On Mon, Sep 03, 2007 at 08:11:29PM -0500, Bo Peng wrote: > > The only comment I have is that during ordinary development time no 'two > > OKs' are formally needed. > > The problem here is that I have less confidence in this patch. It > looks more like a temporary fix than a formal solution. > > > has a tendency to shorten after committing lots of uncontroversial > > stuff and lenghten as soon controversial stuff went in without proper > > discussion. > > This was what I have done. I split the patch and committed the obvious > part (and left with a uncompilable trunk). > > > > This is not to say that waiting for explicit nods is discouraged or even > > forbidden. > > I will wait for another day. > > > Just go on and try to solve the problems you encounter in a timely > > manner. And try to keep trunk compilable ;-) > > Let me explain the problem again: > > I need a good signal handling mechanism to update the embedding > dialog. Because addition/removal of embedding files only happens to > certain insets, it should be easier than the Toc dialog. This patch > emits the embeddingChanged signal along with structureChanged, and > does not update the dialog *accurately*. For example, if you remove a > figure from a paragraph, the embedding dialog is not updated because > there is no structure change. > > I am not quite sure what to do: emit the signal with buffer edition, > or in the constructor/destructor of some insets. If there is no good > suggestion, I will submit the current patch and revisit this problem > later.
Would making the dialog modal be an acceptable solution? This way the buffer is less likely to change behind the dialog's back... Andre'
