>The frightening list of critical bugs shrunk thanks to the fearless >fighters (Vincent, Enrico, Richard et al.), and I think it's time to >let the users participate from that. > >I have just stamped my OK on some open issues on trac. Apart >from that, I still see the following (critical) bugs: > >#6055 Cant accept changes > >This is something which needs some investigation, I think. This is a >problem of how lyxfind selects found changes, because this method >does not work properly in tabular cells
You probably already know that I don't agree with this conclusion. I think the problem is that we retain "inserted" paragraph breaks at positions where there is no paragraph break. This is inherently wrong. The fact that lyxfind can't find the change correctly is then just a symptom and not the bug itself. I've been trying to prevent those "ghost" paragraph breaks, but I didn't finish it yet. There are still some cornercases left. Moreover, I don't know how the crash can reproduced, I only know that it crashed on me while experimenting with the file attached to the bug. >#6251 Crash while quitting LyX. Happens in lyx::Buffer::getChildren > >I do not see any progress. Yet again, I have done some work on this, but I was a bit too ambitious. I wanted to make a solution that would be intuitive and correct in all very complex cases of documents with numerous childs with different parents in multiple views. I think I can come up with a simple solution that just prevents LyX from crashing. >#6451 Crash when interacting with buffer-dependent dialog with no buffer > open > >This is fixed in trunk by closing all buffer-dependent dialogs when there >are no more buffers open. Is this the correct approach? > According to Abdel it is not. However, it has always been the design that we close inset-dependent dialogs when the inset's document was closed. (I hope I'm not mixing up 1.5.x, 1.6.x and 2.0.0svn here). At some point, we lost this connection, so I think this would be correct for branch. In trunk we now have the new InsetDialog design (and I hope the problem is gone there). >Anything else? I'm really really really bothered by the scrolling issue (bug 6454). My colleagues have been cursing and yelling my name over and over again that the scrolling problems makes their work like hell.. This seems to be introduced in 1.6.5, but I don't really know how that could have happened. Do other people see the problem too (at least Pavel saw it) ? Could you help me to find some reproducible case ? Any ideas how this could be caused by looking at the changes towards 1.6.5 ? I need to have a serious look at this. Vincent
