| *** curlymorphic has joined #lv2 | 00:28 | |
| rgareus | drobilla: just fixed http://dev.drobilla.net/ticket/1080 (patch attached there) | 00:32 |
|---|---|---|
| rgareus | drobilla: the X11-center/resize fix is https://paste.debian.net/311272/ | 00:33 |
| rgareus | drobilla: I didn't yet open a bug report for the X11-resize (well falktx who asked for it didn't :). can you take it from here? | 00:34 |
| *** curlymorphic has quit IRC | 01:11 | |
| *** NickSB2 has quit IRC | 01:45 | |
| *** tytel has joined #lv2 | 02:04 | |
| *** frinknet has quit IRC | 02:09 | |
| *** frinknet has joined #lv2 | 02:13 | |
| *** frinknet has quit IRC | 02:19 | |
| *** tytel has quit IRC | 02:46 | |
| *** tytel has joined #lv2 | 02:47 | |
| *** tytel has joined #lv2 | 02:47 | |
| *** edogawa has joined #lv2 | 05:15 | |
| *** curlymorphic has joined #lv2 | 06:28 | |
| *** curlymorphic has quit IRC | 06:30 | |
| *** tytel has quit IRC | 07:08 | |
| *** tytel has joined #lv2 | 07:08 | |
| *** tytel has quit IRC | 07:33 | |
| *** tytel has joined #lv2 | 07:38 | |
| *** mlpug has joined #lv2 | 07:56 | |
| *** sigma6 has joined #lv2 | 08:24 | |
| *** ventosus has joined #lv2 | 08:44 | |
| *** ricardocrudo has joined #lv2 | 08:45 | |
| *** falktx has joined #lv2 | 08:51 | |
| *** ventosus has quit IRC | 10:18 | |
| *** NickSB2 has joined #lv2 | 10:30 | |
| *** mlpug has quit IRC | 10:46 | |
| *** ventosus has joined #lv2 | 11:50 | |
| rgareus | drobilla, falktx: http://dev.drobilla.net/ticket/1085 | 11:54 |
| rgareus | should probably be "minor" sorry, I just left the default. | 11:55 |
| drobilla | rgareus: thanks | 13:40 |
| rgareus | drobilla: y/w | 13:41 |
| *** ventosus has left #lv2 | 13:43 | |
| *** NickSB2 has quit IRC | 14:05 | |
| *** frinknet has joined #lv2 | 15:04 | |
| *** ventosus has joined #lv2 | 15:10 | |
| *** frinknet has quit IRC | 15:27 | |
| *** son0p has quit IRC | 15:28 | |
| *** NickSB2 has joined #lv2 | 16:18 | |
| *** sigma6 has quit IRC | 16:40 | |
| *** tytel has joined #lv2 | 16:42 | |
| *** LAbot has joined #lv2 | 17:31 | |
| *** falktx has quit IRC | 17:55 | |
| *** rncbc has joined #lv2 | 18:14 | |
| *** HarryHaaren has joined #lv2 | 18:15 | |
| *** NickSB2 has quit IRC | 18:21 | |
| *** ricardocrudo has quit IRC | 18:49 | |
| drobilla | rgareus: WWhat if the X11 UI resizes itself? | 19:00 |
| rgareus | drobilla: using the resize-extension? | 19:02 |
| * HarryHaaren has been reading discussions, nothing to contribute, so have been quiet as a mouse. | 19:03 | |
| rgareus | HarryHaaren: while you're here.. I tried to test with your plugins (they're usually small) | 19:04 |
| rgareus | HarryHaaren: but you do not return a X11 Window ID | 19:04 |
| rgareus | HarryHaaren: openAV-ArtyFX filta/gui/ui.cxx says it's an X11UI but sets LV2UI_Widget * widget to some Widget::Widget (FLTK thingy)C[C[C[C[C[C[C | 19:06 |
| rgareus | HarryHaaren: it should really return a X11 Window | 19:06 |
| drobilla | rgareus: I guess so, which means the host does something, but.. just wondering | 19:07 |
| drobilla | I will probably just blindly apply this regardless, but manually ramming the X11 window to the center of the parent seems messy and I assume basically anything changing size is going to screw it up completely | 19:07 |
| *** HarryHaaren has quit IRC | 19:07 | |
| rgareus | drobilla: I'll test in a minute | 19:07 |
| rgareus | drobilla: meters.lv2 can be compiled as X11UI and they can resize themselves | 19:08 |
| *** HarryHaaren has joined #lv2 | 19:08 | |
| rgareus | drobilla: without resize-extension the "plugged" window resizes. the container does not. | 19:09 |
| * HarryHaaren hates when his battery dies mid-convo | 19:09 | |
| rgareus | drobilla: it's cropped in this case | 19:09 |
| HarryHaaren | ah just caught the not return X11 Window ID, was there more? | 19:09 |
| drobilla | HarryHaaren: How goes AVTK2 anyway? | 19:10 |
| rgareus | HarryHaaren: http://pastebin.com/bkqTjZak | 19:10 |
| drobilla | ... how could returning an FLTK pointer as an X11 handle not have made anything and everything blow up catastrophically? | 19:10 |
| drobilla | rgareus: Propagating that change magically up through Gtk sure would be nice, but never figured that one out | 19:11 |
| rgareus | drobilla: suil detects it' not a valid X11 window and does nothing. FLTK does some magic | 19:11 |
| rncbc | rgareus, drobilla: because suil takes care of it | 19:11 |
| drobilla | I don't think resize extension is really a thing that should exist | 19:11 |
| rgareus | drobilla: x_window_is_valid() fails for harry's openAV-ArtyFX | 19:12 |
| rgareus | HarryHaaren: https://github.com/harryhaaren/openAV-ArtyFX/blob/master/filta/gui/ui.cxx#L82 | 19:12 |
| rncbc | rgareus, drobilla: i mean, suil x11_in_whatever will convert it to a whatever widget | 19:12 |
| drobilla | But........... it's not an X11 window | 19:13 |
| rgareus | rncbc: yeah the host is fine with it. but passing events around fails | 19:13 |
| rncbc | rgareus: events from who and to whom? | 19:14 |
| drobilla | I guess the host just happens to not actually use the thing, but it ends up in the parent because that is passed | 19:14 |
| HarryHaaren | rgareus, i'm not looking at the details right now, thanks for pointing it out - honestly, it works and I was happy. | 19:14 |
| HarryHaaren | Right now I'm working on AVTK, and have ArtyFX ported to that, which is PUGL based, so I presume it gets it right.. | 19:14 |
| HarryHaaren | ;) | 19:14 |
| rgareus | drobilla: it's a magic FLTK construct: https://github.com/harryhaaren/openAV-ArtyFX/blob/master/filta/gui/widget.cxx#L40 | 19:15 |
| HarryHaaren | I've been lacking time, but this evening is an OpenAV evening, I'll try push the ArtyFX based on AVTK with prio, to solve | 19:15 |
| rgareus | drobilla: first thing in it: Fl_Double_Window* o = window = new Fl_Double_Window( | 19:15 |
| rncbc | drobilla, rgareus: yep, suil, takes the x11 window id from the parent, nothing else | 19:15 |
| rncbc | drobilla: btw, any comment on http://dev.drobilla.net/ticket/1084 | 19:17 |
| rncbc | drobilla: sorry for poking again :) | 19:17 |
| rgareus | http://dev.drobilla.net/attachment/ticket/1084/suil-0.8.3svn5719-qt5-2.patch.gz looks funny :) | 19:18 |
| * rgareus can't read gz. | 19:18 | |
| rncbc | rgareus: yeah, i had to gzip it otherwise trac would consider it spam | 19:18 |
| rgareus | falktx might have. he's one of the few waving the QT flag | 19:18 |
| rgareus | HarryHaaren: pugl will get it right, indeed | 19:20 |
| rgareus | HarryHaaren: do you know other people using NTK/FLTK for plugin GUIs like you did that may have this issue? | 19:20 |
| HarryHaaren | win. Then I just need to get my act together and release that | 19:20 |
| HarryHaaren | yeah the Infamous plugins IIRC | 19:20 |
| HarryHaaren | perhaps bsjones' PSI plugins too | 19:20 |
| HarryHaaren | even though I think he's moved to Juce or something since | 19:21 |
| HarryHaaren | so perhaps updated/backported the UI's, not sure | 19:21 |
| rgareus | oh dear. infamous stuck just uses fl_embed( self->ui,(Window)parentXwindow); and does not even set the widget | 19:22 |
| * rgareus checks if suil initializes it with NULL | 19:22 | |
| HarryHaaren | there were a few approaches, none of them ideal for some reason or another IIRC | 19:23 |
| rgareus | surely you can get a X11 window ID out of FLTX | 19:23 |
| rgareus | NTK/FLTK | 19:23 |
| drobilla | rncbc: Only made a cursory glance. I'll apply it this weekend | 19:24 |
| rgareus | oh, oh. | 19:25 |
| rgareus | ssj71: did you never have crashes with your Stuck GUI (in ardour)? | 19:25 |
| rgareus | suil just uses malloc() instance->ui_widget is never initialized. but suil uses it when sending X11 events (keyboard, resize, ..) | 19:26 |
| rgareus | and https://github.com/ssj71/infamousPlugins/blob/master/src/stuck/stuck_ui_main.cxx#L12 never assigns anything to https://github.com/ssj71/infamousPlugins/blob/master/src/stuck/stuck_ui_main.cxx#L12 | 19:26 |
| rgareus | and https://github.com/ssj71/infamousPlugins/blob/master/src/stuck/stuck_ui_main.cxx#L12 never assigns anything to LV2UI_Widget * widget | 19:26 |
| rgareus | granted the X11 window ID is just an integer. and sending X11 events to some random client is OK as long as that client-ID does not exist | 19:27 |
| rgareus | given that there are usually no more than 100 active X11 clients on the chances are slim with a 32bit int. | 19:29 |
| drobilla | So unbelievably sketchy :) | 19:32 |
| rncbc | drobilla: thanks | 19:34 |
| rgareus | drobilla: suil should use calloc, or zero-initialize the ui_widget. | 19:34 |
| drobilla | rgareus: Speaking of sketchy, manually resizing the plug's X11 window... | 19:35 |
| rgareus | drobilla: or we need to track down all plugin authors and | 19:35 |
| rgareus | drobilla: manual resizes (container window) works fine | 19:35 |
| rgareus | drobilla: plugins that don't resize are re-centered, others grow/shrink | 19:36 |
| drobilla | rgareus: Right, misread, you're just getting the display from that | 19:36 |
| rgareus | drobilla: what is not yet done: forward the aspect-constraints | 19:36 |
| drobilla | rgareus: Surely the UI window is already the "right" size anyway though? | 19:36 |
| rgareus | drobilla: yes it is | 19:36 |
| *** tytel has quit IRC | 19:37 | |
| rgareus | all plugins that I've tested call wrapper_resize() which calls gtk_widget_set_size_request() | 19:37 |
| rgareus | hence we don't need a "size-request" handler | 19:38 |
| *** curlymorphic has joined #lv2 | 19:40 | |
| * drobilla wonders why trac doesn't follow lad commits anymore | 19:42 | |
| drobilla | Sigh. So tired of dealing with Trac, but lack any better options | 19:42 |
| drobilla | rgareus: http://dev.drobilla.net/changeset/5723 | 19:44 |
| *** rncbc is now known as rncbc|AFK | 19:47 | |
| *** flexus has joined #lv2 | 19:53 | |
| HarryHaaren | drobilla, do you know patchwork? I've only heard about it recently through work. Its pretty ok. No commenting feature built-in, ML parsing only | 20:06 |
| HarryHaaren | http://jk.ozlabs.org/projects/patchwork/ | 20:07 |
| *** nicksb2 has joined #lv2 | 20:09 | |
| drobilla | A mail backed or git backed thing would certainly be nice, but seems this only does patches, not tickets | 20:13 |
| drobilla | There's debbugs but setting that up doesn't look particularly fun | 20:14 |
| HarryHaaren | no bugtracker integrated no... perhaps a neat feature, instead of [PATCH] do [BUG] and have ML based bug tracking... hm. Not that it is currently capable of it :/ | 20:15 |
| drobilla | bugseverywhere is a git-backed solution, but it seems to be a bit dead now | 20:21 |
| *** tytel has joined #lv2 | 20:27 | |
| *** ventosus has left #lv2 | 20:30 | |
| * drobilla finally has a controller that sends poly aftertouch | 20:38 | |
| *** rncbc|AFK is now known as rncbc | 20:39 | |
| drobilla | Great fun. Now to build some Ingen patches... | 20:39 |
| drobilla | and, y'know, per-note curves in Ardour some time in 2138 | 20:39 |
| drobilla | rgareus: Re: http://dev.drobilla.net/attachment/ticket/1080/commit-d0ae272.diff why have a static wrapper instead of just checking for NULL? | 21:10 |
| drobilla | (also global variables that start with _ are naughty) | 21:10 |
| rgareus | drobilla: did the time-machine I fedexed you finally arrive? | 21:11 |
| rgareus | drobilla: what would be your choice of variable name in this case? | 21:12 |
| drobilla | If I had a time machine proprietary audio software would be obsolete already :) | 21:12 |
| drobilla | rgareus: I dunno, suil_something_something, but my main question is why it needs to exist in the first place since a pointer is returned so a sentinel (NULL) is already available | 21:13 |
| rgareus | drobilla: NULL = no wrapper needed . NULL = error occured | 21:15 |
| drobilla | ah | 21:16 |
| rgareus | drobilla: you could change the API int open_wrapper(..., SuilWrapper* wrapper_to_return); | 21:16 |
| drobilla | The no wrapper needed case should probably just be moved out of open_wrapper | 21:16 |
| rgareus | drobilla: indeed. apart from that all other NULLs are errors | 21:30 |
| rgareus | drobilla: that's a much better solution. | 21:31 |
| rgareus | mmh !container_type_uri is also an error, is it not? | 21:31 |
| *** tytel has quit IRC | 21:32 | |
| drobilla | Yes, but required parameter from caller | 21:32 |
| *** tytel has joined #lv2 | 21:33 | |
| drobilla | odd check indeed | 21:33 |
| drobilla | I guess that could be useful as a way to use suil for convenience but just get the unwrapped handle | 21:34 |
| *** nicksb2 has quit IRC | 21:35 | |
| drobilla | Timestamps in NEWS files are starting to get embarrassing, I need to make releases... | 21:38 |
| *** ricardocrudo has joined #lv2 | 21:50 | |
| * drobilla debates trying to figure out all this modern fancy GLSL business to draw a grid, or just stick with cairo and call it a day | 21:51 | |
| rgareus | drobilla: for what? not all machines support all of GLSL, esp when you want antialiasing etc. | 22:04 |
| * drobilla shrugs | 22:06 | |
| drobilla | Because that's how things are done these days and stuff? | 22:06 |
| drobilla | Plus rounded anti-aliased rectangles don't seem like a thing you can really do with old GL anyway | 22:07 |
| HarryHaaren | GLSL is "too new" for actually using as a cross-platform old-machine-compatible tech | 22:07 |
| * HarryHaaren investigated when doing Avtk | 22:07 | |
| rgareus | yeah that's why I asked "for what?" | 22:10 |
| rgareus | some small private project -> cool | 22:10 |
| rgareus | new LV2 plugin -> better not | 22:10 |
| HarryHaaren | yep. | 22:11 |
| *** ricardocrudo has quit IRC | 22:12 | |
| HarryHaaren | I would like to have a teensy small bezier-drawing lib, that operates on GL with antialiasing, that would remove cairo as a dep from AVTK, and make it PUGL, <GL aa drawing lib name>, and AVTK only | 22:12 |
| HarryHaaren | all source included, no external deps bar platform backend | 22:12 |
| * HarryHaaren dreams | 22:12 | |
| drobilla | Text being a PITA will probably cure me of that urge anyway | 22:12 |
| drobilla | Though I'm generally more inclined to shoot for modern, shiny, and portable (including to mobile) than crusty old UNIX machines these days | 22:13 |
| drobilla | This whole ultra nichey irrelevance thing is getting pretty blatantly unsustainable for me | 22:13 |
| drobilla | This just happens to be a very simple static UI where tinkering with that seems reasonable | 22:14 |
| HarryHaaren | I was intending on manually drawing a font with beziers, removing all font handling away from fontconfig etc too | 22:17 |
| HarryHaaren | and caching it at various sizes / levels of AA if needed | 22:17 |
| HarryHaaren | bla bla, premature optimization etc :) | 22:17 |
| drobilla | lol. | 22:18 |
| drobilla | I can think of very few wheels one should reinvent *less* than drawing text | 22:19 |
| * HarryHaaren might do a plugin spec after doing text ;) | 22:23 | |
| drobilla | In vaguely related news, this makes me wonder how a grid "UI plugin" might talk to the world in something like Ingen | 22:26 |
| HarryHaaren | you're thinking about "canvassing" a UI over an existing graph right? I don't see the use case of eg Roomy needing to understand the Ingen graph.. | 22:29 |
| *** nicksb2 has joined #lv2 | 22:29 | |
| drobilla | No. Literally a plugin that's a grid | 22:29 |
| drobilla | A software launchpad/monome/whatever, essentially | 22:30 |
| drobilla | In some hypothetical future where you could actually 'program' with atoms in Ingen and build stuff, like one might with Max and so on | 22:30 |
| *** tytel_ has joined #lv2 | 22:31 | |
| *** tytel has quit IRC | 22:31 | |
| HarryHaaren | aaaah ok, nice yeah gotcha now | 22:31 |
| drobilla | (triggered by reviving my step sequencer now that I have a Launchpad again) | 22:33 |
| *** nicksb2 has quit IRC | 22:33 | |
| HarryHaaren | cool | 22:35 |
| HarryHaaren | that'd be real cool actually, the Akai APC grid controller i have has been wired up to Buska (OpenAV looping project) and some steps squencing going on, fun | 22:36 |
| *** rncbc has quit IRC | 22:38 | |
| drobilla | Indeed. | 22:49 |
| drobilla | Ultimately I want to use this thing as a control center for everything, so I'm not sure how an independent step sequencer will fit in to all that (maybe switching in software), but I'll revive that thing and get it working solidly at least | 22:49 |
| drobilla | 8x8 sucks for sequencing but what can you do... | 22:50 |
| HarryHaaren | 8x5 sucks more :/ | 22:50 |
| HarryHaaren | 32 step 8x4 with one line of 8 "random function" buttons it is then | 22:50 |
| HarryHaaren | anyways, have fun with it, I'll keep you in the loop if I get anything funky going with it | 22:51 |
| * HarryHaaren notes its almost 12 here, bedtime. Laters guys | 22:51 | |
| *** HarryHaaren has quit IRC | 22:54 | |
| *** tytel_ has quit IRC | 23:00 | |
| *** edogawa has quit IRC | 23:16 | |
| *** flexus has quit IRC | 23:28 | |
| *** tytel has joined #lv2 | 23:45 | |
Generated by irclog2html.py 2.13.0 by Marius Gedminas - find it at mg.pov.lt!