right, but the scale was MUCH smaller. And there's usually someone who know someone who has context, so you had someone to call.
I'm working on a repo these days that's larger than _entire companies_ I worked with in the past. A 10-person team maintaining more code than 100s at a previous job. The _scale_ makes this wildly different from what we're used to, no matter how much "legacy code" you had to maintain before.
I used to work on Visual Studio 20 years ago. At the time it was 30M lines of code with a history going back over a decade and about ~1k people working on it. Product cycles were measured in years. Time spent working on new features was measured in months, since no one understood how it all fit together we then had to spend years playing whack-a-mole with bugs, to stabilize it before shipping it. I regularly had to debug through large parts of the stack I'd never been in before. So did lots of other people.
Being able to work effectively in large amounts of unfamiliar code was an expectation for mid-level engineers (L62 if I recall correctly). Its not a new skill, but it is increasing in value again.
Focusing on the scale feels wrong.
These legacy codebases could be incredibly massive and documenting how to run them alone could be a multi month reverse engineering project.
Now ai can do all that part. Dealing with a legacy or unknown codebase is so much easier…