Stefan Monnier <monn...@iro.umontreal.ca> writes: >> Having said that, the current implementation is very dumb (it copies the >> entire buffer every keypress) and it seems to behave fine. > > Of course, this introduces a significant performance problem, tho it may > only rear its ugly head in some rather than all cases.
Yes. Of course, this is improvable -- before and after-change-functions give me enough data to make it work better, although it will make the cloning functions harder to write. It is, however, fast enough for me to test whether it works from a user interface POV which was what I needed. I was rather surprised, but it appears to cause no noticable lag when I type even on my netbook. > But it also breaks the mark-ring, AFAIK. And narrowing, and lots of things, I would have thought. Phil _______________________________________________ gnu-emacs-sources mailing list gnu-emacs-sources@gnu.org https://lists.gnu.org/mailman/listinfo/gnu-emacs-sources
