After 25 Years, 'Lost' OS/2 2.0 Build 6.605 Finally Re-Discovered (os2museum.com)

"In a fascinating example of poor timing, disk images of OS/2 2.0 pre-release level 6.605 from July/September 1991 were missing for over 25 years, only to show up literally one day after after the 25th anniversary of the OS/2 2.0 release," writes the site OS/2 Museum. An anonymous reader writes: It's the last OS/2 2.0 pre-release which didn't use the Workplace Shell (WPS), but "instead utilized the same old Desktop Manager as OS/2 1.2/1.3, which makes it the closest surviving relative of the Microsoft OS/2 2.0 SDK." Featuring a 16-bit/32-bit hybrid kernel and a "DOS Window" icon (as well as a few games like Reversi and Klondike Solitaire), "the look and feel was not quite the same as OS/2 1.3 and in fact was a cross between OS/2 1.3 and Windows 3.1."
The elusive 6.605 pre-release fell between 6.149 and 6.167 -- and "It is not known what possessed IBM to assign it a completely out-of-sequence number."

Comments

Re:OS/2 Warp 4: Better than modern Linux.

[johanw]

> systemd

Well, that's your problem. Try a Linux distribution that isn't systemd infested like Slackware. That's where I went after working some years with OS/2 2.1 - 3.0 on my home machine, after I realised that I was tuning the OS/2 setup more and more in the direction of Linux.

But OS/2 was fun compared to DOS, I could run the simulation programs I used in my graduation work in 4 DOS screens simultaneously at night and ahve them all be finished in the morning.

Mea Culpa

[mykepredko]

I, along with a few people I worked with, can take some credit/blame for that.

IBM Toronto Manufacturing in the early-mid '90s was building memory SIMMs for PS/2s running OS/2. IBM, at the time, had a standard set of seven memory tests but they didn't catch problems in a multi-tasker like OS/2 running on a 286/386 with memory management with multiple processes running so the memory was heavily used - so we booted OS/2 and started up a number of processes, each one accessing memory continuously and see if there were any defective memory chips on the SIMM that only failed during very heavy usage.

The issue was trying to figure out if the failure was caused by an instruction or data operation and where exactly it was in the address space which was problematical because the actual page where the failing bit/data was obfuscated due to memory management paging. When a SIMM failed the OS/2 test, it would be brought to a debug station where attempts would be make to recreate the problem and, using a custom OS/2 build, return to that sequence where the problem lay and, using a processor emulator, determine what the address was failing.

As a side note, memory at the time cost $150/Mbit, so it was worth the time and effort to find and replace the defective chip on a 1-2 Mbyte SIMM.

I'm not sure exactly how our code got into the main branch, I suspect it was because there were other things that were fixed in that recovery code and situations where there was a failure resulting in the loop of death wasn't part of the test sequence.

I do remember that there was a fairly simple way of cold booting the system from so that you could avoid this failure loop (we called it the "loop of death") - If memory serves correctly it was pressing 'F8' when the OS/2 logo comes up and then selecting a cold boot from a menu (this is going back more than 25 years ago so don't shoot me if I'm wrong). OS/2 support at that point in time was pretty good, they would have explained how to get out of the loop of death as it was a pretty common problem, especially with badly behaved Win 3.11 apps.

Re:OS/2 was obsolete

[Anonymous Coward]

Indeed. Often times a dead platform is killed off simply because it's parent company went out of business due to overly aggressive, potentially illegal, business tactics by competitors.

Other times a platform ends up dead simply because an inferior, less desirable, replacement was conjured up by it's parent company who is having trouble selling the new system simply because the old is still around.

Oh wait. Perhaps there's no need to leave it dead after all! :P

Re:IBM Tech Support was horrible

[Greyfox]

Apparently the support line was the top rated one in the industry early on, but the penny pinchers started pinching pennies. When I got there, they would pretty much support you for life if you'd ever bought OS/2 or an IBM PC. They started out with screeners taking the call initially and then transferring it to level 1 support, but they did away with the screeners and added that responsibility to the call center guys. They also got a lot more picky about you having a set amount of support per license key and they started enforcing that. And they also added a 900 number, which apparently got a fair bit of traffic. They did have a network support team as well, and that covered all the shifts. Those questions were queued up and answered as they came in, and the guys who answered them were considered to be "level 2 support" if I remember correctly.

Around the time Windows 95 came out, there was a push for all the people in the call center to get the "OS/2 Certified Engineer" rating, but IBM shut down OS/2 before anything much came of that. I got mine at the '95 Comdex, while doing volunteer support for Team OS/2. Still have the little plastic card...

But yeah, most of the level 1 guys didn't have any experience with OS/2 and a few didn't have any experience with computers, when they started. About 90% of the problems that came in were for similar issues though -- printer stuff and video problems seemed to be the most of them. I still have the command line command to reset the video drivers to VGA burned into my brain. I could actually fix your shit for a wider range of problems, if you were lucky enough to get me, but fixing your shit is time consuming and I was frequently in trouble for not answering as many calls as I was supposed to be. A lot of the techs just wanted to throw a reboot-requiring command at you and make you go away so they could keep their numbers up.