Submitted by It was now 1984, and IBM had a different problem: DOS was pretty much still a quick and dirty hack. The only real new thing that had been added to it was directory support so that files could be organized a bit better on the IBM PC/AT’s new hard disk. And thanks to the deal that IBM signed in 1980, the cloners could get the exact same copy of DOS and run exactly the same software. IBM needed to design a brand new operating system to differentiate the company from the clones. Committees were formed and meetings were held, and the new operating system was graced with a name: OS/2.
Fantastic article at Ars Technica about the rise and demise of IBM’s OS/2. OS/2 is one of those big ‘what-ifs’ of the technology world, along the lines of ‘what if Apple had purchased Be instead of NEXT’ or ‘what if Nokia had opted for Android’ (sorry). Our technology world could’ve been a lot different had OS/2 won over Windows 3.x/95.
I reviewed OS/2 as it exists today (eComStation) six years ago.
From what I read in the article it might have meant that IBM had locked up the pc industry. They might have done even more damage than MS did.
Exactly as there is a REASON why OS/2 lost out that everybody seems to forget. Pull up a seat kiddies and let this old codger fill y’all in on some history.
You see kids IBM was PISSED, we are talking finger biting mad that their deal with MSFT, followed by losing the BIOS court case (for those that don’t remember the court ruled that reverse engineering was legal as long as they weren’t copying code and were using COTS parts) meant that anybody and their dog could make a PC. You see kids IBM had gotten used to fat rat profits because “nobody got fired for buying IBM” and suddenly all these upstarts were undercutting them left and right and we are talking hundreds of dollars difference folks.
So first they tried to “wrest control” AKA get back their monopoly status by coming up with a new bus that they would charge like crazy for to any of the OEMs while they of course could use it free, it was called th MCA bus if you want to look it up. The later named “gang of nine” which was the biggest non-IBM OEMs at the time got together instead and came up with their OWN standard which they called EISA and just to rub salt and save money IIRC they used the MCA connectors, simply wired backwards so that IBM couldn’t sue and the hardware manufacturers wouldn’t have to make a new socket to support it.
It was in THIS climate that IBM came out with OS/2 and at a time when the OEMs were paying as little as $15 a pop for Windows/DOS IBM wanted….$200. Remember that with inflation that was probably closer to $600 in USD today and again IBM of course wouldn’t have that boat anchor tied to THEIR PCs. Needless to say it was viewed like plague blankets to the OEMs. Towards the end of the run IBM heavily discounted the price but after trying to backstab the OEMs with the MCA bus frankly there was ZERO love for IBM at that point and besides, why would you want OS/2 when all the software by that point was written for Win9X?
So there ya have it kids,the REAL reason why OS/2 never went anywhere. Back in the day (mid 90s to be exact) I ran OS/2 and it was really nice…until I had to boot into Windows because the software I needed didn’t have an OS/2 version, the same was true of BeOS which I also ran. I said then and I’ll say now if OS/2 had been owned from the start by anybody OTHER than IBM? It would be what Windows is now as it was incredibly stable and fast, but with IBM at the helm it was doomed from the start.
Actually, EISA was a little more than just “MCA wired backwards”. It was a compatible upgrade to the ISA bus, which was the standard before MCA came along.
Do you have any actual source for the claim that Microsoft was pushing windows for $15 a pop?
In any case, the main problem of OS/2 wasn’t cost. That OS was DOA because it never had any defined value proposition against the systems it was supposed to supersede or compete against. Not only that, but OS/2 initially was not a good design or product really.
Yes I do Tyler, its the MSFT DOJ court case. Feel free to look it up but one of the charges they were convicted of is tying their discount prices to whether or not you sold any other OSes on your systems, as well as tying license costs to how many units you sold NOT how many units sold with Windows. These two factors made it virtually impossible to sell a PC with anything else without taking a hell of a hit to the bottom line.
And I’m sorry but you are wrong, I was in PC sales even back then and there was a LOT of value in OS/2 as frankly before Win98SE Windows was buggy as hell. Even Gates himself was famously bitten by the BSOD during a presentation when he plugged in a USB printer only to have the system crash in front of everyone. OS/2 was insanely stable compared to Windows pre XP, folks take a stable running Windows for granted now but back then? Daily BSODs were the norm.
From this article what i get is that if IBM had bought out the Amiga instead of the licensing deals with them (with all the rexx and rt video stuff) They would have had the UI (and in my mind the best in the world at the time by far) and they would also have had the application base if still source compatible with other worskspace applications. If you add the later improvements to the DOS contemtability box (where it finally became a better DOS than DOS) it would be a very potent windows killer.
Many “ifs” there but i think that would have atleast postponed the Microsoft dominance. Windows 1,2 and 3.* could not hold a candle to AmigaOS in any task.
Amiga was a good combination of hardware and software.
Not sure what IBM would have done with Amiga’s hardware design.
Besides I can’t imagine the demoscene graphics community mingling with the boring suits of IBM.
I had an Amiga. I loved my Amiga. But as advanced and as fancy AmigaOS was at the time meant it was really a dead-end when taking into account some features that we now take for granted in a modern system. In a sense that making the system evolve would be highly difficult.
All of AmigaOS was about code running in “user”. No isolation at all. Traversing and modifying liked lists of system structures. Runtime patching of dynamic libraries by a process for the full system. Adding memory protection on top of that without throwing backwards compatibility out would have been really difficult (the only way being probably running virtual machines and integrating clipboards and the like between the virtual machines). On the other hand if there was a company able to do that it was IBM, though.
And as flexible as AmigaOS was at the time, its reliance in doubly-linked lists for many of the system structures (even the filesystem!!!!) would have come back and bitten big time. Not to talk about the security nightmare it could have been (remember that back in the time viruses were quite common in the Amiga).
Just got into eCS homepage and i cannot see any updates from may. Seems like things getting worse for them
Unfortunately the 2.2 release has been stuck in some kind of endless beta programme since seemingly forever.
I have 1.2R running in a VM on my MBP. I ran OS/2 as my primary OS at home right up to and including Warp 4 and I love the OS, but I have to wonder whether Mensys (the company doing the technical work) have the resources to ever push this out. I hope so. If they do, I might well upgrade.
Ah, happy memories of coding in Visual Smalltalk Enterprise on Warp 3. Sigh…
eComStation is updated periodically,
SATA driver, ACPI sub-system, universal video driver were released recently
http://en.ecomstation.ru/commentnews.php?id=2325
Especially when OS/2 Warp came out.
While Microsoft actually showed it products, IBM just had a bunch of commercials staring at a screen and marveling at it. It really failed to interest the General Consumer.
Interestingly enough, this is due to the success of the marking towards apples iProducts. Each commercial shows you how to use the product. vs Others who just have people star at their phone and say how cool it is.
I had forgotten how bad the OS/2 Warp commercials were. Thankfully, YouTube reminded me:
http://youtu.be/G6YZbrhBhjQ
I’m no marketing genius, but I’m pretty sure that one of the most important things in marketing a product is to, you know, feature the product in your advertising.
http://www.os2museum.com/wp/?p=2144
I just read that so-called rebuttal, and, well, I only see him arguing with things because he either didn’t read the article well-enough or he simply has a different opinion about something. Like e.g. the following snippet
The article itself mentions that NT 3.1 never became popular, so I don’t understand why this guy goes on a tangent about it. It’s not a rebuttal of anything the article said about NT in the first place. It was exactly as the article says: stable, multiplatform and fault-tolerant, and yes, not exactly a smashing hit.
The only thing worth mentioning in that “rebuttal” is the mention of Microsoft’s anti-competitive actions, like e.g. forcing OEMs to bundle Windows on PCs and not accepting competitors’ OSes. That should’ve been mentioned in Ars’s original article, but nothing else about this “rebuttal” hits the mark.
What can I say. (Ex)OS/2 users can get very emotional. There’s a lot of bitterness still out there.
Also it feels like I’ve read the same story from the same author. Déjà vu.
In a sense they’re like the Amigans of the x86 world. It’s fascinating the irrational emotional reactions people have to technologies which were mainly out of their scope of control or creation.
maybe this bring some lights on Jeremys article
“Writing is a funny thing.
As a young nerd, I was fascinated by personal computers and operating systems, and became a huge advocate of a funny operating system called OS/2. I spent a lot of time arguing about its merits on Usenet forums like comp.os.os2.advocacy. I moved on to Windows 95 when it was released, but always had a soft spot in my heart for IBM’s failed OS.
Twenty years later, I felt like the need to tell the story was welling up inside me until I was about to burst. I wrote the entire first draft in two days. Today, the article has been published on Ars Technica.
So did I take 20 years to write it, or two days? I guess it depends on your point of view. But I’m glad I wrote it.”
Posted by: Jeremy Reimer
http://jeremyreimer.com/m-item.lsp?i=219
Edited 2013-11-27 09:15 UTC
Exactly. Microsoft pioneered the then-revolutionary idea of being able to choose from multiple OEMs and later being able to even build your own computer to run the OS. Everything other (well-known) OS in the personal computing arena back then was wedded to overpriced hardware.
And there is no reason to believe IBM wouldn’t have done everything MS did, considering IBM’s track record, with the addition of the wedded to hardware thing MS didn’t do.
Edited 2013-11-25 23:07 UTC
I dare say, IBM were already on a consent decree with the DOJ by then for their previous shadowy business practices, so might not have had an opportunity to “do a Microsoft”.
Had OEMs been free to license OS/2, they might have done and at a price competitive with DOS+Windows (which were two products then, remember).
That’s not to say IBM would have done so, mind. They might have blown it anyway, but they did at least try courting the OEMs.
And anyway, being able to choose from multiple OEMs was hardly a pioneering MS idea. It was the OEMs themselves that made the hardware compatible with IBM PCs (and hence DOS) and CP/M before it was an example of a single OS with multiple OEMs.
Now to read the actual article….
I recently installed OS/2 Warp v3 on a 486, just to play with it. After using it for a night I can say that back in ’94 I would have taken either Windows 3.1 or Windows NT 3.5 over OS/2 without a second thought.
OS/2 veteran Michal Necasek disagrees with some of the statements in the article and he gave his view in a blog post @ OS/2 Museum:
http://www.os2museum.com/wp/?p=2144
Already posted above…
Kochise
oops sorry didn’t see that.
I’m with OS/2 Museum on this one. The Ars Technica article is ham handed.
OS/2 was quite popular in the BBS scene 🙂
Most Fidonet Mailers (FrontDoor, BinkleyTerm) and BBS software had native OS/2 versions. And together with Ray Gwinn’s IO drivers, made any BBS node “internet-capable”: it emulated COM ports for both OS/2 and DOS applications, which could be mapped to TCP ports. This was quite awesome for it’s time.
I reviewed OS/2 as it exists today (eComStation) six years ago.
Today == 6 years ago?
(Though I do get your point.)
IBM got GUI tech from Amiga, in return Amiga got the Arexx scripting language from IBM. That explains why you found the proprietary closed source IBM language Rexx in Amiga.
Mr. Reimer writes of OS/2’s backwards-compatibility as though it was a bad thing. In many cases it can be. The most common example cited is the Commodore 128, since developers preferred to write C=64 software that would run in C=128’s compatibility mode instead of writing native C=128 apps.
At the same time, backwards-compatibility has often made the transitions to new systems easier. It was a major selling-point when I bought my PS2 and WiiU, for instance.
The difference here is that PS2 still gave developers a compelling reason to write software for the new and improved platform. (Sadly the jury’s still out on whether developers will start writing good WiiU games en masse.) C=128 failed because it was actually a step backwards for the C=64 game developers; its enhancements were squarely aimed at business users, who were mostly using IBM and its clones by that point.
OS/2 didn’t give developers a compelling reason to write OS/2-native apps, but they tried selling it on its strengths at running existing popular applications. Backwards-compatibility was far less a factor than IBM’s incoherent business strategy in OS/2’s early struggles and ultimate downfall.
The downfall of IBM’s personal computer division is that it wanted to dominate the industry from both the hardware and software aspects. This is just impossible. Microsoft came close to dominance with over 90% control of the OS market at one point, but MS was wise enough to stay out of hardware (until now, with largely-disappointing results in the PC/mobile market.)
OS/2 and PS/2 might have been able to succeed if IBM had released them together as inseparable aspects of a superior computing experience. The PC clone market was ripe for a new OS during the latter half of the 80’s. It wasn’t until Win3.0 in 1990 that the PC experience could measure up to the Mac (and the PC didn’t really measure up to Amiga until Win95 came out!)
IBM should have realized that they might be able to control about 7-10% of the hardware market if they concentrated on making the entire PS/2 experience superior to the beige-box clones, and made OS/2 an exclusive but inseparable part of that experience. Instead they overreached, lost the OS battle with former partner Microsoft, and eventually declined as a PC maker until they sold out to Lenovo.
Of course, it’s hard to tell if the courts of the late 80’s or early 90’s would have allowed IBM to use the restrictive licensing terms which have made it possible for Apple to make OS X exclusive to their machines in more recent times…