The technical documentation was really good. Complete BIOS listing (that as fun to read), and IIRC there were circuit diagrams (since there was very little VLSI -- just the processor, memory, and a bunch of gate-level TTL chips).
I think that BIOS listing sealed IBM's fate to lose control of the platform; it made it near trivial for a competitor to analyze and shove under the doorway of a "clean room" team to reimplement.
A lot of things had circuit diagrams then, I did chip level repairs on Kaypro clones, and CP/M machines at the time, and it was all TTL off the shelf chips. You could follow the signals through the diagram with a logic probe, clip out the broken gates, clean the board, and pop in a new one, it was very satisfying.
On the other hand, all that information was available for the Amiga too (and probably also Atari), in fact it was quite normal on home computers to get the complete hardware schematics in the box, and the ROM were quickly disassembled and commented by hobbyists. The IBM PC was probably an outlier for the "business world" though.
That's interesting. My first computer was a TI-99/4a, and I don't remember seeing stuff like BIOS listings in the documentation. I don't have any of that stuff anymore, so, I could be wrong.
No, you're right. The TI-99 was notoriously closed and anti-hobbyist. It was my first as well, and I recently got my hands on one again (though a later, reduced-cost beige-case one). It's remarkable how opaque the system is.
Contrast this with my Commodore 128 or TRS-80 Model 100, which ship with schematics, opcode info, etc. and had a healthy book ecosystem around them.
I hit my limits with the relatively weak TI-BASIC, but having nowhere else to go (even with the technical "green manual"), I wonder what I would have gotten up to had there been more there like a built-in assembler. Then again, the thing did ship in 1979 initially and was the first home 16-bit machine (technically).
The ROM listings were usually "third party" efforts, for instance in Germany there was a fairly popular (among home computer enthusiasts) "xxx Intern" book series where most of the book consisted of the annotated ROM disassembly. For instance "CPC Intern":
Atari published the listings of the ROMs for their 400/800 line of computers. They also published the BIOS for the Atari ST (which I helped write, and wrote the documentation for).
In order to avoid copyright violation, Compaq did clean room reverse engineering of the IBM PC BIOS in order to build the first clone. They didn't use any code listing.
IBM top brass considered the PC market something they had to be in that could use to upsell other more expensive IBM kit.
They were very surprised when it became a hit and established a new market but then they lost it when the closed spec MCA (Microchannel Architecture) PC.
The cloners (Dell/Compaq) banded together and did EISA (Extended ISA) in reponse and went their own way.
Afaik it was an internal skunkworks project led by a rather visionary for IBM leader that knew it would be dead if it did things the IBM way.
IBM made the market legitimate and the bus spec was open and anybody could make expansion cards without paying royalties.