Hi Werner, I've now set up a little comparison test (see attachment). On the left, you can see three flags (8th, 16th and 32th) with a tight surrounding box in a markup \column so that the slight differences in width are clearly visible. On the right, there is one bar of quavers with alternating stem directions and maximum horizontal compression (I used system-count = 1 and line-width = 10).
Above, there is the unified width version with stemthickness# / 2 removed, and below, you can find the same example with the original widths: flag-widths.png <http://lilypond.1069038.n5.nabble.com/file/t3887/flag-widths.png> The red guiding lines demarcate the slight differences in spacing. After all, I agree to go for unified widths because there is no reason at all why tails of quavers/semiquavers should be treated differently from all the rest. All the best, Torsten -- Sent from: http://lilypond.1069038.n5.nabble.com/Dev-f88644.html _______________________________________________ lilypond-devel mailing list lilypond-devel@gnu.org https://lists.gnu.org/mailman/listinfo/lilypond-devel