On Sunday 31 July 2005 01:05, Paul Davis wrote: > > based on the current layout of jack, writing an application that has an > > effect on all sound streams generated looks a bit hacked: > > of course, because if the user didn't route them through a single point > of control, then there is no single point of control - this is > deliberate, its not an error or an oversight. and here i am, pointing out a disadvantage of this concept :/ > > this mixer application would have to plug itself to the main outs and sit > > idle waiting for new connections. when a new connection connects itself > > to the main outs, the mixer app would disconnect the ports and reconnect > > them to the in-ports of the mixer application. > > "main outs"? what are they? my RME interface has 26 outputs. a delta > 1010 has 12. what are "the main outs" ? exactly these 26/12 outputs are the main outs. the same thing could btw be done with the main ins as well. in terms of the jack api, i could also call them ports with the hardware flag set. most jack programs automatically connect to the first hardware ports available, as far as i mentally gathered. > > instead, what about a final amp routine within jack that reads its amp > > value from a float shared in memory. other apps could read from/write to > > this float, and synchronization does not matter in this case. > > this was discussed at length during the design of JACK and rejected by > almost everyone. "naaaa" "naaaaaaaaaaaaaaaa" "hm ... na." "but..." "SHUT UP!" "ok" :P -- -- leonard "paniq" ritter -- http://www.paniq.org -- http://www.mjoo.org
