make better run smarter [feature] (#11709)

[AdrianV]

IMHO the new better run behavior with nim c -r feature is too optimistic now.

For example there is no recompile when I change from -d:debug to -d:release or when I change the compiler or cpu target or compiler parameters.

I would suggest that these parameters should be tracked too.

[timotheecour]

• that's indeed more on the line of what I was suggesting in [tooling] cache compilation to avoid recompiling when possible RFCs#107:

A1: when nim c -cache foo.nim foo bar gets run, nim creates a hash string hashid of all compiler flags (including cmd line ones and implicit ones from nim.cfg, config.nims, but excluding program arguments given after foo.nim, these are not cached so we can rerun with different such arguments)
... etc
# then see if hash exists otherwise create new one and garbage collect via LRU to keep disk usage low

this prevents nim c -d:foo --stacktrace:on bar.nim to invalidate what was cached by nim c -d:baz bar.nim for example; both are kept, until garbage collection kicks in

also I'd prefer this cache to be triggered by a separate --cache flag, not by the existing -r flag, for several reasons:

• separation of concerns
• the new -r (as shown above) is still too optimistic, and using nim c -r -f is not equivalent to the old nim c -r since -f will do more work
• sometimes, user wants to run nim c -r foo separately from nim c foo followed by ./foo (eg for logging separately compiler output and runtime output, or to avoid Error: Spawning of process failed caused by choosenim using a wrapper process (shim) instead of symbolic link dom96/choosenim#126), but that would prevent them from using the smarter caching feature