Hi Kieren, On Sun, Mar 22, 2015 at 7:39 PM, Kieren MacMillan < kieren_macmil...@sympatico.ca> wrote:
> Hi David, > > > Try the attached out. It will only center an object if the contents of > all staves can reasonably be centered. > > So far, I would call that "perfect”! =) > oh, it will be broken! Anyway, I'm a little unsure of the procedure for determining whether an object is alone in its measure. There's a C++ procedure that determines this and maybe that ought to be made accessible from Scheme. [...] > Any reason why this can’t be submitted as a patch? What bugs me is that the centering is done after measure-widths have been determined. I would imagine that in most cases this routine will end up moving objects to the right. When it does, that is infringing on what LilyPond has decided is an appropriate space between the objects in our column and the column holding the barline (and whatever else). I would think that a good algorithm would ensure that the space after our objects would be increased, to match the space before. Basically, distances should never fall below the ideal distance in either direction, which will involve resizing the measure. Now, I haven't seen results that look bad (in my limited study!) and certain negative outcomes I can imagine could be guarded against. Also, I don't know how feasible the more integrated approach is, whether it would slip nicely into the program as is, or would require extensive changes. --David
_______________________________________________ lilypond-user mailing list lilypond-user@gnu.org https://lists.gnu.org/mailman/listinfo/lilypond-user
