Comment #2 on issue 1477 by percival.music.ca: suppress output for expected
warnings
http://code.google.com/p/lilypond/issues/detail?id=1477
For regtests:
- some people might do "make test" and think that something is wrong when
they see 500+ lines of errors and warnings.
- the regtest comparison is quite verbose, and generally contains tons of
false "warnings". To quantify this, the last three regtest comparisons
(.50, .49, and .48) contained 18 useless and 3 useful text diffs (4vs1,
6vs1, 8vs1 respectively). Admittedly, those "useless diffs" are already
reported in other issue numbers.
For docs:
- some people might do "make doc" and think that something is wrong then
they see over 2000 lines of errors and warnings.
I think it would be nice if we could tell programmers and doc writers "if
your patch produces an error message when compiling regtests (and maybe
even docs), don't push it". But clearly that's not possible if we already
have so many warnings -- even if it would be technically possible (by
keeping a complete log of the previous doc build and doing a diff?), I
think it would be horribly unfair to new contributors to insist that their
work prints no errors when the existing code-base prints so many errors
already.
_______________________________________________
bug-lilypond mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/bug-lilypond
