>>>>> "Michael" == Michael Gerz <[EMAIL PROTECTED]> writes:
Michael> Jean-Marc, thank you very much for stepping forward! I had the impression that nobody was going to do it :) Michael> A consequence of this patch is that the list of authors grows Michael> with every load-buffer-save-buffer cycle. Luckily, hard disk Michael> space is longer an issue :-) This is why we should recycle the authors. This is not very difficult, I think. >> Finally, I think that a policy should be defined in AuthorList to >> reuse empty author entries. >> Michael> With our current implementation, the active user (i.e., the Michael> one editing the document) has always author id = 0. Michael> Unfortunately, our code base does not permit a more flexible Michael> approach without significant code changes. What do you mean? That it is saved in the file with id 0 for all his changes? Then I do not understand what happens when somebody else edits the file... This is precisely an area I do not understand. >> Nevertheless, the patch seems to fit the bill. Please please test >> it, I am not sure at all I did not make mistakes. Michael> I will do so before the weekend. Question: Does your patch Michael> also scan nested insets? I can't see any related code. ParIterator does this for us. JMarc
