Halt and Catch Fire (AMC Series based on the founding of Compaq)

[VK3DRB]

Er, no. The PC for a niche market was the one before the IBM PC we all recognise today.

Wrong. The PC before it was the 5100, 5110 and 5120 machines. They were designed for professionals and small business, not for hackers. I used to repair these machines and none of my customers were general consumers or hackers. The cost was prohibitive for a start.

It had a proprietary IBM CPU and you could only use an interpreter on it (some kind of BASIC, I think).

Correct. Except most models had TWO interpreters on them - BASIC and APL.

The IBM PC we all know was a result of IBM worrying about the number of small businesses who found they could do their basic accounts and database activities on the Z80 based CP/M machines which were becoming popular.

I was present at the very first internal announcement of the IBMPC at IBM Australia - at 211 Sturt St, South Melbourne in 1981. The senior marketing manager clearly stated to us engineers the machine was aimed at a very small market that could grow potentially make up 2% of IBM's business - the hackers market (the term that was used at the presentation). Hackers in those days being people who were interested in modifying what was under the bonnet, unlike the bastardised original term "hacker" used by the media today. Very quickly, the official launch aimed it at small businesses and home owners who knew nothing about computers (via the Charlie Chaplin advertisements). IBM sold the machines through VARs, VADs and VAMs, which limited their aceessibility to the home user.

Not many business ran CP/M machines back in 1981. IBM never saw the CP/M machines as a threat in 1981. In fact, few businesses owned a computer at all in 1981.

That was actually a very smart move by IBM. They didn't want to stop competitors. Anti-trust concerns meant they only wanted to slow their competitors down. They did enough to ensure that, while making the machine open enough that third party cards flourished, and the PC ended up in all sorts of unexpected applications, broadening its market.

The IBM antitrust case was dropped in 1982. IBM specifically intended publications of the technical details was for OEMs to make I/O boards and peripheral, not the main system. This OEM "partnership" paradigm was spelled out on numerous occasions in the early 1980's within IBM at technical and marketing meetings and in publications. It backfired over time. IBM fought tooth and nail to keep market share by offering exceptional quality in manufacture (and they did this very well), but by the the mid 1990's, IBM realised that people want low cost above quality. Today, you cannot buy an IBM PC or even a laptop. They lost the market completely.

Microchannel was a very stupid idea, but OS/2 Warp wasn't around in those days. The original OS/2 was a screwup, mostly because it was locked to IBM's microchannel  hardware. That is what seriously limited its appeal. I don't remember it being very expensive. The whole machine was more expensive, as a full scale OS like OS/2 needed a lot more memory than was being fitted to most machines at that time.

Wrong. Microchannel was technically quite superior to ISA... better specs, smaller form factor, 16 AND 32 bit buses and unlike with ISA the I/O board could become the bus master.

And the original OS/2 was NOT locked to microchannel. I was the first person in Australia to install OS/2 in 1987. I got a copy of OS/2 1.0 prior to release the very morning the first proof copies were received at IBM's Diskette Replication, for engineering purposes. It was installed on an IBM 80286 machine with 2MB RAM. The bus was ISA. It ran without a hitch and I found it much more useful than DOS, especially with applications running on a Token Ring network.

Warp was more expensive than Windows in the earlier days. Arguably the ease of piracy of Windows made it popular, as did its relatively lower price and better marketing. By the time IBM dropped its price on OS/2 Warp, it was too late. They lost the battle badly and surrendered in the late 1990's with their tail between their legs.

There are a time when OS/2 Warp was released that Warp may have defeated Windows, but IBM screwed up with lack of innovation in marketing above everything else.

[nctnico]

I'm in the US, and even I had issues with the AMC website.  Turned it was adblock plus.  Turn if off for that page (and the eevblog of course!) and all is good.

So far it seems like Mad Men with a touch of the Social Network.  But without the amazing dialog of Aaron Sorkin.  Also if you're into great dialog, and not an American hater  - check out "The Newsroom"....

'The Newsroom' is nice too. A lot of good criticism on American society!
When I hear or read about a new TV series I just check the first two episoded to see if it is something worth watching. The first episode of 'True detective' is just crap (skip it) but in the second episode the story starts to catch on.

[Bassman59]

Interestingly, I had this conversation with another group of forum-goers elsewhere and mentioned Dexter and Homeland as examples of similar, but IMHO, far superior dramas.  It seems our taste was 180-deg out of phase on that one.  I also thought The Newsroom was (is) excellent.  May be just a different cup of tea.

My wife and I started re-watching "The Sopranos" from the beginning, and it's impressive how well that show holds up. I mean, it's about a bunch of psychopaths who live by a code which they regularly bend and break, but it's fascinating. The writing is snappy and the acting goes right up to the line of being a ham sandwich but doesn't cross it (too often, at any rate).

Plus, I'm from northern New Jersey, so I recognize all of the landmarks.

As for "Halt And Catch Fire," there are technical errors in the show that make me want to throw things at the television. And beyond the glaring errors, there were things which were obviously there to make the reverse-engineering process seem more visible to the (not-techy) audience. (Hello, HP1630 logic analyzer, anyone?)  Plus I hate the sales guy character, and I don't remember any female hackers like Cameron. (Girls like her were always at the punk rock shows I went to the night before Circuits exams, though.)

[SirNick]

It's dramatized, so I forgive the technical errors.  On a whole, it does a pretty good job of being technical enough without losing the general populace.  I roll my eyes from time to time (like the idea that a power surge from plugging in a vacuum cleaner is going to destroy floppies), and then raise my eyebrows when they go into territory I didn't expect them to touch (like talking about PSK modulation).  The writing would never hold up to the scrutiny of a bunch of engineers, but that's OK.  It's fiction.  Just be close enough that I can suspend disbelief and I'm good.  (Read Dan Brown's Digital Fortress for an exercise in WTFery.)

Similarly, I disagree that the truth about cloning PCs and battling it out for market share would be a more interesting series.  A more interesting documentary, perhaps, but fiction?  Nah, cutting-edge, seat-of-the-pants, hardware hackers giving birth to something that can change the industry -- that's a fun plot.  The fact that it vaguely resembles actual history is a bonus.

Cameron is a tough sell, I agree.  How you are expected to jump straight out of early college courses with her level of experience ("Who knows anything about VLSI?") is beyond me.  However, I did work with a coder that was fresh from high school, and showed a maturity well above many of the professional programmers I've met.  I attribute that to his working on a few OSS projects, and learning practical skills from that ethos.

Finally -- The Newsroom -- for anyone that hasn't seen this show, you should find the intro to episode one and watch it.  It had me hooked in the first five minutes.  You don't have to be pro-American nor anti-American.  It's a sobering, critical view of America's faults, but it's all completely fair.  It's not bashing for the sake of it.  At some point you have to stand back and understand where your weaknesses are, to prevent being hoodwinked by them.  The fact that this level of honesty has to come from a fictionalized TV show, and not the genuine news media, is a little sad, but welcome nonetheless.

[westfw]

Cameron is a tough sell, I agree.  How you are expected to jump straight out of early college courses with her level of experience ("Who knows anything about VLSI?") is beyond me.

I dunno.  I was almost her, sort of.  I mean, I followed personal computers in HS (late 70s), and poured over the Intel 8086 manuals when it came out while I was in college.  My college was gifted a bunch of SDK86 development systems (kits!) from Intel, and we assembled and fiddled with them in my Senior "Digital electonics" class.  Me more than most, cause I had the programming bug even though I was an EE.  I did my Senior Design Project on that board; all in x86 assembler (cross assembler running on a CPM system.)  That was pretty much all it took to be an x86 "expert" on graduation (81); you didn't need to know about the chip internals and VLSI, you just needed to have been paying more attention than the "pros" who were building Qbus cards or programming IBM Mainframes in PL/1.  When the IBM PC came out and my first employer bought one with "incentive funds", I was all over it.  Unix to x86 cross assembler from MIT: cool.  Published BIOS interface: excellent!  H19 terminal emulator: done.  XMODEM upload/download from the mainframes that were my "real job": done.  Try out Microsoft C, v1, in the days when that meant having your source code on one floppy drive, and the various passes of the compiler on the other (switch floppies for each pass!):  yeah (eww!)  Go off and become rich and famous selling PC communications SW:  well, that part didn't work out so well, but I did try.  (learned a lot about SW publishing legalities.  Spent money on lawyers.  Sigh.) (and I stuck with the mainframes for another couple of years instead.  Work entirely on PC-class machines?  Nah; no Internet, no multiuser stuff; mainframes were a lot more fun.  Then got heavily into networking, largely because of ... my x86 experience.)
The point is, you only needed to know a *little* more about x86 than the people who were busy doing other things to be a "x86 wiz-kid", and that wasn't all that hard.  And you do get that class of people coming straight out of universities...

[SirNick]

OK, I guess I could see that.  My only eye-rolling moment really was the part where he meets her in that classroom.  The roll-call of technical disciplines seemed over the top, but you're right -- the systems were far less complicated, so learning "a lot" about several things wasn't quite the career-long process it is becoming now.  Aside from that scene, she's a talented assembly hacker (no problem there), and owned up to not being a hardware expert when she tried recruiting Donna for her start-up.

Now, Claudia from Warehouse 13?  I suspect her skills are being exaggerated. 

[German_EE]

Power surges and floppies.

Well

If you had either an Apple ][ or an Apple //e with floppy drives you knew to ALWAYS open the floppy drive door before powering your system on or off. If you did something with the power and the heads were in contact with a disc they were nearly always corrupted. So, given that this is about the same time period it's possible that the surge from a vacuum cleaner could do something nasty.

I think I'm showing my age 

[SirNick]

Wow.  That is tremendously bad design.

[nctnico]

Not really. Floppy drives are a standalone component and they have an active low write enable (IIRC). If the power sequencing is slightly off then you could activate the write head during power up or power down. I still make sure there is no disk in a disk drive when I switch a device on or off.

[SirNick]

So am I the only one who tries to think about what happens to logic levels that control some other device when the voltage goes below normal operating level?

[SeanB]

Many designers are caught out by supply rails that either power up out of sequence, or that drop out too fast or too slow. Some IC's have very specific requirements for power supply, in that they need the supplies applied in the correct sequence, and that they ramp up within a defined time. A lot of microcontrollers with brown out detection will not reset properly if the initial supply comes up too slow and the on chip reset circuit times out various parts at different voltages and then it operate erratically or not at all.