On Nov 24, 2014, at 1:03 AM, Pierre Perol-Schneider <pierre.schneider.pa...@gmail.com> wrote: > Here both are ok; you simply omit to put any bar check: So a | (b c d) is > wrong, and a( | b c d) is right.
So to be clear, I am asking whether the behavior we are seeing is intended for some reason that I haven't understood yet. It is obviously the case that the behavior is as it is. >From the perspective of a user who just wants to write music, there is no >reason at all for me to think that a | (b c d) is wrong and a (b c d) is okay. > For instance, a | b (c d) works, and a (b | c d) works. So it doesn't make >any sense that a | (b c d) doesn't work. It may be that making it work is really hard because of the way bar checks are implemented. If so, it may be that there is no way to make things consistent in the way I am suggesting they should be. I can even see how, from a data structure perspective, making a | (b c d) work consistently is difficult. But if consistency is desired, then the fact that a | (b c d) doesn't work is a bug that is hard to fix, not a feature that doesn't need to be fixed. The question I was asking is whether this is a feature that I just don't understand, but your response isn't really answering that question. If in fact a (b c d) is wrong, but just happens to work, and a( b c d) is right, then it would be helpful for users if a (b c d) threw an error, instead of working, and if a | (b c d) threw the _same_ error. _______________________________________________ lilypond-user mailing list lilypond-user@gnu.org https://lists.gnu.org/mailman/listinfo/lilypond-user
