Re: New-Gen "IT" People ... Think DISASTER

For me, the nice thing about the university courses was the access it gave me to computers, although through a family connection I had gotten my hands on a couple of small office machines before that. The first machine I actually played with (at the tender age of 15) was a Univac 1004, a hopped-up electronic version of the IBM 407, so I even got a taste of plugboard wiring.

Alongside university course assignments, I wrote a lot of fun programs for myself, and learned as much from that as from the actual courses. However, disenchantment set in when I realized how far Computer Science was from the Real World [TM] (at the time, the CS department was still under the department of mathematics). Pretty soon it turned into a lot of airy-fairy theoretical stuff (one course hit you with a new programming language every two weeks), with a lot of time spent contemplating the whichness of what with no practical applications. Meanwhile, I found I enjoyed getting down to the metal with assembly language, which didn't go over well with the faculty.

Fortunately, between my second and third years, I managed to land a part-time job programming at a small service bureau. That's when the real learning began. When it came time to go back for a third year of university, I arranged my schedule so that I had Thursdays off. I worked at that service bureau on those days, while during the rest of the week my disenchantment with Computer Science deepened, and I discovered that theoretical math really wasn't my forte. At the end of my third year (a disaster aside from the programming part), I dropped out, went on full time with the service bureau, and have been programming real-world applications ever since.

I've always said that this will be the epitaph of our society (spoken in a nasal whine, of course).

Two words: s*it happens. Those who refuse to believe this should get what's coming to them - but unfortunately too often don't.

It matters not if you win or lose, but how you lay the blame.

Is that anything like the Canadian Phoenix federal payroll screw-up?

I've always said that payrolls should never have been computerized, because computers are logical and payrolls aren't. But people have become so enamoured with complexity - confusing it with sophistication - that things become messy very quickly.

Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away. -- Antoine de Saint-Exupéry

Give them a break, programming in Roman numerals was a bitch.

Yes. Computer science is relatively speaking bollocks. Software engineering however is extremely valuable

Nope. In my day there was a 370 and a Sigma 6 (engineering department), I missed Titan by a few years.

Victoria University of Wellington (NZ).

From memory ours had 32 KB main memory and another 32Kb, in a separate cabinet, that operated as fast disk store: After it was booted, all standard executables and, IIRC the Algol standard library, were loaded into it. Any unoccupied space in it could also be used as a sort of paged memory for holding and operating on large arrays. It had a line printer as well as paper tape readers and punches. By the last time I saw it, it had acquired four 1/2" inch mag tape drives, mainly, I think, because it booted faster off mag tape and anyway, I'd imagine that would be more resistant to wear & tear than paper tape,

Luxury. We had perhaps 8k of store and the only I/O device was a paper tape read/punch. The machine was in fact an 803 (I misread your OP) and so slow that it couldn't drive the reader at full speed. And it only had autocode - no Algol.

It supposedly was intended for use by the optics group in the Physics dept, for ray tracing, but we undergrads who had some slight interest in Astronomy wanted to use it for orbit calculations, but we didnt really know how.

You should spend less time paying attention to culture war nonsense.

It is very hard not to be overwhelmed by it Richard.

When every single part of your life is affected by it.

Frankly I feel your comment feels like saying to a 1930s German Jew 'you should pay less attention to that national socialist nonsense'.

The Elliott 503 preceeded the 803: I've seen 803s in NMOC but never seen one running.

The 503 was physically huge, being four cabinets, each about the same size as a two-door wardrobe, an engineers monitoring panel which was about 4 feet by 3 feet, a lineprinter, two paper tape readers, two paper tape punches and a control console. The latter was a small desk with keyboard and teletype-style printer inset flush with the desktop. IIRC had two

It used 39 bit ferrite core memory and needed those big cabinets because the logic was built from discrete transistors, so the 503 must have been one of the first transistorised computers.

Transistors had been around at least 10 years prior to its introduction.

The late 60s had them cheap enough for 'transistor radios' to be the i-pods of their time.

"Following the invention of the transistor in 1947—which revolutionized the field of consumer electronics by introducing small but powerful, convenient hand-held devices—the Regency TR-1 was released in 1954 becoming the first commercial transistor radio. The mass-market success of the smaller and cheaper Sony TR-63, released in 1957, led to the transistor radio becoming the most popular electronic communication device of the 1960s and 1970s.¨ So building a 'puter out of DTL or RTL using discrete semiconductors was utterly feasible in the the late 1950s. Slow, extremely bulky and pretty power hungry, but feasible.

"In January of 1954, supported by the military, engineers from Bell Labs built the first computer without vacuum tubes. Known as TRADIC (for TRAnsistorized DIgital Computer), the machine was a mere three cubic feet, a mind-boggling size when compared with the 1000 square feet ENIAC hogged. It contained almost 800 point-contact transistors and 10,000 germanium crystal rectifiers. It could perform a million logical operations every second, still not quite as fast as the vacuum tube computers of the day, but pretty close. And best of all, it operated on less than 100 watts of power. ¨

I don't know about that one, but for sheer venality, bad management, incompetence, (seeming lack of) and purely incompetent system design and development and swindling the users this easily takes the biscuit.

Private Eye published a good description of the scandal:

Indeed: another interesting example of cockup by carelessness turns out to be the recent delivery of samples from the Bennu:

The first computer anything I remember seeing were the 7 transistor radios owned by the (relatively) rich kids in 1956/7 and the single transistor- boosted crystal sets that the rest of us built around then.

I never saw the inside of the Eliott 503's cabinets, so have no idea of hoew its transistors were packaged, only that it was described as not using ICs. The first machines I saw th interior of were the ICL 1902 and

Side comment: I thought the Optical Mark Reader was a great idea: it read A4 paper sheets which were marked up with a pencil. I was involved in a project in Wellington where the hospital's Heart Unit used them to record patient notes: the page were designed by hospital staff, who apparently preferred then to hand-written notes. We just wrote a fairly simple system that red the OMR forms and stored the data on tape as well as translating the marks into printed pages.

Have you ever tried to read a doctor's handwriting ?

There's supposed to be a TV special (docudrama?) about it on soon. Buggered if I can find it though. OK! "Mr Bates vs. The Post Office" to be aired on ITV, starting 9pm New Year's Day

Back to front; the 803 came first (1960 or so), then the 503 in 1963. The 803 was bit-serial, so the cycle time was around 2msec [1]. The 503 was much faster.

It's amazing how much better systems work when the designers allow the users to comment on what they'll be using.

You NEED to have local backups - pref several layers - of all your important stuff. You do NOT rely on 'cloud' storage.

You NEED to have at least locally-usable versions of your day-2-day important software. Can users run Office apps on their individual boxes if the net goes down ? Probably not. Ergo something like LibreOffice should be installed on every relevant box.

All e-mail thru M$ or Goog or its friends ? Nope - need to have a local mail server that CAN be activated. May not work if the Net is badly damaged, but at least you TRIED.

Accounting/payroll ? Can you do it THERE, at least half-ass ? Spreadsheets usually work quite well - in a disaster you will get a break on govt/tax nuances. Oh, can you pay people with CASH MONEY in-hand ?

Users should be able to at least get snail-mail, type up letters/responses and print them. Not as 'efficient' as online stuff, but the world worked on that paradigm thru the 70s and early 80s so it DOES work.

"Cloud" always SOUNDS so great - but the younger set do NOT seem to realize how FRAGILE that model can be even on a good day. Gotta put fair effort into being self-reliant, self-sustaining. It's the 'professional' thing to do, rather than "Blame M$" excuses. Assume everything can be screwed for at least a month and some of your data will NEVER re-appear.

Hey, got any good old LAND-LINE phones ? :-)

Complex systems soon become so "normal" that people just cannot imagine them NOT working. Then they start to make bad mistakes. I was in a serious hurricane area some years back. NO infrastructure remained - all the wires were a tangled mess, water-lines were torn up, even the roads were blocked for about a week. People had to WALK places. No phones. No PCs. Nothing worked - even the banks were down for awhile. The kiddies were in a daze because they couldn't text anyone. Pay in CASH. That was a NATURAL disaster, not 'cyber-war' or giant solar-flare.

Fair enough: I hadn't seen an 803 until I first visited the NMOC , so around 35 years after I last saw VUW's 503. The last time I was at the NMOC there wasn't lot of technical info on the 803 or an Elliott timeline and as I haven't seen one run.

I have heard that the 803 instruction set is identical to the 503: is this true?