[EMAIL PROTECTED] writes: > On Mon, 13 Aug 2007, Abdelrazak Younes wrote: > >> Look at loadChildDocuments(). It's all in the commit logs. > > Being somewhat lazy and ignorant, is there a link to the commit logs?
And you will learn they say * Buffer: load all child documents in one go where it makes sense. This has the advantage to call updateLabels() only once for the master buffer and to get rid of the LyXView dependency. We can think of maintaining a child document list in the future. Now it is up to you to guess where it makes sense. Is the intent that all child buffers are always loaded? Was the impact on really large files been evaluated? Are previews computed on such un-shown child documents? Why have the documents to be scanned once more instead of doing recursive loading directly in Buffer::read? I am ambivalent on this thing, actually. JMarc
