Am Sonntag, 7. August 2011, 11:11:13 schrieb Phil Holmes: > There's no intention of stopping make showing errors. There is an issue > with warnings - make doc produces so many that the output is pretty much > unreadable, and they've been ignored for a long time - so having warnings > turned off by default is the preferred option. We should allow them to be > turned back on, though.
For a documentation build, I might agree (some of the warnings are desired, like in some regtests, while others are harmless), but for a build of the lilypond binary, hiding compiler warnings is a really bad idea, as they typically point to grave coding errors (e.g. variable shadowing, unused parameters, signed/unsigned messups, etc.). Many of them will directly cause bugs. Cheers, Reinhold -- ------------------------------------------------------------------ Reinhold Kainhofer, [email protected], http://reinhold.kainhofer.com/ * Financial & Actuarial Math., Vienna Univ. of Technology, Austria * http://www.fam.tuwien.ac.at/, DVR: 0005886 * LilyPond, Music typesetting, http://www.lilypond.org _______________________________________________ lilypond-devel mailing list [email protected] https://lists.gnu.org/mailman/listinfo/lilypond-devel
