On Sunday 26 October 2003 7:19 pm, Guillaume Laurent wrote: > Agreed. I'd be tempted to implement looping at the meta-iterator > level, though not yet (I'd rather try to stabilize things first so we > can have that overdue intermediary release, even if it contains some > well known bugs).
Well, considering we have some pretty horrible bugs as it stands and this stuff is all pretty close to just working I don't see any reason why one of us can't just wade in and fix it. At the moment all three of us seem to be poking it with our foot as we all don't have quite enough information. > On a side note, Rich, where do we stand on the slice size issue ? I > recall you were looking into this. I think larger slices could help > in several issues, though they would also complicate some others > (like stopping playback, or playtime edition). I've voiced my thoughts to Chris over the last couple of days that this slice thing is just a hangover of the old sequencer architecture. While While I've come to this conclusion however, I can't think of a better mechanism in the short term to get events to and from the sound layer without changing the architecture of that - and that would be a huge job and take as back many months. Also I don't think it'd get us anywhere considering the tight coupling between MIDI playback and JACK audio code in AlsaDriver.cpp. AlsaDriver is currently very complicated actually and getting longer and longer (but that's a separate issue). The work I did on slice sizes was to remove the "fetch latency" parameter from the sequencer which has no meaning now. The two remaining latencies are "read ahead" which is the slice size itself and "playback" which is the initial pause on start of playback. The latter is almost immaterial now and the former still forms the basis of our playback. The main issue with increasing slice size is that if you've already parsed events from the meta iterator and pushed them to ALSA then any mods to the GUI will most likely not get picked up as the sequencer's playhead has moved on too far. So yeah, if we stay with current architecture I'd leave these levels the same. Both Chris and I should be able to do a reasonable amount of work on RG this week so I'm hoping we'll get the mmapped stuff pretty much completed including looping and audio between us. It _can_ certainly all work and I'm sure we can work it out all together with judicious use of Messenger or whatever. R ------------------------------------------------------- This SF.net email is sponsored by: The SF.net Donation Program. Do you like what SourceForge.net is doing for the Open Source Community? Make a contribution, and help us add new features and functionality. Click here: http://sourceforge.net/donate/ _______________________________________________ Rosegarden-devel mailing list [EMAIL PROTECTED] - use the link below to unsubscribe https://lists.sourceforge.net/lists/listinfo/rosegarden-devel
