3 ms·
When I was at Apple (started in '87, about the time the Macintosh II shipped), many of the groups in the systems software and support orgs (development tools, n
by kabdib 3y ago
When I was at Apple (started in '87, about the time the Macintosh II shipped), many of the groups in the systems software and support orgs (development tools, networking, etc.) had a table where there were binders of ROM assembly listings, maybe 2-3 linear feet of paper bound in a metal rack. They were a big help with particularly tricky crashes, or for figuring out ROM bugs, or just learning about the system.
This practice got more and more awkward as the number of Macintosh models ballooned, and was largely abandoned by the early 90s (too many machines, and groups got file servers and ethernet instead of floppy disks and localtalk).
It was fun just reading the code; it was really well-commented, and nearly every line had someone's initials on it (spot a bug? track down the responsible party and talk about a patch...)
Developing software without source control is difficult to imagine now. ("We do have source control, his name is Mike -- just give him a floppy disk with your changes and some instructions on how to integrate them.")
- nxobject 3y agoManaging a common large body of assembler like that, in hardcopy, seems to be a lost art... it reminds me of how the same amount of assembler needed to be managed in hardcopy for the Apollo DSKY/AGC, complete with paper changelogs, line number conventions, etc. [1] But a lot of the bookkeeping there looks like was needed as well to keep track of changes to punchcard decks. [1] https://www.youtube.com/watch?v=-y37tXoBDx0 https://www.youtube.com/watch?v=-y37tXoBDx0