From: Andreas D. <and...@go...> - 2011-11-03 21:15:19
|
2011/11/3 hermann <bru...@we...>: > Am Donnerstag, den 03.11.2011, 20:23 +0100 schrieb Andreas Degert: >> I have some little bugs left on my list: >> >> - we have now a popup message on error, so that you don't miss an >> error message (which normally means something didn't work like you >> expected). But now some "unwanted" errors are visible: > > yep, best is to turn them to simple warnings in the logger I hope there are none left now (after last checkin...) >> - when qjackctl is running but jackd not started, autostart by >> guitarix >> does not work well (don't know what to do here, maybe use >> dbus-send to communicate with qjackctl) -> not very important >> > uh, that works here fine, have just test it with the current checkout > 1270 and it work well. Seems to be an ubuntu problem. They made qjackctl a script which calls pasuspender and then qjackctl.real, might be the second pasuspender call.. I'm not inclinded to research that further... > >> - btw. I never tested midi output mode... > midi out put works, but one cant say it is usable professional, I'm > myself have a lot of fun with it, I use it in a "Band in the Box" > meaner, eg, percussion, sitar, bass, etc. with Qsynth. OK, I just didn't want to destroy it inadvertently. >> - strange bug, not so easy to reproduce: >> o rack in horizontal mode, both racks active >> o compiled with nls support (should be disabled by default) >> o german language environment (-> strings are longer) >> o load a preset with phaser or postamp >> -> the window should start to get a little bit wider, then >> screen update and response on mouse and keyboard >> hangs, until e.g. the window is moved a little bit with >> the mouse. Then window updates happen and everything >> works again. >> can someone reproduce this? >> > I must check that, . . seems to be connected with slow network update when using guitarix UI remote, on local display all is well. I think its too much of a fringe case to research further (and sometime after the release UI code will change quite a bit anyhow... :-) Btw. I think we should keep our solution for jack weak functions (and convert jack_set_session_callback usage also to that scheme) because using -fPIC will give other disadvantages and and we are linking to libdl anyhow. Paul now classified it more as compiler / linker bug.. ciao Andreas |