How the OS/2 Flop Went On To Shape Modern Software (theregister.com) 116
"It's fair to say that by 1995, OS/2 was dead software walking," remembers a new article from the Register (which begins with a 1995 Usenet post from Gordon Letwin, Microsoft's lead architect on the OS/2 project).
But the real question is why this Microsoft-IBM collaboration on a DOS-replacing operating system ultimately lost out to Windows...? If OS/2 1.0 had been an 80386 OS, and had been able to multitask DOS apps, we think it would have been a big hit.... OS/2's initial 1980s versions were 16-bit products, at IBM's insistence. That is when the war was lost. That is when OS/2 flopped. Because its initial versions were even more crippled than the Deskpro 386...
Because OS/2 1.x flopped, Microsoft launched a product that fixed the key weakness of OS/2 1.x. That product was Windows 3, which worked perfectly acceptably on 286 machines, but if you ran the same installed copy on a 32-bit 386 PC, it worked better. Windows 3.0 could use the more sophisticated hardware of a 386 to give better multitasking of the market-dominating DOS apps...
IBM's poor planning shaped the PC industry of the 1990s more than Microsoft's successes. Windows 3.0 wasn't great, but it was good enough. It reversed people's perception of Windows after the failures of Windows 1 and Windows 2. Windows 3 achieved what OS/2 had intended to do. It transformed IBM PC compatibles from single-tasking text-only computers into graphical computers, with poor but just about usable multitasking...
Soon after Windows 3.0 turned out to be a hit, OS/2 NT was rebranded as Windows NT. Even the most ardent Linux enthusiast must c\oncede that Windows NT did quite well over three decades.
Back in 1995, the Register's author says they'd moved from OS/2 to Windows 95 "while it was still in beta.
"The UI was far superior, more hardware worked, and Doom ran much better."
But the real question is why this Microsoft-IBM collaboration on a DOS-replacing operating system ultimately lost out to Windows...? If OS/2 1.0 had been an 80386 OS, and had been able to multitask DOS apps, we think it would have been a big hit.... OS/2's initial 1980s versions were 16-bit products, at IBM's insistence. That is when the war was lost. That is when OS/2 flopped. Because its initial versions were even more crippled than the Deskpro 386...
Because OS/2 1.x flopped, Microsoft launched a product that fixed the key weakness of OS/2 1.x. That product was Windows 3, which worked perfectly acceptably on 286 machines, but if you ran the same installed copy on a 32-bit 386 PC, it worked better. Windows 3.0 could use the more sophisticated hardware of a 386 to give better multitasking of the market-dominating DOS apps...
IBM's poor planning shaped the PC industry of the 1990s more than Microsoft's successes. Windows 3.0 wasn't great, but it was good enough. It reversed people's perception of Windows after the failures of Windows 1 and Windows 2. Windows 3 achieved what OS/2 had intended to do. It transformed IBM PC compatibles from single-tasking text-only computers into graphical computers, with poor but just about usable multitasking...
Soon after Windows 3.0 turned out to be a hit, OS/2 NT was rebranded as Windows NT. Even the most ardent Linux enthusiast must c\oncede that Windows NT did quite well over three decades.
Back in 1995, the Register's author says they'd moved from OS/2 to Windows 95 "while it was still in beta.
"The UI was far superior, more hardware worked, and Doom ran much better."
I was really hoping OS/2 would take off (Score:5, Interesting)
Re: (Score:3)
Not the 16 bit versions, they were a pain.
16-bit OS/2 1.x was an improvement over DOS. While the 286 still limited you to 16-bit addressing it did provided protected mode operation, some memory management improvements, and a better filesystem. Which led to OS/2 1.x finding a little bit of a following in industrial settings.
Re: (Score:2)
Yet, OS/2 1.x cannot run Doom.
Re: (Score:2)
Which is odd because any app under DOS is 99.9% using the BIOS once it gets loaded. But OS/2 was seen as a professional system, so the time to get Doom to work on it wasn't seen as worth it given the market for OS/2 games was very low.
Re:I was really hoping OS/2 would take off (Score:4, Interesting)
OS/2 1.x was not that bad. It was a very robust server OS, for example. It's telling that Microsoft's own Lan Manager server was based on OS/2 1.x (instead of Xenix) for years, before the transition to Windows NT.
Windows/NT ! From the makers of edlin! (Score:5, Interesting)
Unmentioned in the article is how OS/2 was IBM, and Windows was marketed to everyone and also IBM.
Back in the that day IBM was considered the de-facto standard of everything and they were very proud of it. They were very used to dictating what the market would buy whether they liked it or not, (IBM 360/20 anyone?) protected everything with patents, and expected the customer base to be nice and obedient and docile.
Naturally they were unable to adapt the the multiple changes in the market and it is anyone's guess as to whether they could even see those changes in advance much less respond to them. Bill Gates was there to pick up the pieces and you know the rest of the history.
Re: (Score:2)
Re: (Score:3)
But the original name for Windows NT was... wait for it... OS/2 3.0.
Microsoft decided to rebrand it Windows NT because it was designed to be portable and compatible with OS/2 and POSIX and support multiprocessing. They later ported it to many other hardware platforms, including 64 bit DEC Alpha, which was the first mass produced 64 bit chip. The biggest issue was 32 bit compatibility, but if the app stuck to the Windows API, it usually ran fine (barring bugs). The biggest problem was many x86 apps had optim
Re:Windows/NT ! From the makers of edlin! (Score:5, Informative)
It's not just that they failed to adapt. At a time when it was probably mostly, but not entirely over, Microsoft effectively blackmailed IBM's software division into stopping marketing OS/2 to end users by threatening to stop selling Windows licences to IBM's hardware business. Something that only came out by chance many years later in another major antitrust case against Microsoft.
Re:Windows/NT ! From the makers of edlin! (Score:5, Interesting)
This was the end of the line for Gates's frustration with IBM, as OS/2 took resources from projects he and Balmer were convinced would take off. Publicly claiming that upcoming Windows 3.0 would not be "Presentation Manager Lite", MS still death-marched developers to produce the release, while devs allotted to IBM sat on their hands or did code reviews for IBM managers. Win 3.0 Program Manager kicked ass on Presentation Manager, it was definitely not "lite" - and it ditched all the heavy-baggage of IBM SNA requirements.
"OS/2 NT" is a bit misleading. Late in the endgame of the IBM/MS relationship, Gates discovered that Dave Cutler was being cut away from DEC, with a recalibration of Prism and the future of Alpha. Cutler had begun a 64-bit microkernel evolution of his VMS system. OS/2 3.0 was on the boards, still dragging MS resources and tying up IP. Gates hired Cutler to build an alternative, skunk works kernel from his Prism design work, with the hope of porting the Windows System 32 layer with dependencies etc. When the last bitter contract work was delivered for IBM, Cutler and the Windows team ground out the hard work of delivering their kernel, TCP stack, and Windows 3.11 port —Windows NT.
Most of this stuff is well-covered in Carroll's "Big Blues" along with Zachary's "Show Stopper!: The Breakneck Race to Create Windows NT". I had a small part at the NT launch in Moscone Center, working for a ghost-writer on the Sybex NT book that launched at that event
Re: Windows/NT ! From the makers of edlin! (Score:2)
That sounds like bad planning.
Re: Windows/NT ! From the makers of edlin! (Score:4, Funny)
Re: (Score:3)
SNA baggage in OS/2? What is this baggage? OS/2 1.x itself has no SNA baggage, unless we're talking about IBM's OS/2 1.x Extended Edition, that came with lots of bundled extra IBM stuff. The MS OS/2 1.3 used in Lan Manager in particular, has no SNA baggage, except some pieces, known as Netbios, NetBEUI, 802.2 LLC, SMB, DLC and others, but they all were also present in Windows NT later.
Microsoft was an advocate for OS/2 1.x, not Window (Score:5, Informative)
Unmentioned in the article is how OS/2 was IBM, and Windows was marketed to everyone and also IBM.
The summary certainly botched the history: "Because OS/2 1.x flopped, Microsoft launched a product that fixed the key weakness of OS/2 1.x." The truth is that Microsoft was an advocate for OS/2 1.x, not Window. According to both IBM and Microsoft, DOS users were supposed to upgrade to OS/2 1.x. And on OS/2 1.x there would be a GUI called Presentation Manager. Windows, ACCORDING TO MICROSOFT, was a stop gap measure that would allow DOS users to upgrade to DOS plus the Windows GUI. Windows and Presentation manager APIs were nearly identical, making it easy for developers to target both. So apps should be available for both.
The idea was that the upgrade path for early adopters would be:
DOS -> OS/2 + PM
and for the laggards:
DOS -> DOS + Windows -> OS/2 + PM
At lease for the 16-bit world. IBM was working on the Intel based OS/2 2 that would be 32-bit. And in parallel MS was working on the cross-platform 32-bit OS/2 NT.
When sales of Windows 3 took off MS decided to "divorce" from IBM and go their own way. With Windows being the real 16-bit destination not OS/2 + PM, and OS/2 NT would be renamed Windows NT.
And to f*ck up the competition, 32-bit OS/2 2.0, MS canceled its plans to provide the software development tools (compilers, debugger, etc) for OS/2 2.0. So IBM had an OS/2 but no development tools for developers.
This bought MS time to announce the new stop gap measure, Windows 95.
Now the early adopter path was:
DOS -> DOS + Win3 -> WinNT
and the laggard path was:
DOS-> DOS + Win3 -> Win9x/ME -> WinNT
These two paths merged at WinXP.
Re: (Score:2)
At lease for the 16-bit world. IBM was working on the Intel based OS/2 2 that would be 32-bit.
Microsoft was given the task to develop the 32-bit OS/2 2.0, while IBM was creating OS/2 1.3 .When the partnership dissolved, as a part of the divorce Microsoft handed over what they had done with OS/2 2.0 to IBM, while retaining the right to use OS/2 1.3 . And they did use it for several years more for their server products, until Windows NT took off.
Re: (Score:2)
At lease for the 16-bit world. IBM was working on the Intel based OS/2 2 that would be 32-bit.
Microsoft was given the task to develop the 32-bit OS/2 2.0, while IBM was creating OS/2 1.3 .When the partnership dissolved, as a part of the divorce Microsoft handed over what they had done with OS/2 2.0 to IBM, while retaining the right to use OS/2 1.3 . And they did use it for several years more for their server products, until Windows NT took off.
Perhaps I should have written "Intel only" rather than "Intel based". There were two 32-bit OS/2. One IBM and one MS. IBM's was Intel only, OS/2 2. MS' was portable to different CPU architectures, during development MIPS and Intel, OS/2 NT. Which in its WinNT 4 incarnation would add PowerPC and Alpha. And later 64-bit Intel and 64-bit ARM. OS/2 NT was to be the upgrade from OS/2 2.0.
Re: Microsoft was an advocate for OS/2 1.x, not Wi (Score:2)
My point was that before the divorce mainly Microsoft developed 80386-based OS/2 2.0 . Until the spring of 1991, all known builds of OS/2 2.0 were from Microsoft. After that, Microsoft left the ship, and IBM took over.
The bottom line is that OS/2 2.0 was designed by Microsoft, at least in the beginning.
OS/2 NT was also known as OS/2 3.0; I made no comments about it.
Re: (Score:2)
In short, IBM and MS decided to run the 2.0 and 3.
Re: (Score:2)
MS actually got pretty far with OS/2 2, see https://www.os2museum.com/wp/t... [os2museum.com] and https://www.os2museum.com/wp/m... [os2museum.com]
Re: (Score:2)
Notice in your link how much of t
Re:Windows/NT ! From the makers of edlin! (Score:4, Interesting)
Bill Gates' moment of brilliance was when he realized that "good enough" really was good enough for most customers. While everyone else was trying to create the perfect OS, the perfect office suite, or the perfect DB, Gates realized that customers needed systems that worked adequately most of the time that a technophobe could use, and they needed them now. No, Windows wasn't perfect, but it was good enough and priced at a not-unreasonable percentage of the price of a new desktop machine.
Re: (Score:2)
customers needed systems that worked adequately most of the time that a technophobe could use
Don't confuse Bill Gates with Steve Jobs.
Re: (Score:3)
The genius was in using Microsoft's "in" with the trade magazines to float vaporware. I remember "screenshots" of Chicago the better part of a year before Windows 95 was released, almost certainly renderings and not actually screenshots of a live system. The whole idea was to convince developers and consumers not to jump ship to OS/2 2.1, while Microsoft's engineers desperately tried to get Chicago to actually run.
And it worked. IBM can take some of the blame for shit marketing, but the fact was that OS/2 2
Re: (Score:2)
I did Win95 support for MS when it was still new, and can say with some authority that 90% of issues were (L)user-generated. Can't really speak to networking issues, since most early adopters were individuals rather than corporate users.
Re: (Score:2)
I had the networking croak on both LAN adapters and on PPP. It was probably a lot to do with the way that TCP/IP was implemented in Windows 95 as a 16-bit subsystem, largely taken from BSD source. Even OSR2 didn't fully resolve it, and it wasn't until Windows 98 that the problem disappeared. Later service updates did fix some the other stability issues, but as with everything, it wasn't until Windows 98 that most of the big problems were finally fixed.
Re: (Score:2)
I went from managing Win 3.11 machines directly to NT 4.0 on a LanMan domain, so missed that fun. Then Active Directory and Win2k changed the game entirely, and "plug and pray" became actually functional.
Re: (Score:2)
IBM was short sighted, it was a mainframe company and was never all that happy with microcomputers (or minis). It was very business oriented, and so to them the PC was supposed to be the low end system that could interface with mainframes (a "smart" 3270 sort of), or at the least be small enough to not compete with its own bigger systems. It had its own line of 32 bit systems running a range of sizes.
Really, at the time of OS/2 the 32 bit market was pretty large, outside of the PC world. Ie, Unix worksta
Re: (Score:2)
They were very used to dictating what the market would buy whether they liked it or not, (IBM 360/20 anyone?
In 1966? Nah, this was way before the peak of IBM's arrogance. Yes, IBM was the biggest manufacturer of computing tech, but its way wasn't without risks yet. For example, about the same time delays and "development hell" of OS/360 almost killed the company.
The success of System/360 made IBM way more arrogant. For example, the 1970 release of System/370 models 155 and 165 without memory virtualization was a very arrogant decision, because "Our flock... pardon us, customers, of midrange machines don't need no
Re:Windows/NT ! From the makers of edlin! (Score:5, Informative)
> Unmentioned in the article is how OS/2 was IBM,
I wrote the article.
I did not mention this because it was not true.
OS/2 was not an IBM product. It was _both_ an IBM _and_ a Microsoft product. Both wrote it, both marketed it.
Allegedly, in v1.x, MS did most of the work and IBM paid. But I can't prove that. If anyone can, let me know.
Re: (Score:2)
I would really like you or anyone else come up with a magazine ad (Info World, PC Mag, etc) by anyone but IBM which featured OS/2 with their PCs. A don't recall a one. I don't recall a single trade show booth that did either.
And at that time I was developing and selling motherboards to the myriad screwdriver-tech companies. Nobody ever asked me to certify that OS/2 worked but it was a different story with Windows. We did test it, however, because IBM.
The market in general and end-users in particula
I saw OS 2 back in the day. (Score:2)
Re: I saw OS 2 back in the day. (Score:2)
The article says that os2 had premptive multi tasking.
Re: (Score:2)
Re: I saw OS 2 back in the day. (Score:2)
OP sais "as I remember" I was just confirming for his benefit that he remembered right.
Not everyone on slashdot is a jackass! (Me only sometimes.)
Backward compatible? (Score:4, Funny)
It was. But not anymore!
They pleaded with me, begged me to install Windows 11 on a processor before the Intel 8 series.
But I couldn't! (starts sobbing) I just, couldn't.
Re: (Score:2)
Oh if I had the mod points...
Re: (Score:3)
It was crisp, and very nice. It had real pre-emtive multitasking as I remember. As opposed to Windows 3.1. I know DOS, and know what Microsoft did there.... they kept building on their shit, and everything is still backwards compatible. I guess most people don't like change.
Actually, it's IBM that is fanatical about backwards compatibility. Things may have changed since I was last an IBM customer, but 20 years ago, anyway, IBM supported every old version of their mini and mainframe OSes and required new hardware to support those old OSes as they released it. It's Microsoft that is the one that retires OSes (etc.) and requires users to upgrade (for a price), and which obsolesces not only software, but also hardware as their newer versions aren't backwards-compatible to older ir
Re: (Score:2)
IBM supported every old version of their mini and mainframe OSes and required new hardware to support those old OSes as they released it. It's Microsoft that is the one that retires OSes.
No, MS and IBM are not that different in that aspect. Windows 11 is still the same OS as Windows NT from 1993, just a newer version. Just like z/OS is the same OS as MVS/370. But guess what, you can't install ancient MVS/ESA on a modern z15 mainframe, just like you can't install Windows NT 4.0 on a PC with Ryzen CPU.
Re: (Score:2)
Yet, you can install OS/2, or rather an OEM version of OS/2 (ArcaOS) on a new Ryzen CPU (only 64 cores are supported, with 32 tested), as long as the frame buffer is accessible in the lower 4GB. True that some stuff such as wireless won't work and you're stuck with one monitor and running at 3840x2160 might leave you without enough memory to do much, but it does run, can be installed onto GPT disk or nVME and even supports secure boot with minimal work.
Re: I saw OS 2 back in the day. (Score:2)
Backwards compatible means different things to different people. To IBM it meant O/S running on old hardware. To Microsoft, it means old apps running on the latest Windows. Even for a minor Windows revision, it can mean a trip to the computer store for new hardware.
And Microsoft "backwards compatibility" certainly doesn't extend to the user. We changed your UI around. Your responsibility to figure out how we redefined your keyboard functions.
Re: (Score:2)
Windows 3.1 had preemptive multitasking too, for DOS programs. In fact, its predecessor Windows/386 2.01 had it too in 1987.
As for backward compatibility, current OS/2 still can run DOS and Win16 applications, with excellent compatibility. Current Windows can't do that.
Re: (Score:2)
[Article author here]
> Windows 3.1 had preemptive multitasking too, for DOS programs. In fact, its predecessor Windows/386 2.01 had it too in 1987.
And 3.0, and Windows 2 on all platforms, 8086 and 80286 and 80386.
The thing is that on 8086 and 80286, all your DOS apps had to fit into the same shared 640kB -- which meant it wasn't much use.
> As for backward compatibility, current OS/2 still can run DOS and Win16 applications, with excellent compatibility. Current Windows can't do that.
A very good point.
Re: (Score:2)
The thing is that on 8086 and 80286, all your DOS apps had to fit into the same shared 640kB -- which meant it wasn't much use.
Windows 3.1 on a 80286 allowed DOS programs to use XMS and DPMI, which allowed them vastly more memory than 640Kb. Windows 3.0 on a 8086 allowed DOS programs to use LIM EMS to the similar effect. Of course, there is no preemptive multitasking on a 80286 or 8086.
A very good point.
Of course this is because OS/2 and eComStation and ArcaOS are all 32-bit OSes for x86-32. Sadly x86-64 does not include Virtual86 mode any more, so they can't easily run DOS and DOS apps without emulation.
Virtualization extensions from AMD and Intel bring back the V86 mode, or something very similar.
Re: (Score:2)
Well, you do get a ramdisk out of the upper memory on ArcaOS thanks to PAE.
Re: (Score:2)
It was my main driver for about two or three years. I loved the object oriented GUI, and with EMX I had natively-compiled *nix commands. REXX was also one helluva a kickass scripting language, but I had bash as well. It had its flaws, the single message queue for PM could cause lock ups.
Re: (Score:2)
Also even on NT it was a bit odd. The kernel was nice, but then the application layer was put on top to hide a lot of its usefulness. Ie, there was a POSIX layer too, created primarily so that you could market it to the government who mandated POSIX support. But you could not do anything useful in POSIX really, because all the useful stuff was done by WIN32; so no networking, no GUI, no devices other than basic text and file based I/O (in a window running under WIN32 layer). That is, the kernel didn't do
OS/2 was hindered by IBM in sooo many ways... (Score:5, Interesting)
I worked at a computer dealership from 1988 to 1994.
First, we had a tool to connect to IBM systems to order parts for PCs (and other stuff from IBM). This software worked under windows (and DOS maybe). It didn't work under OS/2. After a couple of months I asked our rep when OS/2 would be coming around and he said "we have no plans to support this on OS/2". Reminding him that OS/2 was an IBM project didn't really seem to matter.
Secondly, IBM had some diagnostics that you could run to to test various parts of a pc, such as memory. Well, it appeared that OS/2 exercised memory more vigorously than other software. Memory that 'passed' diags would fail while running OS/2, and if passed diags you couldn't make a warranty claim. The only reason that this got addressed was some important companies started spec'ing in the quote products that the hardware had to run OS/2 before it would be considered good.
Wow. What a mess
ecalkin
Re: (Score:2)
Trueb about the RAM. On low quality RAM even the setup of OS/2 warp crashed.
Re:OS/2 was hindered by IBM in sooo many ways... (Score:5, Interesting)
We had OS/2 systems in our high school for a year or two. One of the features was that any programs you had running when you shut down would automatically restart after boot. One of the flaws was that the machine would crash and restart if you ran out of memory. To break a machine, you had to 1) select all of the icons on the desktop 2) hit enter. It would be trapped in a crash/reboot cycle. The only way out was to frantically close apps as they opened after each boot until the machine crashed again, eventually getting the memory usage below the crash level after multiple iterations.
You can bet students figured that one out pretty fast.
Re: (Score:2)
That does not sound familiar, but this was over 30 years ago...
ecalkin
IBM MicroChannel would of better with out the fees (Score:2)
IBM MicroChannel would of better with out the fees that they put on it to kill clones.
Re: IBM MicroChannel would of better with out the (Score:2)
Yes. EISA never took off, also. VLB had it's moment. Then finally PCI and later PCIe supplanted them.
Unfortunately Bill Gates took IBM to shool. (Score:2)
Schrodinger's Fortune 500 Corporation (Score:3)
IBM exists and doesn't exist at the same time?
When I knew OS/2 was toast (Score:3)
That was when I knew OS/2 was a dead OS walking.
Re: (Score:3)
Re: (Score:2)
It was a brilliant strategy and it worked.
Re: (Score:3)
I recall reading about how great OS/2 was, so reliable, so well-designed, yadda yadda. And then I saw IBM machines that came bundled with Windows. And I thought: if IBM won't put their weight behind their own system, then who will?
Re: (Score:2)
IBM at the time was (maybe still is) a bunch of fiefdoms. The PC fiefdom was focused on Windows.
Memory shortages (Score:5, Informative)
There were RAM shortages from the mid to late 1980s to mid 1990s. A 32-bit OS consumes significantly more RAM than a 16-bit one. I ran OS/2 2.0 in 1992. It could walk in 4MB of RAM but you really wanted 16MB for it to work, and that is what I bought, at great cost, with the proceeds of my first programming contract as a teenager. It was a 486 also. Most computers came with only 1MB of RAM back in those days. Many motherboards could not even accommodate 16MB.
Because of these shortages, I don't think that releasing 32-bit OS/2 years earlier would have changed much. The reasons for the failure of OS/2 in the marketplace are elsewhere. The divorce between MS and IBM was a messy one. MS knew how to market software, and IBM did not.
I kept running OS/2 as my primary desktop OS all the way until 2007, at which point I gave up and switched to 64-bit Windows.
I was running a Warp 4.5 VM under Virtualbox just a few days ago, looking to open source some old code that was in the virtual hard drive.
Re: (Score:3)
The only real difference is the pointer size, so of course 32bit pointers will be bigger than 16 bit ones - but...
AmigaOS was a 32bit OS, and could run a multitasking gui in 0.5mb of ram. There is nothing inherent in a 32bit os that makes it require 4mb of ram, that's just the way these particular systems were designed.
A true 16bit OS can only address 64kb of ram, if you want to address more than that you have to start using hacks like segmentation to split your address space into 64kb chunks, and these hac
Re: (Score:2)
Yes, I know all this. The 80386 instruction set uses longer opcodes, though, compared to the Motorola 68k series. A lot of data OS structures tend to move from 16 to 32 bits, also, or get padded on 32 bits at least. Even if you were to write 32-bit DOS, it would still consume more memory than the 16-bit version. Not necessarily a whole lot more. There are 32-bit DOS extenders that typically consume 10-20KB of RAM. Of course, they don't provide 32-bit versions of all the DOS functions/programs.
While AmigaO
Re: (Score:2)
> AmigaOS was a 32bit OS, and could run a multitasking gui in 0.5mb of ram.
I'd call the classic Amiga and AmigaOS a 16-bit OS myself. Fully 32-bit 680x0 chips did not arrive until later (the 68020) and memory management hardware until later still (68030) -- and AmigaOS cannot use any of those extra features.
[Disclaimer: I own an A1200 with a 68030 accelerator in it, but I am no Amiga expert.]
> There is nothing inherent in a 32bit os that makes it require 4mb of ram
Very true.
But other x86 OSes were as
Re: (Score:2)
I started running Warp 3 on a 386/33 with 4MB's of ram. With tuning, including not loading the WPS, it ran pretty good. OTOH, my brother bought Warp 3, installed everything on a 4MB machine and it was useless as it spent all its time swapping.
Re: Memory shortages (Score:2)
The Workplace Shell was an excellent UI, so you missed out. But it was definitely RAM hungry. It was far ahead of its time, with its object oriented nature, something that still hasn't been surpassed in any MS offering, or Linux desktop offering.
Too bad Presentation Manager suffered from the single input queue problem.
IBM (Score:2)
IBM really effed up back then. They could not to do PCs, there was too much internal conflict and internal resource-wars with the VPs in charge of mainframe hardware and software. They also couldn't scale their marketing from being able to sell to big corporations vs smaller companies. And also, internally it was really hard to get anything done. Someone figured out that it would take IBM one year to ship out an empty box with no product in it (one side effect of that was that the IBM PC was clonable becaus
Re: (Score:2)
I was there. It was the software. (Score:5, Informative)
I was there. I wrote software for Windows 3.0. I had an OS/2 development system. The OS/2 coordinate system was not compatible with Windows. OS/2 was not compatible with the Windows Graphic Device Interface. We would have had to have forked the code and separately supported the OS/2 code. At that time Microsoft had an awesome evangelical program to help sell Window software. IBM did not. IBM had no established distribution method to sell OS/2 software. There was a very small market for OS/2 software. We developers looked at money selling Microsoft Windows apps, or no money selling OS/2 apps. We made our choice. And then IBM canceled OS/2.
Re: (Score:2)
Re: (Score:3, Insightful)
Although Bill Gates was involved in monopolistic practices, Windows had software distribution that could make software developers money. Sleezy practices helped sell Windows copies but so did a plethora of supporting software. This is also why VHS beat Betamax.
Re: (Score:2)
Re: (Score:3)
- Release mismanagement -- IBM was making major changes to internal kernel structures during patch releases which would result in IBM's own code breaking itself
- Insistence on assembler -- IBM required Microsoft to write all OS/2 code in assembler
- Diagnostic tools where internal IBM proprietary only. -- You had to "know someone" to get your hands on diagnostic tools
- "Red Book" AP
A shame (Score:3)
Windows 386 was the turning point (Score:5, Insightful)
I had Windows 1.0, which was pretty useless. Windows 2.0 was only a tiny bit better. But when they introduced Windows 386 (based on 2.1, IIRC) it became pretty useful. It had a protected mode kernel, and I could run multiple DOS apps in separate windows. It started selling pretty well, and a ex-Microsoft guy told me it really got their attention, and they realized they might not need IBM after all. They put a huge effort into Windows 3.0 and began to lose their drive for OS/2. IBM was making it hard for MS because they knew MS wanted to sell OS/2 to everyone and IBM wanted some features for themselves to help push IBM hardware.
Using Windows 386 as a DOS task switching platform created the opening for Windows apps. Being able to run Ami Pro (word processor), Micrografx (drawing) and later Excel was a bit of a game changer. Before, I waited months for drivers to appear for a graphics board I got for AutoCAD, and had to run 640x480 for Word Perfect, Lotus 123, etc. After about 6 months, I could run all three at 1024x768. Then I got a NEC Laser printer and waited months for drivers. That was pretty painful. With Windows 3.0, once I had drivers, all the apps worked. That sold me. When Pagemaker arrived, it worked perfectly with my graphics card and drivers. All I needed for hardware upgrades was Windows drivers and AutoCAD (DOS) drivers. Life got so much simpler. I read Gordon Letwin's book "Inside OS/2" and figured when the DB-based file system arrived, it would win and easily beat Windows. That file system never arrived and there was never a reason to leave Windows. But MS seems to trying to give us lots of reasons today.
Re: (Score:2)
> Then I got a NEC Laser printer and waited months for drivers. That was pretty painful. With Windows 3.0, once I had drivers, all the apps worked.
I read a story that the OS/2 DDK (Driver Development Kit) cost in excess of $1000; while you couldn't go anywhere near a Microsoft booth at any tech conference without a Windows DDK being thrown at your direction.
No wonder Windows had drivers for more hardware.
Re: (Score:2)
Not just the DDK, 32 bit compilers were similarly priced, though the DDK used old MS compiler/assemblers. When I got interested in OS/2 towards the end, the DDK was free though you had to register.
Re: (Score:2)
I had Windows 1.0, which was pretty useless. Windows 2.0 was only a tiny bit better. But when they introduced Windows 386 (based on 2.1, IIRC) it became pretty useful. It had a protected mode kernel, and I could run multiple DOS apps in separate windows. It started selling pretty well
Windows/386 2.01 was the first released version of Windows/386, and it predates all versions of Windows 2.x . (Actually Windows 2.x is little different from a Windows/386 without its WIN386 kernel). It had a "protected mode kernel", but no memory protection, so it wasn't that useful. And no, its sales numbers weren't something to be proud about or something to make you think of ditching IBM.
and a ex-Microsoft guy told me it really got their attention, and they realized they might not need IBM after all.
That didn't happen before they saw what the protected-mode preliminary version of Windows 3.0 was capable of. (Yes, it
Re: (Score:2)
You're probably talking about Windows/286. Windowx 1.x and 2.x had the ability to run on an IBM XT PC, but Windows 2.x
Re: Windows 386 was the turning point (Score:2)
An OS or program that runs on 8086 can also run on 80286? I thought it should be obvious.
And no, the Windows 2.1 a.k.a. Windows/286 2.1 didn't use a special 80286 kernel. It uses the good old real mode Windows 2.x kernel, and its only 80286 feature was the extended memory driver HIMEM.SYS, whose main purpose was to enable the HMA (the first 64K over the first megabyte of memory) for Windows use. The HMA, HIMEM.SYS and 80286 were purely optional, though, and Windows/286 ran quite happily on a 8086.
Re: (Score:2)
OS/2 2.x had an interesting way to do VDM's. They ran in ring 2, could mostly use DOS device drivers like native DOS, things like disk access was trapped by the kernel and used the OS/2 drivers.
It was one of the reasons for a better DOS/Windows. Another was a better file system speeding things up.
OS/2 Warp 3 (Score:3)
I purchase OS/2 Warp 3 back in late 1994 I believe. I loved it and would have continued to use it if IBM would not have dropped the ball on driver support. As time went on it was apparent that IBM had lost interest and I went back to Windows 3.11 for Workgroups after having tried the horrible first iteration of Windows 95 that just kept blue screening.
It was great, but Linux was better (Score:5, Interesting)
I loved OS/2. I am just a humble user and was a starving student at the time it was out, and the current Windows version was Windows for Workgroups. The OS/2 Workplace Shell was amazing! Desktop objects were actual object that could be interacted with via right-click. Everything about it was better than Windows. The guy who gave me a box of floppies to install it told me "It will turn your PC into a supercomputer". That was obviously BS, but it kinda felt that way, especially compared to Windows.
But there were no apps. Or rather, no free, "shareware" or reasonably priced apps that normal people could afford. I guess OS/2 was targeted at Enterprise customers, not consumers,
But then apps ported from Linux started to become available. I'd never heard of Linux, but when at last I got sick of OS/2 crashing and freezing, I switched to Linux and never looked back
It went like this: Windows would freeze and/or crash every day, sometimes multiple times per day, and you would lose hours of work before learning to save after typing ever sentence. OS/2 was much more stable, only crashing/freezing once a week or fortnight. But Linux, even back then, (almost) never crashed. It was free, much more reliable and had tons of free apps!
WTF, IBM?
Re: (Score:2)
But Linux, even back then, (almost) never crashed. It was free, much more reliable and had tons of free apps!
Actually, it took years for Linux (and Windows) to match OS/2 in multitasking while accessing the hard disk. OS/2 never even twitched in that condition, while Linux was a jerky nightmare.
Re: (Score:2)
OS/2 had a trick where the foreground app got a priority and IO boost. Made it feel very smooth
Why it flopped (Score:3)
Gather 'round children and let me tell you a tale lol....
470 billion years ago I had the pre-release and release versions of OS/2 (still do, actually). I installed them both from packs of about 30 3.5" floppies.
The only thing wrong with OS/2 was that there were hardly any PCs capable of running it. I had a somewhat-recent PC/CPU box and OS/2 ran perfectly on it, but it was dog-slow. Dog-fuckin'-slow. A state-of-the-art PC would run it passably, sort of, but anything less than that and it was just slow slow slow.
But OS/2 itself worked perfectly- I had no problem getting 2 DOS programs running in Windows sessions while formatting a floppy in each drive and logging into a BBS, and running some of the OS/2 utility widgets, etc etc, all at the same time. It all ran, it was just maddeningly slow.
The overall experience was excruciating without having bomb-ass hardware (which almost no one could afford). We're talking a $5000 box at the time.
On the other hand, Windows ran acceptably fast on the usual consumer hardware like mine, and at the time it was hard to see what advantage OS/2 had. Yes, OS/2 had *true* preemptive multitasking instead of Windows rudimentary time-slicing shit, but Windows worked well enough. And that was that.
People had a better experience with Windows just because it ran better on current hardware and that's why so many people went to it rather than OS/2.
OS/2 was far superior to Windows in literally every way except speed, and in the end that's what pushed many people to Windows instead of OS/2.
Re: (Score:2)
I started running Warp 3 on a 386/33 with 4MB's of ram. With tuning for memory, including not loading the WPS, it was usually fast unless it started swapping. OS/2 was always memory constrained rather then CPU constrained, really you needed 16MB's of ram when that cost a small fortune. Your system was probably continuously swapping.
Windows and Linux ! (Score:3)
Re: (Score:2)
Re: (Score:2)
eComstation is long dead. If you want to run OS/2 on modern hardware, UEFI, GPT disks/nVME and such, you need to go to https://www.arcanoae.com./ [www.arcanoae.com]
OS/2 lost by design (Score:2)
Re: (Score:2)
Microsoft always planned for OS/2 to lose to window
Not before 1989.
Flop? (Score:2)
So if OS/2 was such a massive flop, and W95 was such a massive success,
why are there still tons of systems (e.g. NYC Underground, ATMs, Airling Booking Systems, and many more) running on OS/2 today,
while all that W95 gave us is a generation of IT support with massive PTSD?
REG shilling for the Microsoft organization (Score:2)
"The demos of OS/2 were excellent, crashing the system had the intended effect -- to FUD OS/2 2.0. People paid attention to the demo and were often suppris
Fry's Electronics sold v.4 as OS/2 WRAP [sic] (Score:2)
Disagree, Microsoft won by... (Score:2)
Least we forget PS/2 (Score:2)
After IBM lost the lead in the desktop pc market they decided to reign in the herd by introducing PS/2, which is a close source standard that you must paid for the privilege of making it. OS/2 was intended to run on PS/2 hardware.
I was an OS/2 user (Score:2)
Back in the day, I was all-in on OS/2 and I still miss it. The shootout video [youtu.be] between OS/2 and NT was a real eye opener, and not just because Microsoft sent a sweaty guy in a suit while IBM sent a gym-bro.
At the company where I worked, I was the only OS/2 user on their Netware server. I ran all the DOS and Win3 stuff without any issues. The OS/2 UI was similar to Windows but better in so many ways, once you learned it.
I know that after a couple of renamings OS/2 is still kinda around today, but I'd like
History later repeated itself (Score:2)
From the article linked to
This reminds me of a recent Slashdot article.
When AMD were working on their Athlon 64-bit processors, the only 64-bit ones Intel were prepared to release were the Itanium line. It was made clear to the Intel en
os/2 Warp (2.0) worked just fine. (Score:2)
Unfortunately, that hit *right* before (fall 1994) Windows (released in '95, with, as PC Mag put it, over 64,000 *known* "issues" (read bugs). And that's when M$ was illegally paying OEMs to bundle.
Re: (Score:2)