> Then instead of testing all of that shit every
time we built something from source, we'd just drag in the pre-existing results and
go from there. It's not like the results were going to change on us. They were a
reflection of the way the kernel, C libraries, APIs and userspace happened to work. Short
of that changing, the results wouldn't change either.
To build a set of objects you need to worry about at least the following:
- build recipes for each of them (which may also depend on other things)
- configuration parameters
- dealing with differences on each platform
- third party libraries & alternatives
- toolchains (& may be cross-platform builds)
- supporting/navigating different versions of the last 3 above
You can't really precompute all this as there are far too many combinations and they
keep changing.
Though you may be able to train a program porting AI model :-)