York, MCM70 & the angst of being there first

Since April, Uncle Bob has been prodding me to venture north, to York University to meet a true blue APL scholar, Zbigniew Stachniak.  So, on Tuesday, I finally did.

It turns out York has a lot of reasons to be interested in APL.  They wrote one, for starters.

Stachniak’s research includes a careful analysis of the successes and ultimate failure of MCM, a Canadian company who came out with the first personal computer in the 1970’s which happened to run APL.  As Stachniak demonstrates, being first, especially when it comes to innovation in technology, is hard and MCM went bankrupt in the 1980’s.

York houses a Computer History Museum which is shepherded by  Stachniak and Scott M. Campbell from the University of Waterloo, whom I’ve put on my radar to track down.

Stachniak is now turning his gaze towards I.P.Sharp and is particularly interested in the “APL community’s” views and attitudes towards micro computing in the 1970’s.

The more I learn about us, the less comfortable I am in generalizing about views, but that’s a puzzle for another day.

In any case, he has a boatload of photos inherited from the folks at Soliton, so I had the pleasure of seeing many of the old guard in their youth, which is, quite frankly how I remember everyone anyway.

Oh, and he’s a little reluctant to show me some of the photos.  Just when I was beginning to fear there would be no competing with the sex, drugs and rock and roll featured in The Social Network!

But seriously, Stachniak was able to interview Ken twice before he died and for this, he deserves a medal.

Thank goodness someone was thinking.


20 Responses to “York, MCM70 & the angst of being there first”

  • I had a pet IBM 5100 in the seventies. See the little white rocker switch on the front?


    We did all our work in APL because BASIC stood for Beginner’s All Purpose Symbolic Instruction Code (from Dartmouth’s freshman computer languages curriculum, it was never meant to escape the campus). We all reckoned we were better than “beginner”s. I was nineteen.


  • aprogramminglanguage aprogramminglanguage

    Andrew, You must be clairvoyant or can you actually identify the 5100 blocked Professor Stachniak’s arm?

    I’m excited because the 5100 actually works, so now I can find out what the heck is on the tape I stashed away I don’t know how many years ago.

  • I was working for MCM when they went under in 1982. Their first machine was the MCM 70. I had a single line plasma display. The next wer the MCM 800, the MCM 900 followed by the last one the MCM Power. All machines ran only APL under a proprietary OS which used disk drives as virtual memory (except for 70 and 800 which used cassettes). The 900 used 8`floppys as storage and the Power had 8″ hard disks. I spent my first 3 years of my professional carreer programing these devices. Brings back a lot of fond memories.

  • aprogramminglanguage aprogramminglanguage

    Well, Gilles, I’ll bet professor Stachniak will want to talk to you! Thanks for letting us know.

    This is getting fun, now isn’t it!

  • This reply to the IBM 5100 “APL/BASIC” switch topic comes with a worning to those with high blood pressure.

    The “APL/BASIC” switch on the IBM 5100 is symbolic. The software base of early attempts at the design of portable IBM computers, such as the Elsie (or the 1570 workstation project) in mid-1960s or SCAMP computer (1973) was APL. Then came the “switch” in the 5100 to offer both the APL and BASIC support. There was the switch on the later models as well (e.g. the 5110. It seems to me that the APL community did not see the “switch” nor was it committed to wrapping a duct tape over it to demonstrate a strong support for the introduction of APL into small systems such as the MCM computers. And, finally, in August 1981, IBM removed the switch all together — the IBM Personal Computer showed up with Microsoft BASIC only. The battle for the domination in the microcomputer software market was over — the APL forces simply didn’t show up (oh, yes, there were notable exceptions… the microAPL from the University of Waterloo for the SuperPET comes to mind but … the SuperPET also had the “switch” to execute the Waterloo’s microBASIC.)

  • I remember seeing an MCM 70 with that single-line display at IPSA Toronto–someone came by with it as a demo. I also remember that Ian Sharp reportedly thought it wasn’t worth bothering with PCs: he regarded them as just a fad, “electronic hula hoops” as I recall. The most amusing thing about that recollection is that, as I see more and more interest to recentralizing computing under the banner of “cloud computing”, I begin to wonder if perhaps Ian was _right_ about that in the long run…(of course, meanwhile our phones are becoming full-featured computers!)

  • Wow, did that picture of the 5100 bring back memories. My first job after graduating college was as the first college-trained (in computer science) employee in Chicago Title Insurance Company’s IT department. The VP had bought a couple of 5100’s, was intrigued with them and APL, and gave them to the college kid to learn APL and come up with something to do with them.
    I taught myself APL, used the 5100’s to build the U.S.’s first mortgage closing system (primarily a word processing application), and spent the next five years installing 5100’s at Chicago Title offices and agents around the U.S. With only 32K of RAM and no one to guide me, I somehow figured out a way to page/swap in additional functions when the application required them.

    I bought at least 25 5100’s (giving us what IBM claimed to be one of the larger inventories of 5100’s), and spent more time than I’d care to remember flying with those ‘portable’ computers strapped into the airline seat next to me on our way to the next installation (a 5100 was ‘portable’ like some 25-inch TV’s were portable – simply because they both had handles). I also got to leverage a bit more of APL’s power when I used it to write a program for Chicago Title to confirm the validity of legal descriptions for property: one would type in a ‘metes and bounds’ property description into the 5100, and it would confirm whether the description properly described the property: that lines didn’t cross, and that the final leg of the description returned properly to the initial starting point for the description.

    To Roland’s note: Ian also referred to PC’s and those who used them as ‘data strippers’: a threat to our database business even as he considered PC’s a fad. And I will admit now to tweaking some of my current colleagues: if we’re going to refer to cloud-computing (and its evolution from the ‘thin client’), why not refer to it by its original, proper name: timesharing.

    Thanks for the memories….

  • Gary, I have read your 5100 recollections with pleasure (taking 5100s on a flight story is exquisite; more on “APL in space” below).
    Returning to Ian and his stand on microcomputing: so, in 1975, he didn’t want to solder an Altair 8800 kit to be able to add, with some creative toggle switch flipping, 2 to 5. But then, in the second half of the 1970s, there were micros quite able to run APL and access IPSANET. Was he (and those around him) still persisting that the micros could not be profitably integrated into the IPSANET structure? I’m ready to pay in advance for any answer to this question with the following APL anecdote.
    In August 1973, MCM sent its employee Ted Edwards to APL Congress in Copenhagen with a prototype of the MCM/70. What was unusual about that MCM/70 was that it was mounted in an attache case and was operating on batteries. According to one of my sources, Edwards was not only able to board the plain with this unusual device but also reviewed his presentation using that MCM/70 during the flight to Copenhagen. This constitutes another “first” for MCM: the first portable computer operated during a flight. And, of course, that “laptop” was running MCM/APL (a preliminary version, I guess). The MCM/70 story was picked up by the Danish daily Politiken and if you read Danish, I can e-mail you a copy. Does any of you remember the MCM/70 presentation in Copenhagen?

  • 0. I would like your permission to use your story about the MCM/70 on the plane in a collection of APL anecdotes. http://www.jsoftware.com/papers/APLQA.htm

    1. Please forward the MCM/70 story in the Danish daily. http://translate.google.com sometimes does a passable job, and as well I have access to colleagues fluent in Danish.

  • Zbigniew,

    My frame of reference (my years at IPSA and subsequent absorption into Reuters) is 1980 – 1990. I can’t speak to Ian’s (and the company’s) view at that time on the feasibility and potential profitability of integrating micros into IPSANET. But from the perspective of the database business: we charged by the data point, and added value (and revenue) through the use of APL applications to process the datapoints we sold to clients. If the client needed the datapoint again in the future, they were charged again to retrieve it from us.
    Micros upended that revenue model. If a micro could retrieve (and store!) that data item (which others at IPSA derisively referred to as data stripping), they would never again need to retrieve that same data item from IPSA’s databases and – worse – could conceivably create their own database of that data, warehousing it for future use by others in the client’s organization.

    In its defense, even several years later, at the time IPSA was acquired, Reuters and other data vendors had not yet figured out how to adjust their revenue models to the brave new world of micros. They refused to simply plug in a data pipe to funnel data to a client firm to distribute as it saw fit to its users, instead clinging to the idea that it could continue to license data terminals (in physical locations) to each individual user in a client firm; each user could then look up discrete data items on Reuter pages, for which Reuter clients needed to memorize the page codes.

    How far we’ve come….

  • The most immediate source for memories of Ian’s take on PCs generally, or APL micros in particular, would probably be…Ian Sharp! He’s on LinkedIn, though I gather he doesn’t like this system as well as Facebook. Interested chroniclers could try to contact him there.

    Nonetheless I can’t resist adding another small memory. The Sharp APL/PC implementation, which is still floating around the Net (and anyone can use if you can find a PC old enough to run it) is, as I’m sure everyone here knows, the mainframe interpreter running on a System/370 emulator. I’m not certain whether it’s also general knowledge, though, that that was a strategic move. IPSA contacts with IBM researchers were aware that IBM had developed a 370 on a board that could run in a PC. It seemed clear that IBM would of course leverage their large investment in, and knowledge of, system/370 development by
    releasing that hardware widely, perhaps even making it the standard PC. So Sharp APL/PC was positioned to take advantage of that anticipated development. Alas, IBM took
    precisely the opposite approach–perhaps they were afraid that, if you could buy a PC compatible with a 370, you wouldn’t need a 370…

    Incidentally, I once gave myself the pleasure of running Sharp APL/PC with its 370 emulator running in SoftPC on a Macintosh. OK, it wasn’t terribly fast, but everything worked! I don’t think I ever got around to trying it on the Macintosh emulator running on my Atari ST, unfortunately.

    By the way, I too have a stack of old IPSA newsletters (and for that matter APL Quote-Quads) gathering dust. I would be happy to contribute most of those to some chronicler or other, if they are of any use.

  • aprogramminglanguage aprogramminglanguage

    Nice stories, folks. Thank you very much for your contributions,continued loyalty and readership.

  • Roland,
    Do you still have a working Ampere computer? It came along after the IBM 5100 and MCM 70, but before the IBM 5150 (the PC). Do you know if the Ampere developers looked at the MCM 70 as a good idea? Curtis

  • Curtis,
    Sorry, not only do I not “still” have a working Ampere, I don’t think I ever did. I referred above to my Atari ST,
    which I kept for a ridiculously long time but eventually
    turned in for recycling…and which did not have a native
    APL at all, so far as I know, though probably some of the
    M68K-oriented interpreters could have been ported easily.

  • Roland,
    Sorry. Wrong Roland. It’s Roland Besserer who mentioned having a couple of Ampere’s in comp.lang.apl.

  • Oh Cool! Thanks, Curtis, for connecting us to the comp.lang.apl discussion!

    I hope you’re enjoying The Computer History Museum’s Revolution!


  • Catherine, I don’t know what hole I’ve fallen into this winter but I’m only now reading some of your blog. With Respect To Ian and microcomputers, I recall being puzzled by his choice of relying on IBM’s direction of combining 8008 and S/370 architectures, a true “disinnovation” move.

    He never really liked me but he did answer my question.

    “You’ve gotta trust somebody”


  • Inventing the PC

    The MCM/70 Story

    Zbigniew Stachniak

    A uniquely Canadian story of the company that promised a new era in computing.


  • aprogramminglanguage aprogramminglanguage

    Canadian businesses face unique challenges even today, given the intense competition with our massive Southern neighbour and our attraction for everything not made at home.

  • aprogramminglanguage aprogramminglanguage

    Interesting forum notes post publication of Prof. Stachniak’s book: “Inventing the PC – The MCM/70 Story”


Comments are currently closed.

Analytics Plugin created by Jake Ruston's Wordpress Plugins - Powered by Laptop Cases and r4.