Something Random

[Mercutio]

Man, I don't think anything Compaq made deserved a decent reputation even into the mid-90s when 486s started to become mainstream.
I should note that was the point when product lines started to segregate and we got consumer and business systems, but Compaq was always one of those companies that could fuck up making ice. Its business machines would be full of different-sized Torx screws and easy access to nothing, which made them an absolute nightmare to work on.

I'm sure I've made this point before, but I remember being able to boot off USB on the very first IBM desktops I ever saw with USB ports. I had an extremely exotic 4GB 3.5" drive in a USB enclosure for doing data transfers in 1998 or 1999. Seeing that it was a boot option on those guys led me to install DOS on it and sure enough, that shit could start that PC. Which is the world of difference between how IBM did things and what Compaq was doing with its engineering. I have an IBM laptop from that era that could also boot from USB. Thinkpad 600, maybe?

(I actually have a collection of Thinkpads in my bedroom closet that's about four feet high. They're the retro-hardware I'm most likely to keep)

Regarding the drive size limit, Compaq (and later HP) had a bunch of BIOS programmers on staff, so when working with their products back in the day, it was extremely common for fixes to be BIOS updates instead of OS patches. It's very possible that your laptop had an update that's been lost to time since I'm sure those updates are long gone from the internet at this point.

 

[sedrosken]

I've seen some ridiculous uptime figures and heard stories of Deskpro ENs in particular living through stuff they really had no business doing. Yeah, the torx screw thing is annoying, and is where you can see the Compaq DNA in HPe machines today easiest, but they're also slotted, and I like the way they enclosed the slot rather than cutting the entire way through the head like IBM did on their hex-head bolts. I've had a handful of Aptivas and PC300GLs and such and they all use that annoying hex head with the slippery slot for a flathead.

I can't really blame Compaq for not doing something that basically no one else did aside from IBM at that time. Look how many times they'd been wrong about the direction of the PC industry by that point. MCA, anyone? Maybe a sprinkling of OS/2? USB booting wouldn't become common until the Core2 era, maybe the late P4 timeframe. And considering I can write a plop image to a floppy or CD and force platforms to boot from USB that really shouldn't (even a 486, if you stick a PCI USB card in one) I find it's a forgivable flaw in $currentYear.

I'm not saying Compaq went south for no reason -- buying DEC was probably a misstep, as much of a sure thing as it seemed at the time -- but I will always lament that it was Compaq that died out, and not HP. Notice how once the acquisition was pushed through, the Vectra line just sort of quietly went away... Almost like it was forgettable in the first place.

Interesting point about the BIOS upgrades, but I bet it's a moot point on this, where the name of the game was clearly to shit it out onto the market and capitalize on the sub-2000 dollar notebook market as much as they could before the internals were hopelessly obsolete 6 months from release. The Presarios were consumer grade machines, I'd be willing to bet the BIOS updates only really came for the enterprise gear where support contracts would have been involved.

 

[Mercutio]

MCA, anyone? Maybe a sprinkling of OS/2?

sed, have you ever actually used OS/2? It really was a great idea for its time, especially if you were looking at it versus Windows 3.1 or even NT 3.5. The big down side at the time was just getting a PC with the resources to run it alongside the DOS and 16-bit Windows software you probably wanted to use on it, but if you had a moderately nice PC, it was fantastic. Tannin can come along shortly to elaborate; I moved from OS/2 3.0 to Linux in 1994, but having access to non-stupid multitasking on a home machine in the early 90s was HUGE and made me really hate having to deal with Windows 3.1.

MCA was over-engineered but it did work and it was fast. IBM saw things wrong with ISA and attempted to address them in the IBM-fashion of the day, which was to just do things and assume everyone would follow along because that's how mainframe tech worked. MCA was both technically faster and offered device-to-device I/O. IBM was proud of it, and it cost more than getting EISA, but remember we're talking about an era when an average PC would've cost $5000 in today's money. Those cards were a drop in the bucket at time of purchase. The biggest hassle was digging up the configuration floppies those PS/2s needed to have something changed.

This whole thing has reminded me of a song that you've probably never heard, so here you go:

 

[sedrosken]

Right, I'm not saying there wasn't technical merit to either one:

- MCA was kind of brilliant especially for its proto-PnP arrangement. It was just too expensive and not enough companies made peripherals for it because the licensing was onerous. The only relatively common cards you'll find were made by IBM themselves, usually Token Ring NICs or SCSI cards, or are modern clones of original, much rarer cards like the Snark Barker MCA. The objectively correct solution to the issue of ISA wouldn't be puzzled out for a few years, and it took Intel to do it, but... MCA walked so PCI could run, I personally think.

- OS/2 is a technical marvel, and it'd better be considering how much DNA Windows NT still has from it today, but the mere whisper of Windows 95 killed it dead in just about everything but ATMs and the homes of superfans. I've used it, but only ever successfully managed to get Warp 4 running -- even on my IBM-manufactured 486 Blue Lightning motherboard with its IBM CPU, I never quite got anything earlier to run without issue. Maybe it's that that system uses VLB instead of MCA.

But in typical IBM fashion they were -- I guess I'd call it arrogant -- about it. As you said, they were used to making a decision and everyone falling in line. But by the time they were making these decisions they did not hold a controlling stake in the PC market anymore, and everyone else rightly told them to kick rocks. On the other hand, if IBM had held such an iron grip over the PC platform from the beginning, it's likely it would have faded into obscurity by 1985 as the clones were what made the industry what it is.

 

[sedrosken]

In other news I've learned the display on the Presario is actually one of those funky Hitachi High Performance Addressing things, a missing link between passive and active matrix that massively improved image quality, but didn't do much to prevent ghosting. The improved image quality is why I initially thought it must have had an active matrix screen -- I've seen real passive matrices before, they look like complete garbage compared to those. Apparently HP used some of those HPA panels in some mid-market Pavilion laptops as well. They're still useless for games where the screen moves... at all... but, they at least don't make me want to rip my eyes out looking at a still image.

 

[sdbardwick]

I like having sed around. Keeps reminding me of all the stuff I used to know and keep updated in the grey matter, but no longer have to deal with on a day-to-day basis.