Is PC DOS the predecessor to MS-DOS, made strictly for IBM? How many versions were there and when did Microsoft change it?
What was more interesting from a software archaeology point of view was when I found Symantec had acquired the assets of the then-defunct Quarterdeck; so the source for QEMM and Desqview eventually turned up in the corporate Perforce server and that was certainly fun to read, having made all kinds of things for QEMM back in the day, like a replacement overlay manager for Turbo Pascal (and Turbo C++) that would run overlay code segments directly out of the EMS page frame - this meant pretty much any program we built with the Borland compilers could be turned into a TSR with only the data segment resident.
The one thing that I never did find, though, was the source for the Whitewater Group's Actor. When it was announced that Symantec had acquired those assets from Borland I tried as hard as I could to see if the Actor source code was included but no-one I was ever able to find knew about that. Actor 3.0 was my first introduction to the Smalltalk way of building software, and I still sometimes fire it up in a VM to play enough with its VM format.
Massively predictable comment making noises about Symantec code-dumping QEMM, DESQview and DESQview/X :)
Besides being a lot of fun to read, the latter (which incidentally got some attention on here just a month ago - https://news.ycombinator.com/item?id=29396561) is a whole DOS-native X server. IIUC only Netware's commissioned port of XFree86 comes close to that, but that runs on top of Netware's kernel infrastructure so doesn't really count.
Naturally these sorts of situations sadly represent communicative chasms between geeks and management. Case studies in "it wasn't the end of the world" can perhaps be useful. My standard such example is the HP-20b and HP-30b calculators. On the surface they look just like other calculators HP made around the same time period (2008), but there happen to be a set of pins inside the battery compartment that allow reprogramming the non-bootloader-locked microcontroller inside. An SDK (https://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/archv018.c...) (consisting mostly of a keypad test demo) filled in the remaining pieces (how the keyboard matrix worked, and how to get code running on the ARM CPU, basically), and the result was the WP-34S firmware that reimplemented calculator functionality completely from scratch (the code samples in the SDK provided zero such functionality).
TL;DR, this kind of thing can definitely work and be constrained to a nice little obscure niche of very happy people with low/zero external side effects.
In this case, I wouldn't even mind if the code was released under a similar license to what D used to use (https://archive.ph/20161022202138/https://github.com/dlang/d...) before Symantec relinquished control of the project (https://forum.dlang.org/thread/oc8acc$1ei9$1@digitalmars.com) and it was made fully open-source (https://news.ycombinator.com/item?id=14060846). Microsoft licensing GW-BASIC (https://github.com/microsoft/GW-BASIC) and early versions of MS-DOS (https://github.com/microsoft/MS-DOS) under MIT is... fascinating but understandable, yet entirely unnecessary. "View only with no repercussions but not open source" works for historical interest just fine too.
What is a very good question is whether adequate precedent was set by relinquishing D that it would be tractably viable for say one or two individual campaigners to internally achieve something similar for arbitrary effectively-abandonware software that is only of historical interest.
(If this turns out to not be possible because QEMM and/or DESQview are actively being supported in some way I think the whole retrocomputing scene would be very curious to hear about that through the grapevine!)
Thanks for the interesting info about PC-DOS. That's very cool!
The history of DOS is actually really complicated, arguably even moreso than UNIX, so apologies for the mess I'm about to throw upon you. Because, before we can even talk about DOS, we need to talk about two other technologies IBM wanted in their new PC: CP/M and BASIC. CP/M was the closest thing to a standard disk operating system in the microcomputer world, and BASIC was the JavaScript of the 1980s. IBM absolutely needed both if they wanted to sell computers.
Microsoft was involved with the IBM PC because IBM needed a BASIC interpreter. The OS would have come from Digital Research. But, there was a problem: IBM wanted to use Intel's 8088 chip. CP/M only worked on the incompatible[0], older 8080, and CP/M-86 had been delayed for so long that a company called Seattle Computer Products (SCP[1]) started writing their own 8086 DOS instead. They called it the Quick and Dirty Operating System (QDOS). Bill Gates got wind of IBM's OS problems, bought SCP and QDOS, and sold it to IBM so the PC could ship on time with an OS: "The IBM Personal Computer DOS", or "PC-DOS".
So, to answer your first question strictly; PC-DOS was not made for IBM. It was made for SCP's own S-100 machines first. However, in a sense, it was made for IBM, in that Microsoft made plenty of customizations for them and it contains drivers specific to IBM machines. Bill Gates stipulated that they'd retain ownership of the OS; but the intended licensing model was for other PC vendors to buy the source code and customize it to their own, incompatible architectures. This was the licensing model CP/M used[2]; such a machine would be DOS-compatible but not run any IBM software.
Later on, of course, Compaq was able to legally reimplement the IBM BIOS, creating an "IBM-compatible" machine; and then everyone else in the industry figured out how to do the same and drastically undercut IBM. So MS-DOS was eventually turned from an OEM licensable into a consumer software product that would work on a "standard PC". Digital Research also got into the consumer OS market with DR-DOS, which Microsoft contemplated killing through underhanded tactics[3] but was able to outcompete by better OEM relationships. There's also legally distinct reimplementations of MS-DOS itself, such as DIP-DOS, X-DOS, and later FreeDOS.
[0] Blame Federico Faggin jumping ship and taking the patents with him. Or the Intel manager who demanded he show up to work on time.
[1] Not a front for the SCP Foundation. Or, at least that's what THEY want you to know!
[2] This licensing model also persisted into early Windows; there are special OEM versions of Windows 1.0 that run on Zenith Data Systems machines. It's also the reason why OS/2 1.x comes in "IBM" and "Microsoft" flavors. Microsoft OS/2 was more of a devkit for OEMs who were expected to customize OS/2 in the same way that they were supposed to customize DOS. But it also shipped with technologies that IBM OS/2 didn't have, like LADDR - pluggable disk I/O drivers that would later be reworked into Windows 95.
I should also note that the "DOS-compatible" business model looks a lot like Android fragmentation if you squint at it a little. Remember how I said this was more complicated than UNIX?
[3] The idea was to have Windows 2 or 3 throw fake, non-fatal error messages at the user if it was running on DR-DOS.
Random Google result: https://www.techrepublic.com/blog/classics-rock/my-dos-versi...