Re: List Creep?
It's well worth *reading* articles - or at least looking at the pictures - before making comments like this.
In the early 1980s, civil servant Bernard Terry devised a 'portable text processor' to make his fellow civil servants more productive in the office and out. Electronics giant Thorn EMI designed the machine with help of a team of former Dragon Data engineers. As the Liberator, it launched in September 1985 to become the first …
Otherwise known as "how to fail".
They should have learned from their experiences of the first one that nailing the hardware spec down before even starting to build the prototype pretty much guarantees that the product eventually produced will be obsolescent at birth. Specifying the exact memory requirements for a new machine to be built any later than next quarter is insanity. I guess that this is yet more evidence of the Thorn monolithic dinosaur throttling the agile and think-on-its-feet development team.
Doing this for two planned future versions? Words fail me. I've often said that anyone in this business trying to design for the future rather than now had better have a bloody reliable crystal ball.
Did you notice when this was happening? In the mid-eighties, there wasn't the sort of market we're lucky enough to have now - you couldn't take the platform for granted. You weren't designing "for the future" you were designing everything, and hoping that your vision would gain market supremacy over all the other largely-incompatible visions being developed elsewhere at the same time.
I have to ask, TeeCee: were you even born then?
Back then the IT business was all gambling really, noone knew what was going to take off, compounded with the facts that materials costs were so much higher than now, overall TCO was much higher, and proprietary incompatibilities were all the rage. So yes, you had to design for the future constantly. The people who marketed their vision of the future won, everyone else lost.
Great series - good read right through to the denouement.
I suspect the people slagging them off haven't tried taking a product to market, and have little idea what's involved. It's great to sit here, with decades of hindsight to benefit from - but as others have said, this was largely uncharted territory. No-one knew where the personal computer market was going, and I doubt anyone (or at least very few) would have guessed just how technology was going to take off and electronics really take off down the better, faster, cheaper rollercoaster.
For those of us old enough to remember, this was all "unaffordable magic" to many of us, and the idea of being able to buy computers from two manufacturers and have anything in common was unheard of - in fact, even buying two computers from one manufacturer didn't automatically buy you commonality. What won and what lost was in part a big game of chance - you throw the dice and see what comes up.
As a side note - and here's an idea for someone at TheReg to follow up on in their history series - I don't believe IBM actually invented the PC. Back then there were multiple competing systems - all incompatible. The major names I recall had to some extent settled down to Apple II, Tandy TRS 80, and the Commodore PET in business - though there were still plenty of others around. As someone commented (I think) to an earlier instalment, PETs were really popular in technical environments because they used the IEE488 bus for I/O - it was pricey, but you could link multiple devices (printers, disk drives, plotters, scientific instruments, ...) to the one port.
Apple was pulling ahead in business because of Visicalc. Visicorp needed to pick a system to write their Visicalc spreadsheet for - and they chose the Apple II (I believe because it had up to 48K RAM as standard). Suddenly, department managers could mangle numbers on their own desktop on their own terms - and buy the kit from the departmental budget. On that, I was friends with an Apple dealer back then, so got to hear about some of the tricks used to get round buying restrictions - yes I saw the wad of computer printout he got for a network of Apple II machines all broken down to separate orders that were within the manager's sign-off limit !
At the time I worked in a large manufacturing business, the typical stronghold of IBM. In computing terms, *NOTHING* happened without the say-so of the Systems Department. If you wanted to crunch numbers, you had to apply for a terminal, apply for access, and pay for the resources you used. In our outfit, you could easily wait over a year for a terminal, if you got one at all.
But the Apple II in particular (and the others to a lesser extent) gave IBM and these Systems Departments nightmares - department managers could bypass them altogether. Another tale from my friend was that he had a stock quote he'd do for managers at this business - and he'd tell them that they'd have their terminal installed within a fortnight. The managers asking for quotes on Apple II machines didn't believe him, but without exception, the terminal they'd been waiting for (sometimes over a year) would appear very quickly once there was a risk of something not approved by IBM going in.
So this situation was desperate for IBM. They risked losing their stranglehold on business computing and they weren't capable of designing/building a "low cost" desktop product - all their expertise was in building big stuff, and productivity was measured in how much code you could write, not how small you could make it. The story as it's been told to me is that some small outfit built a desktop computer system - essentially by implementing an Intel design note which gave an example of how you could put together an 8088 system. IBM bought the company, stuck an IBM badge on the computer, got Bill Gates to provide them with some software - and the rest is history. Apple were virtually frozen out of large businesses because now the Systems Departments had something to fight back with - if you wanted something on your desk, IBM had a product to do it.
So when these guys were designing the Liberator, I really, really doubt that anyone (regardless of what they might say after the fact and with the benefit of hindsight) had any inkling of what IBM (with Microsoft) was going to do to the desktop (and eventually, the portable) computer market.
> I don't believe IBM actually invented the PC.
The first reference that I have to the use of 'Personal Computer' is a 1978 Byte ad for the Apple II. It may well have been used before that.
> Back then there were multiple competing systems - all incompatible
The weren't _all_ incompatible. No more so than today where iPad, Android, BB, PS3, Wii, etc are *all* incompatible. Androids, for example, may be all different but are all compatible. Since 1975 many desktop systems were 8080, 8085 or Z80 computers of various forms running CP/M or MP/M (1978) or compatible systems such as CDOS or TurboDOS and were thus compatible - at least with each other. Even an Apple II could add a Z80 Softcard and run CP/M and software such as Wordstar.
> IBM .. they weren't capable of designing/building a "low cost" desktop product
IBM had been building small computers since 1975. There was the Series 1 desktop system which competed with similar machines from DEC and DG. The 5100 portable (luggable) with a 5inch screen was released in 1975. This was followed by the 5110 and 5120 desktop machines. Then there was the System/23 Intel 8085 based desktop machine.
When IBM saw Apple IIs getting installed in their mainframe sites as personal computers running BASIC, Visicalc and CP/M software with a Z80 card they wanted to counter this with a CP/M machine that was slightly better. They modified the System/23 planar (mother board) to take a Z80 and then again for the 8088. They got MS to supply BASIC and asked DRI for CP/M-86 but MS talked them into taking QDOS which was owned by SCP at the time.
The 5150 IBM PC Model A planar was laid out the same as the System/23 from which it derived. The Model B (which I have here) has a reworked planar with the slots closer together and capable of running more than the 256Kb which was the limit of System/23 (bank switched) and Model A.
> IBM bought the company, stuck an IBM badge on the computer
Not true at all.
I don't believe IBM actually invented the PC
Who does? IBM don't claim to have invented personal computers, as far as I know, and no one who knows anything about computer history thinks they did.
Richard's already mentioned some of IBM's small machines prior to the 5150 PC. I don't know that I'd consider the Series/1 a PC, and the System/23 only came out a month before the 5150, but the 5100 Portable Computer from 1975 certainly qualifies. And its prototype, the SCAMP, was built in '73, making the design even older than the Altair 8800 and the S-100 bus, which some people do claim for the "first PC" title. But then so was the HP 9830, which was released as part of their "programmable calculator" line but contained a full BASIC interpreter; it came out in 1972.
Of course the definition of "personal computer" is debatable, but typically - at least in the '70s and '80s - it meant a microcomputer, a hardware implementation that was reasonably compact, power requirements suitable for home use, and a single-user OS if it had any OS at all. Sometimes people include price as a criterion as well. But the main idea was that it was personal: you could point to it and say "this is my computer" without sounding like an idiot.
In Woody Allen's 1969 Take the Money and Run, there's a job interview where the protagonist is asked whether he's used "an electronic digital computer"; and when asked where, he says "my aunt has one". That was funny in '69 because the idea of a personal computer was ridiculous, an idea out of science fiction. Sometime in the 1970s, it stopped being ridiculous. It wasn't the 5100 or 9830 that realized the idea of personal computing, though - those weren't machines that the typical aunt was going to run out and buy. The 8-bit micros were the ones that opened that door.
 The Series/1 is generally classed as a minicomputer. IBM did create a "portable" version for the US Marine Corps, but if Wikipedia is to be trusted, it came in at least two pieces that weighed over a hundred pounds each. And its main OS was a multiuser system.
 For one thing, the 5100 started at about $39000, adjusted for inflation.
I recommend looking for the IEEE magazine "Annals of the History of Computing" . Comes out bimonthly (or is it quarterly? Can't remember) and has many interesting articles and anecdotes like this in it. Some are really esoteric, behind-the-Iron-curtain stuff, and others focus on systems I remember using...
Biting the hand that feeds IT © 1998–2019