Hacker News Reader

Comment by MrRadar | Hacker News Reader

No mention of the PS/2 and Micro Channel Architecture? MCA was IBM's attempt to put the genie back in the bottle and redefine the PC back to being something they wholly controlled, and it utterly failed. It also didn't help that they were late to the market with 386-based computers (Compaq having beaten them to the punch with the Deskpro 386) which already signaled they were losing their market leadership position that allowed them to define the PC platform. In response to MCA, the PC industry formed an independent consortium to define the EISA bus and later Intel itself eventually took over as the de facto standards originator for PCs by developing PCI, USB, ATX, ACPI, AC'97/HD Audio, (U)EFI, and a large part of the other foundational standards on which modern PCs are built.

View on HN

According to this video[1], Compaq's portable also bested IBM. It was Compaq's first product, and everyone thought IBM's upcoming portable would wipe them out, but they held strong and counted on IBM to have production problems, which they did, so Compaq held onto this segment.

[1] https://www.youtube.com/watch?v=HEMhpInIACk

So handy for pirating software at PCCUG meetings.

PS/2s were overpriced and underpowered, and people I knew told me that a surprisingly high percentage of them arrived DoA. That was basically the end of "no one ever got fired for buying IBM".

Yes, I recall the MCA, it was born a dead horse and never run a furlong. I recall all the companies who went through the lengthy MCA add-in card signup process - that took a year or more to be granted, whereupon they fielded cards that never sold. I recall the surplus tables at the Dayton hamfest being piled high with MCA detritus - cards that were eventually scrapped for gold. IBM tried mightily to force this dead interface, few bought the idea and made boxes - few sold. They were dead before they sold one. The supposed superior MCA bus, was soon humbled by a string of faster and better busses. There were a few old fogies who clung to big blue - I see them from time to time wandering in the wilderness...

I remember NCR making a few Micro Channel machines during its short not-quite-heyday. I always thought the idea of putting card configurations onto floppy disks was downright stupid. I wondered what the EISA crowd was thinking when they adopted that for EISA systems.

Yes, Eisa, Vesa and MCA whose sun rose and set so rapidly. I had EISA and VESA bus boxes, but soon left them. I had 7 build your own box retail ops in the 80's, so it came in and went out, and as soon as it faded I stopped buying. Never bought MCA - terrible sticker shock, my student base would not go for it. I think that IBM ran with floppy config and EISA followed like a puppy dog and it died a quick death.

Relevant Computer Chronicles episode from 1988: BUS wars

https://youtu.be/aYt4ZPxVkSs

IBM engineer responding to the accusation that MCA is proprietary (around 14' 15"):

"I really don't understand the term proprietary. To me that would mean that it's not available [snip] and it's available for use by anyone."

That's like when Intel switched to SECC (single edge card connector) with Pentium2/Pentium3 and stuck it on an Atari(NES?)-like catridge: this was to thwart clones and outstanding cross-licensing with AMD. What a weird decision. I'm not even sure if it worked tactically because it only lasted until Pentium 4. Weird days.

The side connector lasted until Socket 370 in the Pentium III product cycle. The new socket took Pentium III chips that had integrated L2 cache (versus earlier PIIIs which had cache on an external chip).

Pentium 4 never had a side connector model, but did have a “socket shrink” in the transition from Willamette (0.18 u) to Northwood (0.13 u) cores. Socket 478 was kept until the Prescott cores (0.09 u). The Prescott cores were also offered with an LGA 775 socket later.

Yup, that's why I said "until Pentium4". P6.0/P6S actually introduced multi-die packages at Intel, not the SECC. Well, actually that's not true either, because I worked on Sidewinder: a 486 that fit in a 386 socket with an ASIC on the top of the package that matched the busses. So it was technically "multi-die".

Wow. Transistor sizes measured in um instead of nm. That takes me back...

I think the SECC1 design was more based on practicality of cache that was available at the time; While the connector was 'proprietary,' the Bus design itself was probably a more protect-able part of locking AMD out.

FWIW AMD's original Athlon used a Slot connector as well, for similar reasons.

Intel and AMD actually used the same physical connector, with the slot keys on the connector opposite for AMD than Intel.

Slot 1, or slot 2 for Xeons. AMD had slot A for a hot second in the early Athlon days, before they went back to a PGA.

It's weird, yeah, but is it worse than aligning an expensive bunch of easily bent metal pins into a grid of small holes?

In terms of usable data/electrical connections, the slot is worse. This could mean you have to use multiple clock cycles to send the same data, due to fewer direct connections.

They did talk in the article about how slow IBM was to adopt the 386.

And that then helped to kill OS/2 only because it delayed the release of OS/2 2.0 as that version needed a 386 to run.

OS/2 2.0 used features found in the 386 to greatly improve the MS-DOS experience.

IBM instead kept pushing the OS/2 1.x versions all of which had very bad support for MS-DOS applications.

That delay was a massive bonus for Microsoft as it gave them time to work on Windows 95, which used the same 386 CPU features to deliver similar improvements in running MS-DOS applications.

Yeah, but also remember that Windows NT started as the next version of OS/2 after MS and IBM stopped collaborating. The UI was basically the presentation manager from os/2 until 4.0, and MS provided OS/2 APIs in NT.

I think you may be misremembering. Windows NT 3.51 had a UI that generally mirrored Windows 3.1:

https://guidebookgallery.org/screenshots/winnt351

You're totally right, I am. Thanks.

Relased in late 1987, Windows 386/2.0 supported multiple DOS boxes using virtual 8086 mode.

Win3.1 added some features which made windowed DOS boxes more usable.

Win95 came three years after OS/2 and Win3.1 were released.

Hard to adopt the 386 when they basically bought the whole supply of 286's.

Reply on news.ycombinator.com