On Friday 18 Jul 2003 2:55 pm, Chris Cannam wrote: > Given that most clients will only care about > some sorts of changes (Segments added or removed, for example), we > probably would be better off having the client make those queries > and test just those things it cares about. > > I can take a look at that with the chord-name ruler and the > segment mapper.
OK, I've rather experimentally committed a set of changes that switch the seqeuence manager and chord-name ruler so that they _only_ use the original RefreshStatus API (no change to the API) and the command-history's commandExecuted signal. That seems more like the "correct" way to do things, but I haven't even really tested whether it works properly for the sequence manager yet (you'll be delighted to know). For the sequence manager, we try to get around the not-a-widget-so-no-paintEvent problem by connecting the commandExecuted signal to an update method that fires a user event at the sequence manager, causing it to do the refresh on the next event loop. This is also basically untested so far. Do take a look at sequencemanager.cpp and see if it makes sense to you and go ahead and tell me why it's nonsense. Chris ------------------------------------------------------- This SF.net email is sponsored by: VM Ware With VMware you can run multiple operating systems on a single machine. WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the same time. Free trial click here: http://www.vmware.com/wl/offer/345/0 _______________________________________________ Rosegarden-devel mailing list [EMAIL PROTECTED] - use the link below to unsubscribe https://lists.sourceforge.net/lists/listinfo/rosegarden-devel
