| *** tytel has quit IRC | 00:22 | |
| *** tytel has joined #lv2 | 00:22 | |
| *** tytel has quit IRC | 00:22 | |
| *** curlymorphic has quit IRC | 00:30 | |
| *** tytel has joined #lv2 | 00:39 | |
| *** tytel has quit IRC | 00:50 | |
| *** tytel has joined #lv2 | 01:14 | |
| *** nicksb2 has joined #lv2 | 02:31 | |
| *** tytel has quit IRC | 02:48 | |
| *** nicksb2 has quit IRC | 04:19 | |
| *** tytel has joined #lv2 | 04:40 | |
| *** deva has joined #lv2 | 06:46 | |
| *** tytel has quit IRC | 07:33 | |
| *** tytel has joined #lv2 | 07:36 | |
| *** edogawa has joined #lv2 | 07:39 | |
| *** rncbc has joined #lv2 | 08:12 | |
| *** tytel has quit IRC | 08:14 | |
| *** ventosus has joined #lv2 | 08:21 | |
| *** tytel has joined #lv2 | 08:32 | |
| *** nicksb2 has joined #lv2 | 09:05 | |
| *** tytel has quit IRC | 09:55 | |
| *** edogawa has quit IRC | 10:20 | |
| *** ricardocrudo has joined #lv2 | 10:30 | |
| *** nicksb2 has quit IRC | 12:07 | |
| *** ventosus has left #lv2 | 12:10 | |
| *** rncbc is now known as rncbc|AFK | 12:11 | |
| *** son0p has joined #lv2 | 13:12 | |
| *** son0p has quit IRC | 13:44 | |
| *** son0p has joined #lv2 | 13:45 | |
| *** rncbc|AFK is now known as rncbc | 13:54 | |
| drobilla | rncbc: gtk2_in_qt5 doesn't seem to work for me | 14:48 |
|---|---|---|
| rncbc | drobilla: hi. how are you testing? | 14:50 |
| drobilla | rncbc: I made a jalv.qt5 | 14:51 |
| drobilla | rncbc: X11 works though | 14:51 |
| rncbc | drobilla: which gtk2 plugin? | 14:51 |
| drobilla | rncbc: tried meters, calf | 14:52 |
| drobilla | rncbc: I'll just commit this stuff and you can see | 14:53 |
| rncbc | rncbc: here they work | 14:53 |
| rncbc | drobilla: ^ | 14:53 |
| rncbc | drobilla: one note is that embedding might not really work as with qt4 | 14:54 |
| rncbc | drobilla: you're probably re-parenting the QWidget* returned by gtk2_in_qt5, in jaclv.qt5, right? | 14:55 |
| rncbc | jalv.qt5 | 14:55 |
| drobilla | Same code as qt4. win->setCentralWidget(widget) | 14:58 |
| rncbc | drobilla: so you get two windows, one empty blank and another top-level window with the plugin gui ? | 15:11 |
| drobilla | No, just the one window with nothing in it | 15:21 |
| rncbc | drobilla: and what if you replate setCentralWidget(widget) by widget->show() ? | 15:24 |
| rncbc | drobilla: just 4 testing | 15:24 |
| drobilla | rncbc: Then I get the UI in a separate window as you describe | 15:25 |
| *** falktx has joined #lv2 | 15:25 | |
| rncbc | drobilla: ok. probably the widget has to be re-parented in some other way than setCentralWidget() | 15:26 |
| rgareus | scandal: cute parent throws adopted child out of the window. | 15:34 |
| rncbc | drobilla: try this way -> http://pastebin.com/yhTC2U74 | 15:41 |
| rncbc | drobilla: nb. ^replaces setCentralWidget(widget); on jalv.qt5 | 15:42 |
| drobilla | rncbc: what is 'this' here? | 15:55 |
| rncbc | drobilla: this = main widget | 15:56 |
| drobilla | my main/only widget is a window, but you have a separate window | 15:56 |
| drobilla | mm, n/m, that's the widget's | 15:57 |
| rncbc | main widget = main top-level window | 15:57 |
| drobilla | Doesn't draw contents at all (including background) | 15:59 |
| drobilla | http://dev.drobilla.net/changeset/5725 | 16:03 |
| drobilla | 16:03 | |
| drobilla | http://dev.drobilla.net/changeset/5726 | 16:03 |
| rncbc | drobilla: maybe this simpler way -> http://pastebin.com/1tZqKMcd | 16:04 |
| rncbc | drobilla: ok. thanks | 16:06 |
| falktx | rncbc: do you know of anything to do the embeding the other way? (qt5 stuff in x11/gtk) | 16:07 |
| rncbc | falktx: nope but qt5 should be similar to x11 via QWidget::winId() gives you the (x11) Window id and should be handled by gtk2 window etc. thereafter. | 16:11 |
| falktx | except that doesn't work quite well | 16:16 |
| falktx | it's the only thing preventing carla from going qt5 on linux :( | 16:17 |
| rgareus | maybe a stupid question: Can suil qt4_in_... and qt5_in_... be compiled at the same time in one build? | 16:17 |
| falktx | rncbc: with this suil change will you start recomming using qt5 for plugins + qtractor? | 16:18 |
| falktx | rgareus: what about linkage? | 16:18 |
| falktx | rgareus: the symbols are very similar | 16:18 |
| rgareus | falktx: strictly speaking they're independent. so qt4_in_.. would be compiled and linked with qt4 libs, qt5_in_... with qt5 libs | 16:18 |
| rncbc | falktx: maybe in mid-november | 16:18 |
| rgareus | falktx: I think the problem is rather pkg-config and the build-environment. Include-files, compiler-flags,... which may share the same name | 16:19 |
| rgareus | jalv would have the same issue, I expect | 16:20 |
| rncbc | rgareus: problem is that qt5-plugins crash qt4-host and vice-versa | 16:20 |
| rgareus | rncbc: right, but a gtk host could load either. | 16:20 |
| rgareus | as long as it does not load both | 16:20 |
| falktx | no | 16:21 |
| falktx | gtk3 hosts can't load qt4/5 stuff | 16:21 |
| falktx | because qt has a gtk2 theme | 16:22 |
| rncbc | rgareus: wrong. it can only load one kind; you cannot load libQt4Gui and libQt5Widgets loaded in same process address-space | 16:22 |
| rgareus | rncbc: as I said a gtk host could load either .. as long as it does not load both. | 16:22 |
| rncbc | rgareus: right | 16:22 |
| rgareus | and as falktx pointed out gtk2 hosts onlys | 16:23 |
| rgareus | falktx: do you know what motivated QT to use a GTK2 theme ? | 16:23 |
| falktx | it's the old code | 16:23 |
| falktx | hasn't been updated to gtk3 because theming in gtk3 is a horrendous mess | 16:23 |
| falktx | also latest gtk3 require theming via static css files afaik | 16:24 |
| rgareus | what a mess. I long for the day that all GTK and QT plugin GUIs vanish. | 16:24 |
| rgareus | and FLTK, too | 16:24 |
| falktx | and NTK | 16:24 |
| falktx | sorry spenser.. | 16:24 |
| rgareus | NTK is not unresonable, were it portable | 16:24 |
| falktx | ntk has the same issues that fltk has, but at least supports embedding | 16:25 |
| rgareus | it's pretty much cairo + basic layout packing. | 16:25 |
| drobilla | Is "resizeable but always square" a thing? on all/which platforms? | 16:25 |
| falktx | using a menu or calling a dialog in ntk will block the main thread | 16:25 |
| falktx | qt/gtk also block, but qt re-uses the glib event handler or something | 16:26 |
| falktx | somehow it works out even when the call is blocking | 16:26 |
| rgareus | drobilla: just set a 1:1 aspect ratio. pugl supports that on all platforms,. | 16:26 |
| rgareus | drobilla: at least falktx' and my fork of pugl :( | 16:27 |
| * rgareus used a dirty hack though, compatible with the old API min_width != width means retain aspect. | 16:29 | |
| *** nicksb2 has joined #lv2 | 16:30 | |
| drobilla | That's probably the reason for it not moving upstream, but I forget | 16:34 |
| drobilla | I merged almost all your stuff, but should have kept a record, because most had to be done manually so git doesn' tknow | 16:34 |
| *** nicksb2 has quit IRC | 16:34 | |
| drobilla | rgareus: seems to be x11 only | 16:39 |
| rgareus | drobilla: windows works as well. I remember testing that. OSX I don't remember | 16:40 |
| *** ventosus has joined #lv2 | 16:43 | |
| * drobilla discovers that resize with pugl+cairo doesn't work at all | 16:43 | |
| *** tytel has joined #lv2 | 16:49 | |
| drobilla | ... and wow, is double buffering sorely needed there | 16:53 |
| drobilla | I have a big patch for event stuff on Windows, too, but the key stuff is completely cracked out and I don't know how to properly do unicode text input there | 17:08 |
| *** nicksb2 has joined #lv2 | 17:20 | |
| *** nicksb2 has quit IRC | 17:24 | |
| *** ricardocrudo has quit IRC | 18:36 | |
| rncbc | drobilla: new patch -> http://dev.drobilla.net/ticket/1086 | 18:58 |
| * rncbc -> diner | 18:59 | |
| *** rncbc is now known as rncbc|AFK | 18:59 | |
| *** ventosus has left #lv2 | 19:06 | |
| drobilla | Hmm.. everything I can find online suggests that XPending is blocking, but my loops chew 100% CPU regardless of any events being present | 19:31 |
| drobilla | But this raises a pugl API question... both blocking and non-blocking (and/or blocking + timeout?) might be desired | 19:33 |
| *** NickSB2 has joined #lv2 | 19:34 | |
| drobilla | Some stuff doesn't work if I make it block by using XNextEvent straight away, though. | 19:39 |
| *** rncbc|AFK is now known as rncbc | 19:56 | |
| *** flexus has joined #lv2 | 20:01 | |
| *** edogawa has joined #lv2 | 20:28 | |
| *** falktx has quit IRC | 20:28 | |
| *** nicksbx has quit IRC | 20:51 | |
| *** deva has quit IRC | 20:52 | |
| *** deva has joined #lv2 | 21:12 | |
| *** rncbc has quit IRC | 21:17 | |
| *** deva has quit IRC | 21:37 | |
| *** tytel has quit IRC | 21:38 | |
| *** tytel has joined #lv2 | 21:40 | |
| *** nicksby_ has joined #lv2 | 21:51 | |
| *** tytel has quit IRC | 22:22 | |
| *** tytel has joined #lv2 | 22:38 | |
| drobilla | Hrm. Well, I added a separate puglWaitForEvents() so you can just stick that in your loop.. but I can't find such a thing for OSX | 23:22 |
| drobilla | Surely it must be possible to block waiting for an event somehow | 23:22 |
| *** flexus has quit IRC | 23:30 | |
| *** edogawa has quit IRC | 23:39 | |
| *** nicksbx has joined #lv2 | 23:45 | |
| *** nicksby_ has quit IRC | 23:45 | |
Generated by irclog2html.py 2.13.0 by Marius Gedminas - find it at mg.pov.lt!