| drobilla | rgareus: With your patch, I only get vertical centering | 00:04 |
|---|---|---|
| drobilla | rgareus: e.g. jalv.gtk http://gareus.org/oss/lv2/meters#K20mono_gtk | 00:05 |
| drobilla | er, no, not that :) | 00:05 |
| drobilla | Hm, and related problem with jalv, initial window size might be much larger than UI (if e.g. tiling WM), even if resizable. The UI resizes once you resize the window at all, but not initially | 00:07 |
| rgareus | drobilla: the latter is a not a new problem (due to the patch), is it? | 00:21 |
| *** uncle-j_j1 has quit IRC | 00:21 | |
| rgareus | for testing I compiled meters.lv2 with EXTERNALUI=no => x11 GUI. EBUr128 is fixed size. all others are resizable, the needle meters have fixed aspect, too. | 00:23 |
| drobilla | Probably not | 00:23 |
| drobilla | rgareus: in what host? If they're fixed, Jalv window is fixed too so there's no centering | 00:23 |
| rgareus | drobilla: I tested with jalv and ardour | 00:24 |
| * drobilla updates, rebuilds meters | 00:24 | |
| rgareus | don't use the _gtk variant | 00:24 |
| drobilla | I don't get how jalv can demonstrate this particular functionality | 00:24 |
| drobilla | I guess if the UI isn't actually resizable but doesn't set the hints | 00:25 |
| rgareus | K20 is a good exmple. w/o the patch it's left aligned | 00:25 |
| rgareus | Harry's openAV plugins would be great. those were the initial motivation for me looking into this.. | 00:28 |
| rgareus | but they do return a FLTK window - not a X11 window ID :( | 00:28 |
| * drobilla still doesn't really get how that hasn't blown up in anyone's face yet | 00:29 | |
| rgareus | cast to int (x11 windows ID) just a lot of messages from libx11 "invalid window ID" that nobody sees | 00:30 |
| drobilla | K20mono likes to segfault | 00:30 |
| drobilla | oh, right. | 00:30 |
| rgareus | drobilla: it must be you. if that segfault is a valid bug, I'll send you a bottle of wine. | 00:31 |
| drobilla | rgareus: Initial size problem is indeed a side-effect of the patch somehow | 00:31 |
| rgareus | mmh | 00:31 |
| drobilla | *** longjmp causes uninitialized stack frame ***: jalv.gtk terminated | 00:32 |
| drobilla | 00:32 | |
| drobilla | good times | 00:32 |
| * rgareus is about to do a longjmp into bed. | 00:33 | |
| drobilla | rgareus: You appear to be using pango in a thread | 01:40 |
| drobilla | The boom is always related to threaded use of font business | 01:40 |
| *** aombk2 has joined #lv2 | 01:57 | |
| *** aombk has quit IRC | 01:59 | |
| * drobilla gives up getting X11 anything in Qt | 02:11 | |
| drobilla | BadRequest this, BadRequest that. I suspect the display QX11Info gives you is bunk... | 02:11 |
| *** son0p has joined #lv2 | 02:52 | |
| *** bgola has quit IRC | 03:19 | |
| *** bgola has joined #lv2 | 03:20 | |
| *** tytel has joined #lv2 | 03:42 | |
| *** falktx__ has joined #lv2 | 03:43 | |
| *** falktx_ has quit IRC | 03:46 | |
| *** c3r1c3__Win has joined #lv2 | 04:38 | |
| *** c3r1c3_Windows has quit IRC | 04:41 | |
| *** tytel has quit IRC | 05:51 | |
| *** tytel has joined #lv2 | 06:37 | |
| *** tytel has quit IRC | 06:42 | |
| *** edogawa has joined #lv2 | 07:17 | |
| *** falktx__ has quit IRC | 07:22 | |
| *** tytel has joined #lv2 | 07:31 | |
| *** tytel has quit IRC | 07:36 | |
| *** deva has joined #lv2 | 08:13 | |
| *** tytel has joined #lv2 | 08:26 | |
| *** uncle-j_j has joined #lv2 | 08:26 | |
| *** tytel has quit IRC | 08:30 | |
| *** tytel has joined #lv2 | 09:20 | |
| *** tytel has quit IRC | 09:24 | |
| *** tytel has joined #lv2 | 09:52 | |
| *** tytel has quit IRC | 09:57 | |
| rgareus | drobilla: the readme mentions that you need a version of pango that's thread-safe. | 10:18 |
| *** sigma6 has joined #lv2 | 11:54 | |
| *** falktx__ has joined #lv2 | 13:29 | |
| *** uncle-j_j has quit IRC | 13:41 | |
| *** uncle-j_j has joined #lv2 | 13:41 | |
| drobilla | rgareus: If distro ones aren't, which seems to be the case, 99.99999999999999999999999999999999999999999999999999999999999999999999999999999999999999999999% of the userbase won't have one... | 14:35 |
| * drobilla puts $100 on "more trouble than it's worth" without even bothering to look at the code to see why threads are involved | 14:36 | |
| *** ricardocrudo has joined #lv2 | 14:49 | |
| *** ricardocrudo has quit IRC | 14:55 | |
| *** tytel has joined #lv2 | 15:23 | |
| *** ventosus has joined #lv2 | 15:27 | |
| rgareus | drobilla: debian is fine, as is KXStudio. (cairo/pango was made thread-save in Jan'14, but Ub 14.04 14.10 screwed it). I also provide static builds | 15:29 |
| rgareus | drobilla: the issues is that pango is used in both: GUI (idle-thread) and port_event (should be GUI thread, but depends on host) | 15:30 |
| rgareus | the reasoning is: only update value if it actually changes, detect size of changed area, queue partial draw. | 15:32 |
| rgareus | pango is used during port-event to measure the size of the partial exposure. | 15:32 |
| rgareus | (and prepare a cairo-surface, double buffered) | 15:33 |
| falktx__ | do hosts call port_event outside of the main thread? | 15:52 |
| falktx__ | (the 'idle' one) | 15:52 |
| falktx__ | is that even allowed? | 15:52 |
| drobilla | rgareus: ... depends on host? Which host? | 16:28 |
| drobilla | All LV2 UI methods must be called from one thread, namely the GUI one | 16:29 |
| rgareus | drobilla: then I don't see how you could have pango thread issues. | 16:29 |
| rgareus | libxcb maybe. | 16:31 |
| drobilla | Qt does use xcb somehow. I suspect this is why none of my Xlib code works reliably in qt5... | 16:40 |
| *** Magnus_RM has joined #lv2 | 17:26 | |
| *** tytel has quit IRC | 17:33 | |
| *** deva has quit IRC | 18:01 | |
| *** ventosus has left #lv2 | 18:24 | |
| *** rncbc has joined #lv2 | 18:52 | |
| *** tytel has joined #lv2 | 19:21 | |
| *** son0p_ has joined #lv2 | 19:28 | |
| *** Anchakor_ has joined #lv2 | 19:28 | |
| *** Anchakor has quit IRC | 19:30 | |
| *** son0p has quit IRC | 19:32 | |
| *** edogawa is now known as Shaun007 | 19:41 | |
| *** Shaun007 is now known as edogawa | 19:42 | |
| *** aombk2 has quit IRC | 19:48 | |
| *** aombk2 has joined #lv2 | 19:49 | |
| *** uncle-j_j has quit IRC | 20:31 | |
| *** tytel has quit IRC | 20:51 | |
| *** falktx__ has quit IRC | 21:08 | |
| *** tytel has joined #lv2 | 21:11 | |
| *** tytel has quit IRC | 21:33 | |
| *** tytel has joined #lv2 | 21:33 | |
| *** tytel has quit IRC | 21:59 | |
| *** edogawa has quit IRC | 22:00 | |
| *** rncbc has quit IRC | 23:41 | |
| *** Magnus_RM has quit IRC | 23:57 | |
Generated by irclog2html.py 2.13.0 by Marius Gedminas - find it at mg.pov.lt!