| *** ssj71 has quit IRC | 00:09 | |
| *** edogawa has quit IRC | 01:35 | |
| *** NickSB has quit IRC | 02:31 | |
| *** NickSB has joined #lv2 | 02:35 | |
| *** Yruama_Lairba has quit IRC | 03:40 | |
| *** NickSB has quit IRC | 05:27 | |
| drobilla | I guess something like enum { SUIL_ARG_QAPPLICATION, ... } SuilArg; void suil_init(SuilArg arg, void* val, ...) will do for init | 08:22 |
|---|---|---|
| drobilla | Though that one's a bit pointless since there's a static global for it in Qt anyway | 08:39 |
| *** dsheeler has joined #lv2 | 09:17 | |
| *** rncbc has joined #lv2 | 09:21 | |
| drobilla | Hm, shit. Init information needs to actually be stored somewhere | 09:51 |
| drobilla | Which in an alternate universe could maybe have been SuilHost, but this is created late and often several times by hosts to set different functions or whatever | 09:52 |
| drobilla | I guess static data in suil itself is fine, since it's not a module and all | 09:53 |
| *** edogawa has joined #lv2 | 10:05 | |
| *** trebmuh has joined #lv2 | 10:42 | |
| rgareus | let's just move to VST3 | 12:11 |
| drobilla | yeah, no :P | 12:13 |
| rgareus | drobilla: I thought you would you be among the first to like that it includes a sane way to describe *generic UIs*. | 12:14 |
| drobilla | I haven't really looked at it TBH | 12:15 |
| drobilla | Judging by the horror show that is VST controls I'm not exactly hopeful :) | 12:15 |
| rgareus | on the upside they speficy exactly 1 option for custom UIs per OS: HWND (win), NSView (OSX), UIView (iOS), X11EmbedWindowID (Linux) | 12:23 |
| rgareus | not quite true, on OSX they also allow the old "HIView" (but that's compatibility) | 12:24 |
| *** M-badosu has quit IRC | 12:25 | |
| *** unclechu has quit IRC | 12:25 | |
| drobilla | Meh. It's a mistake to conflate what the API allows which what you should actually do, IMO. | 12:27 |
| rgareus | and VSTs iplugview API is very similar to pugl + idle-interface | 12:27 |
| drobilla | There are other options that are totally reasonable | 12:27 |
| drobilla | with* what | 12:27 |
| rgareus | the only thing that VST completely screwed up is UTF16 | 12:27 |
| drobilla | The only way to not screw up UTF16 is to not use UTF16 :D | 12:27 |
| rgareus | heh, indeed | 12:28 |
| *** unclechu has joined #lv2 | 12:44 | |
| *** M-badosu has joined #lv2 | 12:49 | |
| *** yann-kaelig has joined #lv2 | 12:51 | |
| drobilla | woo sunshine. I need to get a bicycle already | 12:51 |
| rgareus | really? a few miles west it's rather cloudy and windy. | 12:53 |
| rgareus | and I have the feeling it'll start to rain soon. | 12:54 |
| drobilla | Okay, well, relatively brief moments of :P | 12:54 |
| * rgareus 's bike is still in the basement. | 12:55 | |
| rgareus | then again you would not call that a bike either.. omafiets | 12:55 |
| drobilla | heh | 12:56 |
| drobilla | Hm.... I thought I had an experiment for basic math operator plugins around here somewhere, but I think it died with my thinkpad :/ | 12:57 |
| rgareus | drobilla: control port maths? | 12:57 |
| rgareus | like https://github.com/x42/controlfilter.lv2 | 12:58 |
| drobilla | More like "any port maths" | 12:58 |
| drobilla | Though I guess if it's all float it might as well be, and I can implement applying such plugins to buffers in the host | 12:58 |
| drobilla | (Queue regular moaning about connect_port) | 12:59 |
| rgareus | ventosus' OpenMusicKontrollers should have some more elaborate /any maths/ plugins. moony.lv2 can probably do "whatever" | 13:00 |
| rgareus | drobilla: ingen? | 13:00 |
| drobilla | I should play with those more | 13:01 |
| drobilla | Atom inspector looks cool :) I wonder about that OSC bit... | 13:02 |
| drobilla | rgareus: Yeah, general motivation being to have a basic toolkit of low level units to build stuff in ingen | 13:02 |
| drobilla | Probably also along with some kind of pd-ish text-based adding thing for [math.sin] or whatever, tab completion would be nice, etc | 13:02 |
| * drobilla never has time to play with the fun things | 13:03 | |
| drobilla | hrm, sherlock crashes in jalv.gtk right off the bat | 13:06 |
| * drobilla grumbles something about fucking cryptic cmake | 13:17 | |
| drobilla | Uses worker, doesn't list it as supported. tsk tsk | 13:22 |
| *** HarryHaaren has joined #lv2 | 13:24 | |
| ventosus | drobilla: you pulled at the wrong moment, just doing some changes today, ttl's were not updated, yet, try again :) | 13:35 |
| drobilla | ventosus: OK. Other problems if I force it to init the worker... | 13:38 |
| drobilla | 10 points for how to get a &*($#@&*($&*( unstripped binary | 13:38 |
| drobilla | ventosus: loads now, anyway, thanks | 13:40 |
| drobilla | Can't see the itty bitty text whatsoever, but don't get me started on the braindead way our entire ecosystem has dealt with DPI :) | 13:41 |
| drobilla | Cool beans, anyway. I need to make some atom assembling things for ingen | 13:42 |
| ventosus | NK_SCALE=2 jalv.gkt ... | 13:42 |
| drobilla | ventosus: thanks | 13:42 |
| ventosus | dynamic resize is on the TODOlist :) | 13:42 |
| drobilla | pugl could probably use some facilities for this | 13:44 |
| drobilla | anyhoo, math... I really hate control ports so much | 13:47 |
| drobilla | But I *guess* to just basically make cmath.h.lv2 that's the pragmatic thing to do | 13:47 |
| drobilla | I thought it'd be quite nice to have plugins that have individual atom ports (literally, not sequences) for "functional" sort of stuff, then have the host or an adaptor layer apply them to sequences of atoms | 13:53 |
| drobilla | But that screws you for CV because you have to copy into the actual atom | 13:53 |
| drobilla | Pop quiz: a plugin says it is logarithmic and has a given min and max but no default value because LADSPA is fucked. What's the default value? | 14:22 |
| HarryHaaren | drobilla: case by case basis? Or do you wanna just set a "default" something? | 14:24 |
| * HarryHaaren would guess 50% of linear space, and then LOG-ify that | 14:24 | |
| HarryHaaren | ps; hey | 14:24 |
| * drobilla shrugs | 14:25 | |
| drobilla | Seems like a thing that would be specified, but apparently not | 14:25 |
| drobilla | (This is a thing in some of fons' plugins, which I ported, which is why lv2bench crashes on cs_chorus) | 14:25 |
| drobilla | also hi :) | 14:25 |
| ventosus | drobilla: what would this atoms contain? some virtual machine op code? [a Tuple [a op:Push 0.5], [a op:Mul], [a op:Push 0.5], [a op:Add]] | 14:31 |
| rgareus | drobilla: pick a random number between min/max it's as good as any | 14:32 |
| rgareus | drobilla: ladspa does have some hints for defaults though | 14:33 |
| *** yann-kaelig has quit IRC | 14:33 | |
| drobilla | ventosus: ........... numbers? | 14:34 |
| rgareus | drobilla: it even specified the relationship between LADSPA_HINT_LOGARITHMIC + LADSPA_HINT_DEFAULT_MIDDLE (and some other LADSPA_HINT_DEFAULT**) | 14:34 |
| drobilla | ventosus: Nice thought, though :) | 14:35 |
| drobilla | rgareus: Yeah, there just isn't any default anything hint | 14:35 |
| drobilla | Which is slightly odd because Fons is pretty particular about correct default values and stuff | 14:35 |
| drobilla | But anyway, I realize that giving NaN to such controls is obviously a bug anyway, so it doesn't matter | 14:35 |
| drobilla | (which is what happens if you just feed the output of lilv_plugin_get_port_ranges_float into them) | 14:36 |
| *** deva has joined #lv2 | 14:40 | |
| rgareus | drobilla: ardour uses "0" in that case (regardless if it's out of bounds) https://github.com/Ardour/ardour/blob/master/libs/ardour/ladspa_plugin.cc#L173 | 14:40 |
| *** yann-kaelig has joined #lv2 | 14:40 | |
| drobilla | rgareus: That's..... one option :) | 14:51 |
| *** NickSB has joined #lv2 | 16:53 | |
| *** rncbc has quit IRC | 17:42 | |
| *** rncbc has joined #lv2 | 17:43 | |
| * drobilla tries to figure out how the hell ingen should behave WRT state save/restore | 18:33 | |
| drobilla | I suppose I will need to maintain some kind of "graph structurally changed since load" thing | 18:33 |
| drobilla | I hate flakey statey flags | 18:33 |
| *** deva has quit IRC | 20:13 | |
| *** Yruama_Lairba has joined #lv2 | 21:12 | |
| *** HarryHaaren has quit IRC | 21:20 | |
| *** yann-kaelig has quit IRC | 21:52 | |
| *** yann-kaelig has joined #lv2 | 21:56 | |
| *** edogawa has quit IRC | 22:30 | |
| *** edogawa has joined #lv2 | 22:34 | |
Generated by irclog2html.py 2.13.0 by Marius Gedminas - find it at mg.pov.lt!