Thomas Morley <thomasmorle...@gmail.com> writes: > Am So., 3. März 2019 um 08:18 Uhr schrieb Andrew Bernard > <andrew.bern...@gmail.com>: >> >> HI James, >> >> This elusive defect comes up a lot. See the User archives. I have seen in >> on Linux also, > > Yep, me too. Though I'm surprised to see the problem on this fairly short > score. > > ############### > > Anyway, what _is_ the problem? > > As far as I can tell, creating lots and lots of new contexts > overcharges lilypond. With some OS earlier, with Linux obviously > later.
If the score is small, the "overcharge" does not appear like a necessity. It probably just increases the probability of a garbage collection happening at an inopportune point of time. We don't really want to have windows of instability like that. But we haven't had something dependable enough to actually trace down the problem that is happening. -- David Kastrup _______________________________________________ bug-lilypond mailing list bug-lilypond@gnu.org https://lists.gnu.org/mailman/listinfo/bug-lilypond
