On Aug 25, 2011, at 9:08 PM, Mike Solomon wrote: > On Aug 25, 2011, at 3:17 PM, Han-Wen Nienhuys wrote: > >> On Thu, Aug 25, 2011 at 7:03 AM, Mike Solomon <mike...@ufl.edu> wrote: >> >>> I have verified via pacifier prints that the pure heights of the flag are >>> being taken from the stencil function and are going into the spacing engine >>> in separation-item.cc. However, in theory, this patch should have null >>> effect on the minimum and ideal distances of paper columns. This is not >>> the case (see attached). >> >> ? it's not the heights but the widths that should go into separation-item? >> >> Have you verified that the new flag grob gets taken into account in >> note-spacing.cc ? >> > > I found the problem after a couple hours of banging my head against the wall. > > The Stem::width function uses Stem::flag to get the flag, whereas the stencil > uses Stem::get_translated_flag. Thus, the flag used for widths is not > translated. As the flag's bounding box in the font is shifted by a half stem > length to the right of where the glyph actually is, and as the translated > flag was always shifted by half a stem width, this is what made it work > (albeit kludgily). > > I'll work this into the patch tonight or tomorrow. > > Cheers, > MS
Hey all, I've uploaded a new patch that reproduces the spacing calculations from the current master with a TODO item to fix this. I'm comfortable pushing this (after running the regtests again) in its current form even though it does not fix this problem - it simply transfers it to the new code. I would like to fix the problem, though, in the not-too-distant future. Cheers, MS _______________________________________________ lilypond-devel mailing list lilypond-devel@gnu.org https://lists.gnu.org/mailman/listinfo/lilypond-devel
