OS2 World.Com Forum

Subject  :  SOA value may be best demonstrated by focusing on the applications
Author  :  obiwan
Date  :  29 Nov, 2006 on 02:27
I only now noticed this article that was referenced on the "OS/2 in the News" section.


Someone please explain to me how some things (A.) can be published without the least attempt at fact-checking, and (B.) can be stated by an "Analyst" without damaging his professional credibility. I know I'm being harsh but look at what he says:

I recall ... in 1994 ... the [DOS-based] applications we needed most did not run well [on OS/2], and often crashed the OS more than in the old DOS shell... DOS and Windows applications were haphazard on OS/2... Crashes were killing us. We ended up ... swapping out OS/2 for the latest Windows, and then on to Windows 3.3... The business people ... didn't care what the underlying platform, network, or object model was — we wanted our apps to run. Microsoft did very well by this logic; still does.

Now I grant that an administrator of DOS would have a learning curve in optimizing early versions of OS/2 for their legacy DOS applications, and I'd be the first to advise caution and proper testing in such an upgrade (Uh, oops!). But have you ever heard or experienced that Windows of that era crashed less than OS/2? That businesses fled OS/2 for Windows for the sake of stability? In DOS applications?

Then again, I never tried Windows 3.3 which must have been the stable version. The version that never existed!

Ok that's harsh. The point he makes is true, that it was the availability of applications for Windows that made the difference. However, that's not the story he told. I could go from harsh to cruel and conjecture about why these folks tell falsehoods about their personal experience with OS/2 and Windows, but I'll stop here and let someone else comment.

Subject  :  Re:SOA value may be best demonstrated by focusing on the applications
Author  :  zman zirkle@wizard.net
Date  :  29 Nov, 2006 on 11:12
he does not appear to be a pc geek (windows 3.3?); but, unfortunately, i can attest somewhat to his claims.

when my computer company was in business 1985 -2000, i built many computers with os/2 preloaded. unfortunately, too many people expected miracles from a system good for windows 3.x but way, way, way under powered for os/2. it also did take effort to configure dos sessions particularly for many of the people who got into computers for the money but who had no love or talent for computers. the industry is still full of assholes like this.

in 1995, i had a submarine game that took me a year to get runnning. i finallly realized that the memory region of the vga was not excluded by default. once i figured this out i was able to get every dos game i had to work (even myst). remember how the hp scanner card drivers would not load. well, ibm pc dos7 comes with a neat utility that will demand load a driver from the console (it won't load them high). in fact you can upgrade the dos in os/2. well, even loaded low there was enough dos memory to run the hp card and a win3.x session for ocr use! finally a lot of dos programs caused 100% cpu usage. it is a bit of a trick to get the priority settings right particularly with a client server based app. (i'm supporting two such apps even today). i think you can see the recipe for disaster.

after a while, i got feed up with the cheap lamers that wanted the audi of operating systems on the cheapest crap for hardware for a price no higher than windows. most don't know that a windows nda license which includes ms works is only forty dollars. os/2 alone then cost me 90 (all pcs at that time had paid thru intel a dos license fee and the upgrade version was an upgrade from everything). i finally restricted the builds to the best of breed and would only preload on scsi.

my own second os/2 machine (sept. 1994) was a built out alr revolution mp. it has two 90mhz pentiums, 64 meg ram, ati mach64 video, scsi, soundblaster, tecmar tv adapter, ibm m-audio card (for the voice type dictation), and of course os/2 2.11 smp (currenlty warpserver advanced -- though not in service -- still works!). it cost me 15,000 dollars. the memory alone was 3,000.

of course, there was and still is the "sig" problem! yet, once i eliminated the stupid through a miracle of economics. i was able to afford no cost tech support to all my os/2 people because well, generally speaking, os/2 just works. oh, yeah, i got few calls!!! in the hands of idiots, i've even seen netware just be crap ;-}

one of the most important reasons windows95 killed os/2 was how much money could be made fixing the unrepairable. my os/2 people paid a lot upfront; but, the windows people got reamed and reamed!! to bad the article didn't point that out!

in the u.s., the public broadcasting system used os/2 to upload all its time critical feeds. the smithsonian institution in washington dc ran on os/2 including the kiosks. i could go on; but, you know, the article seems to indicate just how lame the Ignorant Technology people at idg might have been? am i being cruel??

Subject  :  Re:SOA value may be best demonstrated by focusing on the applications
Author  :  obiwan
Date  :  29 Nov, 2006 on 17:08
Excellent point zman. OS/2 crashed for me something awful on lousy hardware when I first tried it, in 1994 if memory serves. It would do that more than DOS. Fits well with the scenario he describes.

Swapping OS/2 for Windows in his story might well have involved upgrading hardware at the same time (the other untold inevitable high cost of Windows). There does seem to be a very subtle implication in the article that those DOS applications were soon replaced with Windows versions, either upgrades or replacements.

I would guess that that high cost is part of the reason for folks exaggerating the benefits of switching to Windows, to continue to try, more than a decade later, to justify themselves.

Subject  :  Re:SOA value may be best demonstrated by focusing on the applications
Author  :  zman zirkle@wizard.net
Date  :  30 Nov, 2006 on 08:54
os/2 is extremely sensitive to bad memory sticks. sure we had a memory tester; but, an os/2 mule i kept on the tech bench was used more successfully to test memory from a pc with intermittent lockup and crashes.

however, it is the apps that matter and thats why i don't see any time is available for efforts to redo os/2 . if say osfree makes it to daylight five years from now. where will the apps be. i'd rather see the resources spent maximizing what we have.

i'm no longer a bledding edge person. i've already made my personal sacrifices to the pc industry. i have no interest in 64bit windows til the apps are more than half baked. just like with 32bit os/2, people will keep ms busy with xp and 32bits for years to come.

Powered by UltraBoard 2000 <www.ub2k.com>