On Tue, May 22, 2007 at 04:32:56PM +0800, zou lunkai wrote: > On 5/22/07, zou lunkai <[EMAIL PROTECTED]> wrote: > >+ /// A set of depth for each frame > >+ // > >+ /// Access to this container is mutex-protected > >+ /// for thread safety. See _frameDepthsMutex. > >+ /// > >+ FrameDepths _frameDepths; > > > >I think the set of depths for a frame would tell us: characters in > >these depths should be kept alive when jump backward and forward. And > >the set of depths might be *changed* by tags(eg. REPLACE and PLACE > >with ratio bigger than zero) in later frames. > > > >In other words: this set will not only dependent on tags before the > >target frame, but also tags *after* the target frame. right? > > > Ah, if it is dependent on tags after the target frame(isn't it?) , > then it means we could hardly get the correct behaviour before a > completely loading of the movie when jumping back.
Right. Good reason for not being dependent. See precedent mail for an alternative handling. --strk; _______________________________________________ Gnash-commit mailing list [email protected] http://lists.gnu.org/mailman/listinfo/gnash-commit
