My last piece on OS/2 was in part a mea culpa, a history of my part in its downfall. However, I can't claim all the credit. In fact, if I'm honest, there were hundreds of reasons why OS/2 failed, and most of them had nothing to do with me. So, here are some of the real corkers. Once upon a time, IBM made extraordinary money …

COMMENTS

House rules Send corrections

This topic is closed for new posts.

Page:

Monday 26th November 2012 11:29 GMT Anonymous Coward

uhm...

"Installing OS/2 wasn’t very hard, but you couldn’t guarantee it would work on your hardware, and at about a thousand bucks a go this was an expensive experiment."

Sorry, but that's not entirely true. There were plenty of lists supplied which provided in every detail which hardware components were and which weren't supported. It was relatively easy to look this up before you started. Even the IBM OS/2 website was very easy to use when it came to finding these lists (which was pretty much amazing considering how a lot of other information was burried between huge amounts of totally confusing links and pages).

Still; if your hardware (-components) was (/were) mentioned as being supported then 9 out of 10 cases it would easily work. The only time I had major issues with OS/2 was when I actually bought myself an IBM (iirc an Aptiva) with the sole purpose of running OS/2 on it. I mean; if I got myself a /real/ IBM computer, surely it should do wonders with OS/2? Guess again!

Which was partly my own dumb fault; I assumed IBM cared. Which IMVHO is what really killed OS/2; IBM themselves. They didn't seem to care one bit. Take the Aptiva; did it come with OS/2 drivers? No; only Win95. The 486 Compaq I once had supplied drivers for just about anything; /including/ OS/2 (not directly but on demand; I recall purchasing a huge amount of drivers and support software; 3 boxes of 3.5" disks which also included plenty of OS/2 drivers).

I even contacted IBM support with this (considering how I was entitled to support through purchase of my new PC). Needless to say, but they didn't quite manage to get beyond "We think it should work...".

But its the weirdest thing; on the other hand IBM did quite a bit of good. I for once was very impressed how Sun's Java (back then a rather new environment) had found a rock solid home within OS/2. Not sure about OS/2 3 ('Warp') but Merlin (the one with the nice pull down menu) supplied Java right out of the box. That was impressive for those days, at least IMO.

There IBM did manage to do the right thing; Java support really managed to extend the stuff one could do with OS/2; now you weren't merely "limited" to your common DOS batch, OS/2 cmd or Rexx scripts. (not that you were really limited; man... I still vaguely recall the stuff one could do in a cmd file. Sjeesj!).

No, if only IBM would have cared a bit more; imagine OS/2 under the supervision of IBM as an open source project, but not in a way where it would only cost IBM money (like Sun did), but in a productive fashion... That could have easily become a power player which could have gone toe to toe with Windows today.

IMVHO of course.

6 1
1. Monday 26th November 2012 11:57 GMT dogged
  
  Re: uhm...
  
  I used OS/2 3 Warp. In the Netherlands. At... IBM. They were the only place in the universe that actually ran the fucking thing. How was it? Well. It sort of worked. Mostly. It wasn't any quicker than Windows 3.11 and not even IBM could get it to print to an IBM printer but most of it worked. More or less.
  
  Further, the development we were doing was Java via NetBeans which was such a truly foul experience that it's soured me on Java forever.
  
  IBM. The only company in the Netherlands that (at the time) made you wear a tie to work.
  
  7 2
  1. Monday 26th November 2012 23:06 GMT Anonymous Coward
    
    Re: uhm...
    
    Dunno which bit you were in then but where I was the ibm nl dress code was most definitely beard, sandels and absolutely no ties. Mind you i was working on crusty mainframes and not this new fangled pc malarkey.
    
    0 0
    1. Tuesday 27th November 2012 10:21 GMT Anonymous Coward
      
      Re: uhm...
      
      I don't think I ever saw any ties in Dynatos - I did see moustache, short shorts and sandals (yes, worn by the same person).
      
      0 0
  2. Tuesday 27th November 2012 13:51 GMT Stoneshop
    
    @dogged
    
    IBM. The only company in the Netherlands that (at the time) made you wear a tie to work.
    
    Well, at DEC FS my manager repeatedly insisted very, very strongly I'd wear a tie, but he did stop short of forcing me.
    
    "What do you think a customer prefers, a FS engineer with or without tie?"
    
    "One who fixes his broken stuff, I'd say" (which I undisputedly did)
    
    Which made it the last of his attempts.
    
    2 0
    1. Wednesday 28th November 2012 17:07 GMT Matt Bryant
      
      Re: @dogged
      
      "....my manager repeatedly insisted very, very strongly I'd wear a tie...." I had an ex-IBM manager like that when I was contracting. All the internals lazed around in jeans and t-shirts but he insisted contractors had to wear a dark blue or charcoal suit (only those two colours allowed!), white or light-blue shirt and a dark-coloured tie. On the first day I turned up in jeans and told the manager, if he liked my suit so much, I'd have it couriered in and he could hang it over a chair and ask it how to fix his problems, but if he wanted my brains then he would have to put up with more casual attire. I had a clause in my contract that I'd sneaked past their HR bunnies that I could choose to wear "suitable attire" - his staff dressed down so he couldn't argue a suit was necessary. He refused, I walked, his boss called to ask me to come back the next day. I went back in jeans. Moral of the story - an employment contract is a two-way street, make sure you get what you want. If it's not in the contract it's not enforceable unless required by law.
      
      3 0
2. Monday 26th November 2012 12:09 GMT Anonymous Coward
  
  Re: uhm...
  
  In 1990 I was working for the market leading supplier of graphics chips for PCs. From our perspective, IBM were completely unhelpful on graphic driver development support. We responded with the minimum of time and effort put into performance optimizations for OS/2 drivers. It was perfectly clear OS/2 was going nowhere with this attitude and a relief when IBM dropped out of the picture for OS/3 (aka NT).
  
  6 2
3. Monday 26th November 2012 12:57 GMT DrXym
  
  Re: uhm...
  
  By 2.1 it was usually very easy to install OS/2 on random software. I somehow persuaded IBM to send me a copy for evaluation and I had no issue installing it on a Evesham Micros no-name box at home or on several PCs I tried around my workplace.
  
  The biggest issue was actually not installing the software but keeping the bloody thing stable and up to date. IBM were notorious for shipping broken software and expecting people to download and apply an unending series of Corrective Service Diskettes (CSDs). They seemed to flow like water for OS/2 and C Set++ which I used for development.
  
  OS/2 3 also introduced a weird ass floppy format called XDF which to this day must cause people issues, especially those who fancy the idea of installing OS/2 in a VM for a bit of nostalgia.
  
  3 0
  1. Monday 26th November 2012 15:02 GMT Yet Another Anonymous coward
    
    Re: uhm...
    
    At least you got C++, we (a world top5 university) couldn't get a copy of any sort of compiler out of IBM for OS2.
    
    Their attitude was basically, if you want to buy a mainframe - fine, if you want to buy any other IBM product go **** yourself.
    
    Eventually we gave up, switched to buying Dec-Alpha PCs running NT, wiping them and installing linux. Why Dec alphas running NT were a fraction of the cost of Alphas running VMS or Ultrix is the reason Dec isn't around today either.
    
    5 0
    1. Tuesday 27th November 2012 10:48 GMT SDoradus
      
      Re: uhm...
      
      Right, and the Red Hat 7.3 boxed edition for Alpha came with a really really cool Intel Fortran compiler that our researchers lusted after. And under Linux they could log in remotely so more than one researcher could use the machine...
      
      2 0
4. Monday 26th November 2012 15:37 GMT I ain't Spartacus
  
  Re: uhm...
  
  ShelLuser,
  
  If I remember rightly IBM came up with Aptiva (and later Ambra) as sub brands that would sell to consumers, without cannibalising their much more expensive corporate PC market. So it's possibly that there was no OS/2 support due to it being a different division, accident, incompetence - or even deliberate decision, so you'd have to buy the most expensive IBM PCs to get it.
  
  My first PC was an Ambra, because it was very cheap. That's the last time I bought a PC by simply going through the pages of PCW and ripping out all the adverts that looked to be a decent buy. Took ages, buy going online wasn't an option. That was in the days when PCW was a 600 page behemoth, with only about 100 of those being editorial. Many many adverts, some with things stuck to them, so the magazine would fall open on their page and others with fold out pages (for the same reason). Some cheapskates trying to stick whole catalogues in about 2 pages of 6 pt type, in black on blue. Ugh!
  
  Happy days...
  
  4 0
  1. Monday 26th November 2012 16:01 GMT Gwaptiva
    
    Re: uhm...
    
    It was worse. IBM The Netherlands actually concocted an OS/2 + Windows in DualBoot crAptiva, and sold it to its retirees!!! Disaster! I did tech support on the bloody things, and guess where my nickname comes from :)
    
    There were also quite a number Choose'n'Lose machines: You booted the first time and had to choose an OS: Windows 95 or Warp... guess the shock if someone called in that actually had chosen Warp... invariably by accident.
    
    Mind you, Compaq shipped pre-installed OS/2 Warp long after IBM actually discontinued the product.
    
    3 0
  2. Tuesday 27th November 2012 10:18 GMT Anonymous Coward
    
    accident, incompetence - or even deliberate decision
    
    "it's possibly that there was no OS/2 support due to it being a different division, accident, incompetence - or even deliberate decision"
    
    I used OS/2 for about 3 years before I joined IBM and I did some work with IBM PSG (Personal Systems Group) at the close of the last century; I'm willing to bet no decision was made, incompetence was involved and the only accident would have been if they did work together.
    
    2 0
5. Monday 26th November 2012 20:49 GMT henrydddd
  
  Re: uhm...
  
  I had a project to construct a BBS as well as let people do normal office functions on the same computer. The computer was a 486 and had Windows 3.10 installed. I put RBBS on it and it worked, but when I started Microsoft Word, the modem would not function due to the fact that Win 3.10 had cooperative multitasking. I purchased os/2 (I think 2.1) which had preemptive multitasking. It also had the capability of running Windows 3.10 at the same time. The whole thing worked perfectly. The BBS ran well and the office work could be done at the same time. I remember os/2 as being an operating system ahead of its time. Perhaps the reason for its demise was that Microsoft saw that too.
  
  4 0
  1. Tuesday 27th November 2012 10:38 GMT Allan George Dyer
    
    Re: uhm...
    
    The pre-emptive multitasking encouraged me to choose OS/2 Warp over Win95. The ability to open a DOS box and make a new connection to your Novell server was also very handy when an admin task [i.e. user needs a password reset] cropped up while you had lots of stuff open.
    
    Eventually moved to NT4 when the ability to read incoming Office documents and handle Chinese forced the move.
    
    2 0
6. Tuesday 27th November 2012 09:45 GMT John Angelico
  
  Re: uhm...
  
  ...I understand your broad point, but the writer of the article was not doing what you describe.
  
  He was trying to emulate the ease of installation of Windows with almost any hardware that happened to be hanging around, and which may have been literally thrown around the office.
  
  MS-Windows was then very forgiving of dodgy quality hardware, whereas OS/2 was far more finicky, but once installed it took a while to bed itself down - a bit like the old days of "running in" a car.
  
  I still run eCS 2.x having started with OS/2 for Win (aka "Ferengi"). It's about time I upgraded - this box has been here with minimal attention for about 4 years. The only thing that doesn't work is the SATA DVD writer - testament to my impatience.
  
  And yes, I am an OS/2 zealot.
  
  0 0
Monday 26th November 2012 11:57 GMT Paul Johnston

Terse?

" I was one of the fools who didn’t realise that English is the most terse of all major languages"

I'm learning Mandarin which makes English seem so flowery and over elaborate.

3 0
1. Monday 26th November 2012 12:39 GMT xperroni
  
  Re: Terse?
  
  It depends on your reference... Chinese (and Japanese for that matter) are more economical than English, but compared to most Romance languages it's a paramount of objectivity.
  
  5 0
2. Monday 26th November 2012 12:58 GMT Anonymous Coward
  
  Re: Terse?
  
  "so flowery and over elaborate"
  
  are you sure of that?
  
  are you sure?
  
  you sure?
  
  sure?
  
  5 0
  1. Monday 26th November 2012 13:35 GMT Kristian Walsh
    
    Re: Terse?
    
    Despite their "complexity", the CJK scripts are the easiest for UI processing, even easier than English, because there are no plurals (Japanese "counting" words excepted). Once you've solved the technical problem of managing fonts with thousands of glyphs, and for Korean the relatively straightforward task of of composing Jamo clusters, everything's simple.
    
    (an aside: Korean has one of the most elegant and logical writing systems in the world - a half a day's practice is enough to be able to read text out loud, although you still won't have a clue what you're saying).
    
    The most problematic language for UI layout is Portuguese; it's the most verbose of the Romance languages, and its particles are longer than the next-worst, French.
    
    The most difficult for message formatting are the Slavic languages, with their multiple number cases, which provide no end of punishment for people who still write stuff like printf("%d item%s",count, (count!=1)?"s":""); ... I used that line at the start of a training course once, and the invariable reaction was "Oh, that's a neat way to do plurals". Yes, and all CPUs are little-endian with 32-bit addresses, and characters are bytes...
    
    7 0
    1. Monday 26th November 2012 22:07 GMT ender
      
      Re: Terse?
      
      > The most difficult for message formatting are the Slavic languages, with their multiple number cases, which provide no end of punishment for people who still write stuff like printf("%d item%s",count, (count!=1)?"s":"");
      
      Guess what? Windows still doesn't know how to do plural forms properly, resulting in Explorer telling me "Selected is this number of files: 1"
      
      2 0
      1. Tuesday 27th November 2012 10:10 GMT Kristian Walsh
        
        Re: Terse?
        
        Heh. I can almost see the strings table:
        
        "%d file selected",
        
        "%d files selected"
        
        or worse:
        
        "%d file%s selected",
        
        "s"
        
        ... It's not Windows, it's the developers. Pretty much every API is weak on this (gettext would be okay if it didn't encourage the use of natural-language text keys, and have the enormous arrogance to assume that your source language is always pluralised like English),
        
        Localizability was normally tacked-on a the end of projects, rather than being considered up front, and this attitude leads to exactly the "Items chosen: %d" kludge: localisation never started until after code-freeze, so there was no way to go back and fix the problem at source.
        
        2 0
    2. Tuesday 27th November 2012 10:53 GMT SDoradus
      
      Re: Terse?
      
      He's right you know. I first noticed this while doing my BA (hons) in French language and literature in the 1990s while working as a sysadmin. Every time I picked up a 'rosetta stone' (piece of paper saying the same thing in multiple languages) from a software box, it was noticeable that English was always far and away the shortest paragraph (and French one of the longest), with just one exception: legalese. English legal warnings were much longer than their continental counterparts.
      
      4 0
    3. Saturday 1st December 2012 09:11 GMT Paul Johnston
      
      Re: Terse?
      
      In Chinese you have to decide between er and liang whereas we would just use two!
      
      1 0
  2. Monday 26th November 2012 15:39 GMT Anonymous Coward
    
    Re: Terse?
    
    "so flowery and over elaborate"
    
    are you sure of that?
    
    are you sure?
    
    you sure?
    
    sure?
    
    Not quite the correct form for a haiku, but close. My offering:
    
    OS2 1.3
    
    No one cares
    
    Now long dead
    
    1 1
Monday 26th November 2012 11:58 GMT Anonymous Coward

os/2 pricing to fail

In the late 80s, Microsoft were trying to sell DOS+Windows to OEMs,discourage use of DOS apps and encourage Windows application development. To say Microsoft were paying hardware vendors to ship Windows makes no more sense than saying IBM were paying vendors to ship Presentation Manager. IBM was not the only OEM shipping OS/2; Compaq, HP and a whole bunch of lesser known companies sold OS/2 product.

True, that couple of megabytes or more extra DRAM needed was a big deal at the time for most of the Personal Computer market.

The comment 'Microsoft was paying people to sell Windows' illustrates how out of touch with reality some of the IBM people were back then. Microsoft were pricing for volume and profit simple as that. IBM were indeed pricing to lose the mass market as the author suggests.

3 3
1. Monday 26th November 2012 12:52 GMT InsaneGeek
  
  Re: os/2 pricing to fail
  
  I remember paying someone $100 for 1MB of used memory back at that time. The minimum requirements for OS/2 was a massive factor as to why I didn't run it even though I wanted to, I'm guessing that for a significant portion of the populace was as well.
  
  The rationalization for me was this: I'm transitioning from the DOS world into the cool new GUI based world... I can either run MS Windows and have enough money to buy it and a number of apps, or have enough money to buy OS/2 and upgrade my hardware but have no apps.
  
  2 0
  1. Wednesday 28th November 2012 13:19 GMT Tom 13
    
    Re: I can either run MS Windows and have enough money to buy it and a number of apps
    
    My recollection is similar, except I couldn't even afford the fancy new MS software at the time.
    
    Also, MS did get some luck in the CD-ROM becoming standard gear soon after they introduced 95. I recall the shellacking PC Mag gave IBM for the number of disks required to install the first release of OS/2. But nothing was said when 95 required 4 more disks when it was initially released. While getting the CD-ROMs to work with boot disks was initially a PITA, it was worth it to avoid disk swapping and the ultimate failure of one of the disks. I mean, we almost always had problem installing WordPerfect 6.1, and that was only 6 HD 3.5 floppies.
    
    1 0
2. Monday 26th November 2012 13:09 GMT DrXym
  
  Re: os/2 pricing to fail
  
  I think the difference in attitude between Microsoft and IBM could be most seen in their treatment of developers.
  
  While MS had a developer program you could also obtain a Windows SDK with a compiler for nothing. You could also buy Visual C++ (and Borland C++, Watcom etc.) at consumer affordable prices. The tools were also user friendly and backed up by good quality documentation.
  
  On OS/2 you were expected to pay a large sum of money for C Set++ and another large sum of money for the OS/2 developer connection to get the SDK and inevitable stream of patches. And there was no IDE for the tools. Not even a decent editor. It took IBM until VisualAge C++ to get an IDE and anyone who has used VisualAge knows what an abortion it was. Borland did actually dip their toe in the OS/2 market and shipped a proper IDE, one which was actually quite good but the compiler and support libs were so buggy as to be almost unusable.
  
  So they severely tested their developers while Microsoft was doing everything they could to help them. And this had a knockon impact on the amount and quality of the software that each platform subsequently enjoyed.
  
  1 0
  1. Monday 26th November 2012 15:15 GMT fajensen
    
    Re: os/2 pricing to fail
    
    I was running OS/2 Warp on a one of those 486 machines where you had to pay 70 quid for a co-processor and insert it yourself. I used the VisualAge C++ to compile 6-8 open source programs that I wanted to have and then I finally realised that I might as well install Slackware Linux on the thing, get the goods directly. That installation eventually took me about 3 months to get to work properly with many kernel rebuilds, but it did work, even sound and the extreme speed 19k Modem with PPP and auto connect.
    
    Warp was quite good, but at that time OS/2 was already cirkling the drain. Certainly, one would not drop big money on applications for it!
    
    1 0
Monday 26th November 2012 12:01 GMT Anonymous Coward

OS/3

At the tail of the millennium no-one thought to just go for OS/2000?

0 0
1. Monday 26th November 2012 13:03 GMT Ken Hagan
  
  Re: OS/2000
  
  Perhaps they had the sense to realise that most PC people associated "2000" with the millennium bug.
  
  Then again, around the same time Microsoft produced a version of Windows that was named after both the bug *and* a disease.
  
  1 1
  1. Monday 26th November 2012 13:59 GMT Andy Miller
    
    Re: OS/2000
    
    OS/19100 would have been funnier...
    
    10 0
Monday 26th November 2012 12:02 GMT Matt Bryant

"....The French version was harder to produce than you might think....."

Now that did bring back painful memories! One civil service project took on an hilariously sour angle when French was added to the spec at a very late date. Your's truly thought that it was a problem for the coders but it actually tuned into a project management headache due to politics - the Francophiles wanted the project to be in French so they could gift it to their chums, whilst other civil servants with more anti-European views were determined no British taxpayer cash should be wasted on pleasing our Continental colleagues. It got even worse when the French bod the Francophiles had introduced to the project insisted all the manuals and such items as stickers on the hardware had to be in both English and French. It got to the point where we had to redesign a lid for one piece of kit just to take a bigger bi-lingual sticker!

Whilst the poor coders struggled for weeks to fit the longer French phrases and still meet the original project deadline, the war between the civil servant factions spread to include those wanting German and Spanish editions! Towards the end I was doing three meetings regarding languages for each one regarding actual project progress. Thankfully I managed to find a bod in Charles Hernu's office that confirmed there was no way the French would consider using filthy English code (I'm paraphrasing but that's pretty close to the sentiment!) and the matter ended there. Funny now but very illustrative of how often the biggest project issues are rarely technical in origin.

6 2
1. Monday 26th November 2012 16:58 GMT Destroy All Monsters
  
  Re: "....The French version was harder to produce than you might think....."
  
  > find a bod in Charles Hernu's office
  
  What are you doing with the defense minister of La France?
  
  0 0
  1. Monday 26th November 2012 20:04 GMT Xelandre
    
    Re: "....The French version was harder to produce than you might think....."
    
    He was the minister's joint MI5/KGB controlling agent?
    
    http://www.lexpress.fr/actualite/politique/comment-hernu-est-devenu-agent-du-kgb_497405.html
    
    2 0
  2. Tuesday 27th November 2012 11:11 GMT Matt Bryant
    
    Re: Re: "....The French version was harder to produce than you might think....."
    
    "What are you doing with the defense minister of La France?" Upsetting Dassault, apparently. Allegedly, etc.
    
    1 2
Monday 26th November 2012 12:02 GMT Mondo the Magnificent

OS/2 my perspective

I was in the (non IBM) mainframe game when OS.2 1.0 arrived in our company

We liked it, but at the time we were also toying with Windows 2.1 in the pre "32 bit" era.

From a business perspective, neither OS/2 or Windows offered us much productivity due to both desktop OSes being relatively new and lacking good productivity applications. We opted for DOS and the likes of Lotus, Quatro Pro and WordPerfect to get things done.

Later I joined a distributor and at this time IBM was announcing OS/2 Warp. Hell, yea, something to give Windows a run for its money. A while later, the local IBM office was doing a presentation on the forthcoming OS/2 Warp! and, yes I was invited and I had to be there... The presentation was done on a ThinkPad running Windows.. WTF! That was commercial suicide at corporate level.

It seems as if IBM has something really good up its sleeve but had no idea how to entice Joe Public the way MS did. Also, lack of application development steered OS/2 into the corporate desktop opposed to the layman's home PC, although I was informed that OS/2 had a rather good following in Germany and some of the Nordic states.. I may need to be corrected on that..

Up until a few years back I used to pop into my local HSBC and sit down with a banking consultant whilst grovelling for a mortgage, loan or advice. That was the last time I ever saw an OS/2 screen, the banks were lat stand for OS/2 and as most banks close down branches, it seems that OS/2 drowned with the sinking ships known as local bank branches..

This two part article by Dom Conner does shed a lot of light on the internal incompetancies that doomed IBM's desktop OS. Also, this type of corporate wrangling and bullshit at Big Blue really does make sense..

A very infromative read to the birth and non-start of what could have been Microsoft's nemesis..if only...

7 0
Monday 26th November 2012 12:06 GMT Flawless101

Good article

I enjoy reading articles like this, to get an insight to how others experience(d) the world of IT.

6 0
Monday 26th November 2012 12:07 GMT J. R. Hartley

Amiga connection

Didn't Commodore license Workbench/Intuition for use in OS/2 in rerurn for REXX?

I may or may not be correct. Please advise.

0 1
1. Monday 26th November 2012 14:20 GMT JaimieV
  
  Re: Amiga connection
  
  No, REXX was an open language and AREXX was written without IBM involvement or any need for licensing.
  
  I've never heard of Workbench being licensed out (which doesn't mean it wasn't!).
  
  1 0
  1. Monday 26th November 2012 14:42 GMT J. R. Hartley
    
    Re: Amiga connection
    
    Conflicting reports. Things like this:
    
    http://www.amiga.org/forums/showthread.php?t=22076
    
    0 1
    1. Monday 26th November 2012 23:23 GMT CoAImBRmoMdMore
      
      Re: Amiga connection
      
      RJ Mical's patent 4772882 was referenced by some IBM patents (among others). Perhaps that's the connection someone picked up as significant.
      
      0 0
Monday 26th November 2012 12:08 GMT Anonymous Coward

I bought OS/2 2.0 when it came out ... considered both win 3.0 and OS/2 but for me OS/2 had a price advantage - on my "home built" PC I was running DR-DOS and upto launch of win 3.0 all the reports were "win 3.0 will not run on DR-DOS" so I factored in buying MS-DOS into the win 3.0 price. In reality turned out that MS had found some DOS call where DR-DOS behaved slightly differently from MS-DOS and used this difference to ensure that win 3.0 pre-release would not run on DR-DOS and since DR were not allowed access to pre-release win 3.0 they couldn't comment on this ... in reality they came up with a patch a matter of hours after the release.

Anyway, some people in IBM were keen on supporting OS/2 users ... I remember when one of the beta's came out commenting on usenet that I'd have to spends a evening or two downloading all the floppy images and a few hours later got an email from someone in Hursley saying if I sent them the required 20 floppies they'd copy the images and post them back to me!

4 0
1. Monday 26th November 2012 12:28 GMT Anonymous Coward
  
  As far as I know this deliberate DR-DOS breaking is a myth. Win 3.0 alpha and beta versions were made available to OEMs for 18 months prior to release and certainly some of them used DR-DOS. I never heard anyone on the 3.0 dev team express any interest one way or the other in DR-DOS when I was in Redmond. Windows was so easy to patch around anyway its highly unlikely to my mind anyone bothered.
  
  2 7
  1. Monday 26th November 2012 12:41 GMT myxiplx2
    
    Deliberate breaking of DR-DOS a myth?
    
    A myth? Groklaw says different. This is just one except of many from it's publishing of the court cases involved:
    
    Source: http://www.groklaw.net/articlebasic.php?story=20120711170909394
    
    "FTC investigators also concluded that in order to sabotage DR DOS, Microsoft had carefully written and hidden a batch of code into tens of thousands of beta copies of Windows 3.1 that were sent to expert computer users in December 1991. When someone tried to run one of these Windows 3.1 beta copies on a PC using DR DOS (or any other non-MS-DOS operating system), the screen would display the following message: "Nonfatal error detected: error 4D53 (Please contact Windows 3.1 beta support.) Press C to continue."
    
    To expert beta-testers using DR DOS with Windows, this message would convey that they could continue using the program, but it might cause problems. The effect would be to deter some from using DR DOS further; others would call Microsoft for an explanation of the supposed risks of using DR DOS."
    
    10 2
    1. Monday 26th November 2012 14:44 GMT TeeCee
      
      Re: Deliberate breaking of DR-DOS a myth?
      
      And the problem is what exactly?
      
      You have a Beta copy of 3.1, which is designed to sit on top of MS-DOS. It detects on startup that it's actually on A.N. Other DOS and issues a nonfatal warning. The only issue there would seem to be that they could have been a shade more explanatory in the error message.
      
      That's nowhere near being sabotage. I actually consider issuing a warning and continuing to be quite generous, given that they probably weren't entirely sure what effects might be caused and a positive result from those Beta evaluations was very important. If it had been me, I'd have had it issue an "Unsupported O/S" message and crap out on the spot.
      
      If they got successfully sued over that, some lawyer somewhere needs a kicking. As usual.
      
      3 8
    2. Tuesday 27th November 2012 14:05 GMT Anonymous Coward
      
      Re: Deliberate breaking of DR-DOS a myth?
      
      I stated that DR-DOS breaking for Win 3.0 was a likely a myth, based on my personal knowledge, not what I read in some blog or other. Unsurprisingly for this site, downvoted.
      
      A separate point has been made that two and a half years later, a beta of Windows 3.1 issued a warning when run on a non-MSDOS system (the actual 3.1 release did not show this message).
      
      I was not involved with Win 3.1 at this time but it seems to me entirely plausible that there were technical reasons for limiting a beta to tested MSDOS configs. In particular the VxD and other components used to host the Windows environment and DOS multitasking on top of various versions of MSDOS and DRDOS had grown to be really complicated. All written in cryptic assembler evolved from the Windows 2.x introduction of 386 mode several years earlier. Memory extenders had to be supported. A nightmare in its necessary but obscure use of DOS calls and virtualization I don't envy whoever was tasked with updating that code for 3.1. This code was available to OEMs not hidden in-house, not some ultra secret sauce. Early versions of DOS were not supported either.
      
      I've not a clue whether the 'FTC investigators' studied the code and development process in sufficient depth to find any technical evidence to support their claim. I suspect the findings were simply a lawyer-type hypothesis on the 'no smoke without fire' theory that because senior MS sales people wanted to sell against DRDOS one or more developers must have unprofessionally and unnecessarily broken the beta product. Personally, I find this conspiracy theory rather suspect but without evidence its anyones guess. Is there any published evidence?
      
      Incidentally. When the product specification for 3.1 was evolving in 1990 there was a school of thought this was maybe the last DOS/Windows combo and Windows would move to NT kernel by mid-decade. History didn't work that way and we suffered Windows 95 etc. waiting until 2001 for XP to finally put DOS to the sword. With this context, the kill the DRDOS competition theory sounds a bit thin to me as far as any involvement of the development group goes.
      
      2 4
      1. Wednesday 28th November 2012 13:29 GMT Tom 13
        
        Re: Unsurprisingly for this site, downvoted.
        
        Possibly because many of us have personal experience with how MS broke any non-MS software related to Windows and DOS during that time period. And yes, I was cutting my IT teeth as a power user at about that point in time. I was the DTP specialist running Ventura Publisher and buying Windows and Corel Draw with my own money to support projects. We used QEMM at the time, and MS constantly and needlessly broke their software.
        
        2 1
2. Monday 26th November 2012 12:40 GMT AndrueC
  
  There used to be quite a bit of support via Compuserve. I remember several discussions with a chap who worked on HPFS. Doug Azarito I think his name was.
  
  1 0