From: Dave R. <dro...@co...> - 2005-04-27 04:16:53
|
On Tue, 2005-26-04 at 19:32 +0200, Lars Luthman wrote: > On Mon, 2005-04-25 at 10:48 -0700, Sean Bolton wrote: > > On Sun, Apr 24, 2005 at 10:37:43AM -0700, Sean Bolton wrote: > > Ah, so you're needing the (non-JACK-specific) transport _position_ > > information, not the JACK transport _state_ information, is that > > correct? > > Yes, I was thinking basically of the type of information that a > jack_position_t contains (although all of it is probably not needed). > The BPM value would definitely be needed for effects that want to set > delay times as fractions or multiples of beats, and the number of beats > per bar could be useful for a synth plugin with a builtin arpeggiator. > The actual position information (current bar, beat, frame number) might > not be as necessary, although there probably are situautions where they > could be useful too. > > This information could be packed into a struct and a pointer to the > struct could be passed as a fifth parameter to the run_synth() > functions. I've been thinking lately that maybe just using Jack headers straight-up wouldn't be the worst thing in the world, particularly once Jack MIDI shows up. DSSI already kindasorta depends on the Alsa headers. If we just used the Jack headers, said "yeah, then /all/ the transport information is there to use, and MIDI as well - plus, this would facilitate plugin MIDI output by following Jack's model of MIDI ports. Gasp! Shock! Horror! etc, I know... but now that Jack is becoming truly cross platform, I'm not so sure plugins depending on it is a bad thing.... and we'd only really need the headers anyway. It just seems silly to me to be taking headers from other projects and slightly mangling them for inclusion in DSSI. I'm not actively suggesting this or anything, just a thought... -DR- |