Hans Aberg <hab...@math.su.se> writes: > On 9 Aug 2009, at 21:07, Bernard Hurley wrote: > >>> The default rule is really that one should have a minimum number of >>> time >>> signatures if the repeat construct is expanded*). >> >> That seems like a reasonable rule. But a quick look through my pile >> of full scores (Mostly Dover reprints of 18th and 19th century >> scores) shows it is not always obeyed. I.e. there are plenty of >> cases where the a time signature change is placed after the repeat >> sign even though there is no time change inside the repeat. > > If one of the alternatives of the repeat end in a meter different from > what it begins, then if the expansion rule should prevail, there must > be a time signature at the beginning of the repeat after the |:. And > it is then unnecessary to have one extra before. > > But not even that was followed in the Bulgarian score. It looks like: > 11/16 |: ... | 8/16 ... | 11/16 ... | 8/16 ... :| > >> As a composer/arranger I always put the time change after the repeat >> sign! Rather than talk about the right or wrong way of doing this it >> should just be a user's option. > > > I figure LilyPond might support what people want to typeset.
Yes, if I tell it explicitly. Short of that, Lilypond should pick the way that best reflects the art of typesetting, as recognized by the developers. It is ok if there are manual overrides (and style overrides for a whole document), but the defaults should not require music typesetting taste and choices. -- David Kastrup _______________________________________________ bug-lilypond mailing list bug-lilypond@gnu.org http://lists.gnu.org/mailman/listinfo/bug-lilypond
