Slashdot Mirror

← Back to Stories (view on slashdot.org)

Updating the Computer, Circa 1969

Posted by ryuzaki0 on from the first-move-the-larger-stone-doorstop dept.

Coudal points out a "Swell article from UK Magazine 'Design' from 1969," excerpting "Designing a computer is a continuous process in which technological breakthroughs must be matched by new hardware, and new hardware by new software, without invalidating the systems already in use."

9 of 124 comments (clear)

  1. Conversational Computing by (1+-sqrt(5))*(2**-1) · · Score: 5, Interesting
    From TFA:
    The 1903A [...] can handle conversational computing on nine remote consoles.
    “Conversational computing” is a fantastic euphemism for command-line-interaction; more sophisticated, in any case, than the point-and-grunt interface of today's hoi polloi.

    My theory is that computing and humanity interrelate: in an environment where Latin is taught alongside math, your users and developers are sharper and more humane.

    1. Re:Conversational Computing by Tri0de · · Score: 4, Interesting

      Respectfully, I think Latin is a very effective language for all kinds of communication. Many Latin words have more universal meaning then English, French or Russian phrases. "Ad hoc" is a great example of a concise term with a precise meaning that needs no translation.
      There is an advantage to a dead language; unlike English meanings aren't mutatingl thus you can often get a better feel for exactly what someone from a different era and different culture was saying without the problem of words such as "gay", "Stupid" or "sick" having very different meanings a few generations later.
      'Good' language is that which communicates what and as you intend, be it technical jargon, slang, Oxford English, Spanglish; if the sender and receptor send and receive the right message it's good language; withing that context Latin can, and often is, highly effective.

      --
      "Everyone is entitled to their own opinion, but not their own facts."
    2. Re:Conversational Computing by Anonymous Coward · · Score: 3, Interesting

      Actually, that's debatable. English draws much of its vocabulary from French (thanks to the Norman Invasion of England in 1066), while it gets a lot of its grammar and structure from its Germanic roots. English is probably more accurately considered a Romance-influenced Germanic tongue. Interestingly, the language we know as English is really a creation of the linguistic forces that existed in England a thousand or more years ago. A little Anglo, and little Saxon, a little French, and a good measure of the local Celtic tongues to give it some flavor. How many other languages can you think of that have three words to describe one concept, each derived from a different source (anger, ire, and wrath, for those of you who are wondering)?

  2. Sales department puffery by wbean · · Score: 5, Interesting

    That article is a typical pice of sales department puffery. If you really want to know what it was like to design a computer in those days read Tracy Kidder's Soul of a New Machine. It chronicles the efforts by Data General engineers to create a new computer. At the time I was working as an engineer for Honeywell's EDP (Electronic Data Processing) division and I can vouch for the accuracy of Kidder's reporting. I recognized all the problems and all the actors even though it was a different company.

    At a given point in the development of computers a lot of people end up working on the same problems and often come up with similar solutions. While I was at Honeywell they bought GE's computer division and we got to see the design documents for GE's new computer. It was very interesting reading since we could look at each turning point in the design and say: "Oh, they decided to do it that way." All of the problems were ones that we'd worked on and the solutions were all ones that we'd considered. For the most part they'd made the same decisions we had. It was an experience that's given me a real respect for the notion that an invention is "in the air." It isn't necessarily because the problems are being widely discussed but more that a given state of technology dictates certain questions and that the solutions follow logically from the questions.

  3. Real Soon Now by hob42 · · Score: 5, Interesting

    I found this other article even more interesting - 1974, issue 311, "In Praise of Hydrogen." It talks about how easily the School of Automotive Studies converted a traditional internal combustion engine to hydrogen, and how with only one major area of research (storage of hydrogen) we should expect our dependance on gasoline to be quickly and easily eliminated.

    Talk about vaporware (pun not intended, though also funny).

  4. Re:Let's not forget the user-interface... by NoMaster · · Score: 3, Interesting
    Plastic was "cheap" and steel meant quality. If the case wasn't heavy enough to kill someone with, it wasn't quality.
    Tell that to the Honeywell Rosy 26 teleprinter in my garage. Plastic case, 20 years younger than the 2 Model 100's sitting next to it, much the same feature set, but 3x the weight. I was going to throw it out today, but damned near killed myself just trying to lift it!

    FWIW, I suspect the real reason that Teletype Model 33 looks so ancient is that, from looking at the internals, it appears to be a clone/ripoff of a Siemens Model 100 or a Creed Model 47 - both much earlier models - updated with an "electronic" keyboard. IIRC, Teletype Corp bought (or maybe partnered with) the UK-based Creed.

    (Slashdotters with a mechanical bent really should look into the old electromechanical teleprinters. They're amazing machines; a real tribute to the ingenuity of their designers. Given a motor spinning at 3000 RPM, and no electronics, how would you convert a 5-bit code to printed text?)

    --
    What part of "a well regulated militia" do you not understand?
  5. Re:nothing to see here... by Ash+Vince · · Score: 2, Interesting

    And if you do happen to work in IT proffesionally, or want to, then knowing what mistakes people have made in the past or what they have done that has worked well can help you immeasurably.

    But then, who wants to do something well when you can do a half arsed job, spend twice as long ironing out the bugs and then get a reputation for being a fuckwit.

    --
    I dont read /. to RTFA, I read /. to offend people in ignorance.
  6. Re:I Grew up with those things by igb · · Score: 2, Interesting

    I always enjoyed #UPPER and #LOWER. The 1900 had such short instructions (24 bit words, 4 6-bit bytes) that you could only access beyond the first 1024 (I think) bytes by using a register (accumulator in those days) as a modifier. #LOWER memory was directly addressable, #UPPER required the addition of a 24 bit accumulator to the base address. And doing IO by initaiting the card (or whatever) moving, then doing as much as you could before calling SUSBY to wait for the IO to complete. Happy days. I'm not old enough to have done it for a living, but I had access to a 1902A (access in the sense of walking into the machine room with a deck of cards) in my early teens in the late seventies.

  7. Re:The metric I want to know by 91degrees · · Score: 2, Interesting

    But there were only a certain number of companies big enough to use a mainframe, and most of these would have been in the US an Europe until very recently. Other companies would have outsourced anything that needed a computer. I'd say world computing power definitely was higher than a 4GHz Pentium 4 by the time the home computer market had started, but the mainframes do mess things up

    But this is where the problem is. With a little research, it's quite possible to get a good estimate for the speed of, say a PDP-8, but unless you know of a good resource for sales of computers between 1950 and 1980, it's difficult to estimate how many machines there were.