rgheck wrote: > Yes, that is right. The new patch doesn't try to ask about that: Vincent > didn't like where I was doing it, and I can't see any other place; there > were Pavel's worries; etc. So this one just asks when the Buffer itself > is destroyed.
I see. Then I misread your description of the patch. > Longer term, I think Abdel's idea, that we should show a list of open > buffers, with check marks, etc, is a good one, but I don't think it > should delay 1.6.4. Yes. For branch, we need an easy solution that fixes the dataloss. > And, as I said elsewhere, even if we implement > Abdel's idea, I think this is yet worthwhile, to try to catch anything > we may have missed. The only thing to keep in mind is that, if the user > chooses not save the buffer, or if for some other reason it ought not to > be saved, then it should be marked clean---clean in the sense that it > doesn't need saving. It occurred to me, for example, that we need to do > this in BufferList::emergencyWrite(): If we fail, then we have to mark > the Buffer clean, so as not to try to save it again when the Buffer is > destroyed. > > So here's the current version. One remaining glitch: If the dirty buffer is a new document, the method should bring up a dialog to enter a proper file name. Now it just saves the newfileX.lyx, which is not very obvious (I first thought the document was lost). Apart from that, it's OK from my POV. Jürgen
