61–70 of 85 posts
Earlier quoted context omitted.
The Model M's buckling spring mechanism has absolutely nothing to do with the mechanics of the Selectric keyboard.
Your not wrong, but the design of the model M sought to emulate the feel of the Iconic Selectric, as well as the model F with the solenoids under each key, equally dissimilar to the Selectic's internals.
It had one solenoid that knocked the case.
Earlier quoted context omitted.
Well, then I might as well point out, that IBM did make a 8-bit adapter for VGA, so I could use my Sony Trinitron with my PC XT 10Mhz clone. Beat the daylights out of my roommates 6Mhz AT, and ran Xenix, which most unfortunately turned me into a Xenix expert of crap ware. After a bad compile, it destroyed both drives file system, so I went back to just Windows 1.0.6, and DOS.
IBM itself or a 3rd party? Because the last IBM 8086 PC, AFAIK, was the PS/2 Model 30 and it didn't ship with VGA. It's an important difference.
PS/2 failure is also one of the reasons why PC compatibles stayed around, its main goal was to regain control of the PC.
The PS/2 was PC compatible. The term means "can run IBM PC software". The PS/2 could. It was compatible. Same OSes, same apps. Remember this was the 1980s. There were almost no protect-mode 32-bit OSes around, and the few that there were then had PS/2 versions. DOS and DOS apps ran fine on PS/2s, unmodified.
https://en.wikipedia.org/wiki/Micro_Channel_architecture
> It was thought that by creating a new standard, IBM would regain control of standards via the required licensing. As patents can take three years or more to be granted, however, only those relating to ISA could be licensed when Micro Channel was announced. Patents on important Micro Channel features, such as Plug and Play automatic configuration, were not granted to IBM until after PCI had replaced Micro Channel in the marketplace. The overall reception was tepid and the impact of Micro Channel in the worldwide PC market was minor.
https://en.wikipedia.org/wiki/IBM_PS/2
> The PS/2 line was created by IBM partly in an attempt to recapture control of the PC market by introducing the advanced yet proprietary Micro Channel architecture (MCA) on higher-end models. These models were in the strange position of being incompatible with the IBM-compatible hardware standards previously established by IBM and adopted in the PC industry. However, IBM's initial PS/2 computers were popular with target market corporate buyers, and by September 1988 IBM reported that it had sold 3 million PS/2 machines. This was only 18 months after the new range had been introduced.
Earlier quoted context omitted.
I thought OS/2 Warp was really neat. I worked IT at Bloomingdale's and at the time they were a strictly IBM house. They used IBM computers running OS/2 Warp in their Bridal Registry kiosks. Since this was a kiosk they had touchscreens and they worked very well and had a "sparkle" in the screen, as the early touchscreen CRT's did. If you lost power, or shut them down incorrectly you would get a bunch of DLL errors and…
re-image? did that exist then?
No doubt the pivotal importance of IBM's PS/2 range and I was instrumental in my organization acquiring many but that wasn't without problems as we already had many ISA machines so PS/2s weren't compatible at a board level. That increased the cost of maintenance amongst other things. However, when I think back the two most disruptive problems for us were the development of the 80286 and the failure of OS/2. The 80286…
Can you expand on the explanation as to why Intel said they designed the 286 as they did?
Also the PS/2 keyboard and mouse connectors. Before that mice were speaking to serial ports or required an add in card with a nonstandard port. the people buying PS/2's were often very concerned with mice, they'd seen them on TV and they wanted one.
I was always jealous of the Mac's ADB. You could daisy chain it, plug any device into any port, it was nice.
Not to understate the importance of the PS/2, but... To say OS/2 failed because it didn't run more than one DOS app at a time is not really accurate. Well before Windows 3 (or NT) was established, OS/2 had its own app ecosystem and it was very pleasing to use, if you had a fast computer. Early OS/2 was meh, but from Warp and on, it was a very nice desktop OS. OS/2 died because Microsoft cut deals with all big OEMs wh…
Are you talking about Microsoft/IBM OS/2, i.e. 1.x, or IBM-only OS.2, i.e. >=2.x here? For 2 and above, the IBM only versions, well, yes, kinda sorta. But that was too late. The game was already over. OS/2 2.0 came out in April 1992. Windows 3.0 came out in May 1990, 2 whole years earlier. It already had established an ecosystem before 32-bit OS/2 appeared. Secondly, OS/2 2 really wanted a 386DX and 4MB of RAM, and a…
Oh yes. IBM, it seems, was late in getting the news Microsoft wasn't as committed to OS/2 as IBM was. I imagine what was said at IBM's top management back then.
What Microsoft saw that IBM didn't was that having a continuous path without breaks was more valuable than having an objectively better product. That's also why we are using PCs today and not Amigas, Ataris, or Suns.
At least we can also use NeXT machines. ;-)
Earlier quoted context omitted.
Are you talking about Microsoft/IBM OS/2, i.e. 1.x, or IBM-only OS.2, i.e. >=2.x here? For 2 and above, the IBM only versions, well, yes, kinda sorta. But that was too late. The game was already over. OS/2 2.0 came out in April 1992. Windows 3.0 came out in May 1990, 2 whole years earlier. It already had established an ecosystem before 32-bit OS/2 appeared. Secondly, OS/2 2 really wanted a 386DX and 4MB of RAM, and a…
> Windows 3.0 came out in May 1990, 2 whole years earlier\ Oh yes. IBM, it seems, was late in getting the news Microsoft wasn't as committed to OS/2 as IBM was. I imagine what was said at IBM's top management back then. What Microsoft saw that IBM didn't was that having a continuous path without breaks was more valuable than having an objectively better product. That's also why we are using PCs today and not Amigas,…
In my then-job, my boss sent me on a training course for 3Com's new NOS, 3+Open, which was based on OS/2 1.0.
I did not realise it was a clever evaluation strategy. He knew I might have enthused about it if given a copy to play with. Instead, being trained on it, I was told some of the holes and weaknesses.
I came back, told them it was good but only delivered with OS/2 clients and had no compelling features for DOS clients, and they were very pleased -- and the company went on to start selling Novell Netware instead.
Looking back that was a good choice. In the late 1980s Netware 2 and 3 were superb server OSes for DOS, and OS/2 with LAN Manager wasn't.
But yes, I think from soon after OS/2 launched, it was apparent and widely reported to IBM that MS was not happy, and as soon as MS started talking about launching a new updated version of Windows -- circa 1989 -- it was very clear to IBM that MS was preparing a lifeboat and would soon abandon ship. By the time Windows 3.0 came out MS had left the project. Its consolation prize was to keep Portable OS/2, also called OS/2 3.0, later OS/2 NT.
This wasn't secret, and everyone in the industry knew about it. The embarrassment to IBM was considerable and I think that's why IBM threw so many people and so much money at OS/2 2.x. It was clear early on that although it was a good product, it wasn't good _enough_.
NT 3.1 launched in July 1993, almost exactly 1 year before OS/2 2.1, and NT made it pretty clear that OS/2 2.x was toast.
I deployed NT 3.1 in production and supported it. Yes, it was big and it needed a serious PC. OS/2 2.0 was pretty happy on a 486 in 4MB of RAM and ran well in 8MB. NT 3.1 needed 16MB to be useful and really wanted a Pentium.
But NT was easier to install. For instance, you could copy the files from CD to hard disk and then run the installer from MS-DOS. OS/2 had to boot to install, and it had to boot with CD drivers to install from CD. Not trivial to achieve: ATAPI CD-ROM drives hadn't been invented yet. It was expensive SCSI drives and a driver for your SCSI card, or proprietary interfaces and proprietary drivers, and many of those were DOS-only.
NT didn't have OS/2's vast complicated CONFIG.SYS file. It had networking and sound and so on integrated as standard, while they were paid optional extras on OS/2.
And NT ran Windows 3 apps better than OS/2, because each Windows 3 app had its own resource heaps under NT. Since the 64kB heap was the critical limitation on Win3 apps, NT ran them better than actual Windows 3.
If you could afford the £5000 PC to run NT, it was a better OS. OK, its UI was the clunky (but fast) Windows 3 Program Manager, but it worked. OS/2's fancy Workplace Shell was more powerful but harder to use. E.g. why on earth did some IBMer think needing to use the right mouse button to drag an icon was a good idea?
I owned OS/2, I used it, and I liked it. I am still faintly nostalgic for it.
But aside from the fancy front-end, NT was better.
NT 3.5 was smaller, faster and better still. NT 3.51 was even smaller, faster and stabler than that, and was in some ways the highpoint of NT performance. It ran well in 8MB of RAM and very well in 16MB. On 32MB of RAM, if you were that rich, we NT users could poke fun at people with £20K-£30K UNIX workstations, because a high-end PC was as fast, as stable, and had a lot more apps and a much easier UI.
Sad to say, but the writing was on the wall for OS/2 by 1989 or so, before 2.0 even launched. By 1990 Windows 3.0 was a hit. By 1992 Windows 3.1 was a bigger hit and by 1993 it was pretty clear that it was all over bar the shouting.
There was a killer combination that had a chance, but not a good one: Novell Netware for OS/2. Netware ran on OS/2 and that made it a useful non-dedicated server. IBM could have bought Novell, combined the products, and had a strong offering. Novell management back then were slavering to outcompete Microsoft; that's why Caldera happened, and why Novell ended up buying SUSE.
(For whom I worked until last year.)
OS/2 plus Netware as a server platform had real potential, and IBM could have focussed on server apps. IBM had CC:Mail and Lotus Notes email, it had the DB2 database, soon it would have Websphere. It had the products to bundle to make OS/2 a good deal as a server, but it wanted to push the client.
Earlier quoted context omitted.
The PS/2 was PC compatible. The term means "can run IBM PC software". The PS/2 could. It was compatible. Same OSes, same apps. Remember this was the 1980s. There were almost no protect-mode 32-bit OSes around, and the few that there were then had PS/2 versions. DOS and DOS apps ran fine on PS/2s, unmodified.
No it wasn't, it had the MCA architecture for the motherboards, which was out of reach to the clone makers. https://en.wikipedia.org/wiki/Micro_Channel_architecture > It was thought that by creating a new standard, IBM would regain control of standards via the required licensing. As patents can take three years or more to be granted, however, only those relating to ISA could be licensed when Micro Channel was announc…
The term "PC compatible" is about software, not hardware. It's not about slots. For instance, laptops were and are still PC compatibles and 99% of laptops have no expansion slots.
The PS/2 range ran bog-standard unmodified PC DOS and MS-DOS, and unmodified DR Concurrent DOS. It ran plain unmodified standard Windows 2, 3 and 3.1.
I put it on them. I deployed them in production. I have personally done this; this is not research or hearsay. I did it myself. In my basement right now is a PS/2 Model 80 with 16MB of RAM that runs Windows NT Server 3.51.
PS/2s were PCs.
Furthermore MCA was not "out of reach to clone makers".
I personally installed at least one Apricot VXFT server with SCO UNIX, and MCA slots.
http://insight.actapricot.org/insight/products/servers/vxft/...
There were other models too:
https://ardent-tool.com/Apricot/
Tandy, ACT Apricot, Fujitsu, Siemens-Nixdorf, Olivetti, AST, and others all sold MCA machines.
It cost money for a licence, sure, but it was not IBM-only.
The list of members was not insignificant:
https://en.wikipedia.org/wiki/Micro_Channel_Developers_Assoc...
MCA lost out to VL-Bus, EISA, and ultimately to PCI, but it did OK for a while.
There wasn't really anything wrong with MCA. Its issues were two fold and relatively minor:
[1] EISA, VLB and PCI could co-exist with ISA. MCA couldn't. [2] Manufacturers had to pay to licence it.
If IBM's OS/2 plan had worked, and OS/2 had been a 386 OS and thus been compelling to DOS users, then in 1987, the PC industry would have had a pre-emptive multitasking OS, and by 1988, it would have had a GUI too.
Windows 3 would never have happened. There's a chance that 386BSD might not have happened. Linux almost definitely wouldn't -- it was 3 years later.
Then IBM's package of a fancy new PCs with shiny new ports and graphics and memory, with a fancy multitasking OS, plus a fancy new 32-bit bus, would have been compelling.
But it dropped the ball. It insisted on the 286 being included.