Federico Bruni <[email protected]> writes: > Can you announce here the release of 2.16.1 at least a week before the > actual release? > So translators will have time to complete their works. > > I have a patch which will be ready next week and I hope it won't miss > next stable release.
No danger. By far the most important regressions requiring a release of 2.16.1 are the line_count fixes, and those have not been completed and have not seen exposure in an unstable release yet. I don't see the point in releasing 2.16.1 without them, so there are still at least several weeks of buffer. In the mean time, I am investigating garbage collection problems: the "needs-work" problems of issue 2815 <URL:http://code.google.com/p/lilypond/issues/detail?id=2815> point to parts of the parse stack keeping marked objects around after their proper logical demise, possibly in areas allocated with alloca and considered "released" by Bison's parse stack management: this might have always been the case, or be related to other fixes I want to cherry-pick because of performance reasons, but the patch in 2815 might exacerbate it. We also have no clue about the GhostScript-related problems, but it does not appear like there is active work going on, so that is, somewhat unfortunately, not something considered delaying 2.16.1. So the collection and selection of patches into the stable branch is proceeding quite leisurely at the current point of time. I don't expect a release of 2.16.1 before end of September. -- David Kastrup _______________________________________________ lilypond-devel mailing list [email protected] https://lists.gnu.org/mailman/listinfo/lilypond-devel
