Jean-Marc Lasgouttes wrote: > Since the auto-open feature is a `big' change, code-wise (should > auto-open be a property of the cursor or of the inset?)
It's not such a big change. Most of the change is status_ -> status(). > and UI-wise > (do I want closed footnotes to pop-up on me when I do char-forward?), That does not happen. char-forward skips closed footnotes, as it used to be. > I propose a much less ambitious patch. > > I removes André's patch as you did, but for the reason that everything > is done in BufferView::setCursor now. Although I took the occasion to > use the same code that André used (on the premise that he knows what > he does), the main fix in the patch is to remove a use of cursor() > where dit should have been used instead. Without having actually tested your patch, I think this introduces two problems: (1) all insets are forced to status Open (even inlined ERT insets). At least this should be fixed, the manual reversion of this change is too annoying. (2) one of the most-hated behaviour is back: all insets remain open after s&r _and_ spellchecking. We had fixed at least a big part of it (s&r), so the patch is a ui-regression wrt 1.3 (while Alfredo's fixes the problem completely, i.e. also after spellchecking, next-note etc.). > As far as I know this fixes 1921, but I'd be interested to see what > was André's patch supposed to fix, to check that I did not break > anything. André might correct me, but the comment indicates that the patch also aimed at bug 1921. Regards, Jürgen
