On 5/26/20, David Kastrup <d...@gnu.org> wrote: > 0.5 would likely result more often in beam thicknesses getting rounded > up sometimes, down at other times, making for an uneven look after > digitisation.
You’re right in theory, but that doesn’t sound like a convincing argument within the LilyPond ideological framework. See for example the debate we had in late 2007 for using slightly rounded barlines so that music scores would appear more nicely in many PDF viewers; that was dismissed by Han-Wen because the point of LilyPond is to produce good print quallity, not to compromise and accommodate improper software implementations. If really needed, it could be .4999 or .500001 or anything, but Valentin’s (heeyyyy) point stands: .48 is perceptibly thinner than most engraving practices. (I actually like his example of .55 much more than I’d have expected myself to.) Valentin (heyy), your question has actually been asked in the past for what it’s worth: https://lists.gnu.org/archive/html/lilypond-devel/2012-01/msg00265.html Cheers, -- (other) V.