My OS/2 barf bag story
When I worked at Borland way back in the early 90s most of the company wasted a year or so building OS/2 versions of our products. The PC OS story hadn't finished yet, but most of us new it was over. Even so, it didn't feel completely wrong to be writing software for OS/2.
Heck, I was hired a year prior primarily because I had "extensive" OS/2 experience. At that time Borland had just struck a deal with IBM to provide OS/2 versions for many of their products. It was a ton of cash. But also a crappe tonne of time. Time Borland did not have while in Microsoft's crosshairs. But I digress.
The project I was on worked with an IBM liaison. He'd fly out periodically to hand deliver OS/2 beta updates on floppy disks. It was amazing. I don't recall the exact number, but the OS consisted of about twenty 3.5 inch floppy disks. Even back then it seemed insane to install anything with that many floppies.
Note, we were basically building a new product using our own early beta C compiler for an early beta OS. Was the bug in our code, the compiler's code, the OS code? But I digress.
The software usually came in nicely tailored, reusable boxes that sat upright. Not quite 21st century Apple quality packaging, but respectable for the time. Well, the last hand-delivered OS/2 update came in a barf bag from the IBM liaison's flight. He dropped it off on my desk, I thought he was joking. No.
I don't recall the excuse, but the mood concerning OS/2 at that time wasn't exactly upbeat. I called it the barf bag omen. The rest is history. This is hilarious and simultaneously sad. Thanks for the story. I remember working on OS/2 from my internship in college through the first few commercial software jobs after graduation though we used IBM's C/C++ compiler, which may have been shared with Microsoft but the exact origin I don't recall. I very much preferred Borland from my days in Turbo Pascal and Turbo C++ though I can't remember if that was just Intel or Atari ST (68K)? Lots of game-changing software was built on OS/2 at the time because getting that kind of thread performance was nearly impossible since it predated everything else in the late 80's so things like ATM machines, stock trading systems, even things that are still being sold today though I'm sure OS/2 is long gone from the releases: https://www.broadcom.com/products/mainframe/product-portfoli... Pretty amazing for the time given there are things we did there that just haven't lived beyond NT or the like. It's sad that the entire industry seemed to forget about CUA en masse. It would be really useful now with the return of tiling to prominence. > I very much preferred Borland from my days in Turbo Pascal and Turbo C++ Turbo Pascal was stunning for its time. A fully functional editor and fast compiler operating from a single 5.25 floppy, which occupied one of the two A/B drives on my university's PCs before hard disks were affordable. But I have to admit OS/2 felt more like the "adult" OS compared with Windows. To be fair Windows had to load on top of DOS, so it didn't have the luxury of using the PC to its full extent. But still. Wow, reading the summary of CUA it seems like a dream (even if a few standards like cut/copy/paste, seem awkward) in today’s UI hellscape. Swipe left for the menu. Then tap Settings, then that spawns a modal. Swipe up and down to scroll it if you realize there’s scrolling with an invisible scroll bar, oh nope, you swiped too vigorously, we interpreted that as a close gesture because we animate in our modals from the bottom so that seemed natural to us. Modern software was a mistake for sure. I worked for an IBM subsidiary in the early 90s while they were building and releasing HPTS. Our sub did ancillary products for HPTS and 3890 sorters, all in “c”, assembler, and running on OS/2. The OS/2 API was, IMO, more logically designed and documented than anything since. It was a blast. Sadly, OS/2 didn’t survived. But, it was fun while it lasted. >The OS/2 API was, IMO, more logically designed and documented than anything else since. Absolutely. I was disappointed when I began writing Windows apps after having used Presentation Manager etc. Windows APIs felt like children’s toys. Sometimes I hear these stories, and wonder how people in the future will find these kinds anecdotes and understand them. Will they even be able to discover them? Is there going to be a modern version of the jargon file? I hadn't thought about that, but you're right. My story probably makes little sense to non-programmers of that time. Going deeper, most of whatever the 20th/21st century leaves for future historians may very well be indecipherable, if at all recoverable.