say pre-1970) were so hot and error prone just due to having so much hardware (possible points of failure). Finding that balance in composition of chips (i.e. this single chip can now do what these 10 chips used to do) was key to a simpler-to-build system (that could then be put on an assembly line and stamped out in mass).

I found this nice list of Intel chips:
Time Line 1969-1979 - Intel Vintage (intel-vintage.info)

[jecel]

The poster is great and the animation is even more informative. One problem I don’t have a solution to is how to present prices. The best option is to show the values at the time they were current, as you did. But it is how for people now to have a good idea of what $750 in 1973 really meant. But showing adjusted values would add too much confusion and would also become obsolete since someone watching this 15 years from now would not have a good idea of what values adjusted for 2023 meant.

The name of the Kenbak-1 is spelled with an “i” in the poster.

One interesting detail about the TMS9900 was that it was implemented in a unique bipolar technology called I²L instead of MOS like all the others. I think its successors, like the TMS9995, were implemented in MOS.

[jdreesen]

The well know Swiss collector Robert Weiss has been selling his computer poster for years now. It is gigantic at 0.9 x 2.5 meters …
Check it out here : Poster digitale E

[oldben]

Sigh… You can’t by the old computers any more, just the poster.

The one thing all this information does not show, is when the newest parts became common for the Average Joe, Woz or Bear. My current cpu design,uses 2901A’s
22v10’s and undefined Proms at 100 ns, with a DE1 FPGA as the test platform.
A new hexadecimal front panel is planned for the system, after I get my software
designed, over the led and switch pannel I have now. As a retro design, I have no
idea when to have the release date of the fictional computer. I can’t design software
untill I have a idea of ram I have. 32Kb for the user and 32Kb for the OS looks ample.
I am usng a cross assembler In C so I have no idea of how big the native assembler will be.16K data? 16K code?

• 1969 core memory & state decoding and SSI TTL. 64Kb max memory
• 1971 core memory & 74181’s and 256x4 proms.
• I972 surplus chips , core , seven segment displays. New core 96Kb max
• 1975 4K mos ram,LS TTL. 128Kb max
• 1976 2901’s, 82s100’s,MK4096 drams,
• 1983 2901A’s 22v10’s 16K static ram

.

[voidstar]

I tried ordering that a couple months ago, and there was some issue going through at that time. I’ll try again later in the week.

The Swiss one is a great poster, but it overwhelmed my daughter - so she put together this arrangement as a more concise “intro” to the topic (it’s her signature on the bottom left of ours). Thanks for pointing out the Kenbak-1 spelling!

So far I can’t convince my daughter to do the audio narration on this We’ve thought about maybe “hiring”/renting one of those AI ones. Still mulling it over, but we’re still examining the timing of things - goal was to keep initial video to 10 minutes (not a 2 hour documentary).

EDIT: another idea was maybe doing multiple videos, ~10min on each main section. But we just get weekends, so that could take all year to finish.

[voidstar]

I’m thinking about just removing the prices altogether - and putting them in a summary at the end maybe? Or just link it as another reference (webpage).

Because aside from the issue of “initial brochure” prices, the other issue is amount of memory and exact configuration – for most of the systems, just a few accessories can more than double the initial price. And removing price sort of de-clutters the presentation and discussion.

Like the initial IBM PC, $1500 just got you a tape cassette interface and 16KB. To really make use of the system in any meaningful way, one would really need to spend closer to $4000. Same story for the Altair - $400 kit was basically like taking an electronics self-paced class, adding anything and the kit was quickly over $2000 (like 4KB memory, a serial card, drive controller, etc) – of course, some people were savvy enough or connected enough to find suitable accessory parts themselves (if could write the code to integrate whatever-it-was).

Another thing I like to do is show the average cost of a car and a house during those years, that helps me gauge the relative cost of the system. e.g. a $20k system in 1970 was basically like buying a small house (of course that’s regionally subjective, especially across continents).

[oldben]

For the people across the pond, a CASTLE might be better than a house.
Perhaps it might be better for 3 parts. Mini computers, personal computers (word processing/spread sheats/acounting),home computers (gaming/ toy - trainer).

[voidstar]

Take 2 version (still unlisted, narration still TBD).

Domesticating the Computer (take2, beta 1080 @ 30fps) - YouTube

Some cleanup and refinements throughout, now extended to ~12min.

Hopefully typo corrected. Also replaced Tandy 1000 with actual 1984 vintage (single black disk drive).

The main theme is how these computers crawled into our homes. Like cats and dogs, it was a process of domestication over time decades Maybe more like horses? It’s why the Odyssey is there - it was the first step into people’s homes.

My daughter has a couple easter eggs placed in the video. Think one will be found, the other probably never

A few touchups left still (and high quality final render), but think the overall flow is ok - but will take some time to figure out the narration.

[drogon]

What would also be nice is chapters with dates.

Thanks,

-Gordon

[voidstar]

In the final, YT lets us put timed chapters in the video description - so that will help when we get to that point (still a draft so far).

But also - since we lose the row or years at the bottom while zoomed and panned, maybe we could add a “year marker” in one of the corners (that’s always “current” to what the discussion is). Will give it a try and see how it looks.