On 20/06/06, D. Michael 'Silvan' McIntyre <[EMAIL PROTECTED]> wrote: > > The rationale is that the program for an instrument is associated with that > > instrument, and that RG's GUI only has one fixed association between > > program and instrument (in the instrument parameters) so the GUI can't do > > anything meaningful with mid-track program changes. > > That is true, which is why I had to argue like hell to convince Rich to allow > mid-track program changes, and why they're still not actually terribly > useful. (Mid-track program changes really need to have a bank as well as > program, but when I looked at adding that, it took me deep into places in > base/ I was not comfortable going. I ignored it, because the only time I > ever actually use a mid-track program change is to switch between GM > instruments, like flute for piccolo, or straight violin for pizz. That > doesn't mean the way we do it is actually right. It's clearly not.) > > > I guess the principle > > is that it may be less confusing overall to refuse to accept mid-track > > program changes than to accept them and transmit them but not reflect them > > anywhere in the GUI. > > But we *would* reflect them in the GUI the same way any other MIDI sequencer > would. In the event list.
If I were coming at this without any history, I would expect that the GUI would let you select a program of "[none]" for a track, which in effect tells RG "I'm doing my own patch changes the hard way, thanks." The selected instrument then only determines the MIDI device and port, no implicit program change is sent. (The user would put the patch name/number on the track name or on the segments. I have to do this anyway because I have gear which doesn't have patch memories... :) _______________________________________________ Rosegarden-devel mailing list Rosegarden-devel@lists.sourceforge.net - use the link below to unsubscribe https://lists.sourceforge.net/lists/listinfo/rosegarden-devel
