| *** falktx has joined #lv2 | 00:00 | |
| rgareus | dsheeler: I do contract work for harrison (working on ardour) | 00:00 |
|---|---|---|
| dsheeler | nice. you deserve some $$$ for that work ! | 00:02 |
| *** gianMOD has joined #lv2 | 00:21 | |
| *** frinknet has quit IRC | 00:53 | |
| *** falktx has quit IRC | 01:01 | |
| *** falktx has joined #lv2 | 01:04 | |
| *** gianMOD has quit IRC | 02:03 | |
| *** Spark[01] has quit IRC | 02:14 | |
| *** gianMOD has joined #lv2 | 03:13 | |
| *** gianMOD has quit IRC | 03:18 | |
| *** ssj71 has quit IRC | 03:31 | |
| *** falktx` has joined #lv2 | 03:38 | |
| *** falktx has quit IRC | 03:42 | |
| *** gianMOD has joined #lv2 | 03:53 | |
| *** gianMOD has quit IRC | 03:53 | |
| *** gianMOD has joined #lv2 | 05:41 | |
| *** gianMOD has quit IRC | 05:46 | |
| *** gianMOD has joined #lv2 | 07:30 | |
| *** gianMOD has quit IRC | 07:34 | |
| *** edogawa has joined #lv2 | 07:51 | |
| *** gianMOD has joined #lv2 | 07:54 | |
| *** gianMOD has quit IRC | 07:59 | |
| *** ensonic has joined #lv2 | 08:59 | |
| *** ricardocrudo has joined #lv2 | 09:22 | |
| *** falktx` has quit IRC | 09:36 | |
| *** ricardocrudo has quit IRC | 09:41 | |
| *** ricardocrudo has joined #lv2 | 09:55 | |
| *** ensonic has quit IRC | 10:22 | |
| *** falktx has joined #lv2 | 10:44 | |
| *** gianMOD has joined #lv2 | 11:01 | |
| *** falktx has quit IRC | 11:07 | |
| *** falktx has joined #lv2 | 11:08 | |
| *** falktx has quit IRC | 11:15 | |
| *** falktx has joined #lv2 | 11:19 | |
| *** gianMOD has quit IRC | 12:52 | |
| *** unclechu has joined #lv2 | 13:06 | |
| *** unclechu has quit IRC | 13:11 | |
| *** unclechu has joined #lv2 | 13:12 | |
| *** gianMOD has joined #lv2 | 13:17 | |
| *** son0p_ has quit IRC | 13:25 | |
| *** son0p_ has joined #lv2 | 13:27 | |
| *** gianMOD has quit IRC | 13:34 | |
| *** gianMOD has joined #lv2 | 14:52 | |
| *** dsheeler has quit IRC | 15:16 | |
| *** gianMOD has quit IRC | 15:32 | |
| *** dsheeler has joined #lv2 | 15:43 | |
| *** gianMOD has joined #lv2 | 15:59 | |
| *** gianMOD has quit IRC | 16:09 | |
| *** fredex has joined #lv2 | 16:27 | |
| fredex | hi guys, what is the correct lv2 plugin dir on x86_64? | 16:28 |
| drobilla | Depends on distribution and/or personal philosophy | 16:31 |
| drobilla | IIRC Redhat is of the opinion that /usr/lib is 32-bit until the end of time, others (including Debian) think /usr/lib should be native (i.e. 64 bit in this case) and /usr/libnn reserved for other stuff (32 bit in this case) | 16:33 |
| fredex | well would it be /usr/lib64/lv2 then? | 16:33 |
| fredex | on 64vbit | 16:33 |
| drobilla | Well, on Fedora. | 16:33 |
| drobilla | But /usr/lib64 doesn't exist on amd64 for many distros | 16:33 |
| drobilla | (Yes, this sucks) | 16:34 |
| fredex | the is slackware multilib | 16:34 |
| fredex | I have /usr/lib and /usr/lib64 | 16:34 |
| fredex | will they be seen wherever I put them? | 16:34 |
| fredex | is what I'm trying to discover | 16:34 |
| drobilla | That is controlled by LV2_PATH | 16:34 |
| fredex | hmm ok | 16:35 |
| drobilla | The default is compiled into lilv and settable at configure time | 16:35 |
| fredex | ok | 16:35 |
| fredex | I'll need to look at lilv then | 16:35 |
| drobilla | You can just set LV2_PATH in your environment to whatever you want. It works like PATH. | 16:36 |
| fredex | I see | 16:36 |
| fredex | ok thanks | 16:36 |
| * drobilla personally thinks lib64 on 64-bit is completely insane, particularly since the preexisting body of things that just install to PREFIX/lib is absolutely huge, being, y'know... the obvious thing to do everyone's been doing for 30 years | 16:37 | |
| *** fredex has quit IRC | 16:43 | |
| *** gianMOD has joined #lv2 | 17:00 | |
| *** Spark[01] has joined #lv2 | 17:02 | |
| rgareus | drobilla: you should know that there's http://lv2plug.in/pages/filesystem-hierarchy-standard.html | 17:19 |
| rgareus | if distro's don't follow the standard... not our problem. | 17:19 |
| rgareus | likewise http://refspecs.linuxfoundation.org/FHS_3.0/fhs-3.0.html if distros don't follow specs use /lib64 .. not our problem | 17:22 |
| drobilla | Meh. If they wanted everybody to follow the FHS they shouldn't have made it stupid :P | 17:30 |
| drobilla | That's what happens when you let Redhat in cahoots with proprietary software companies and suits influence your specs | 17:31 |
| drobilla | (Though IIRC it doesn't mandate the stupid way, it's just ambiguous, leading to the current situation) | 17:32 |
| drobilla | But yeah, the LV2 FHS says /lib | 17:32 |
| drobilla | We haven't really dealt with multilib at all, though | 17:32 |
| drobilla | In practice it's not always so | 17:33 |
| drobilla | Someone once made the suggestion that we can avoid having to mark up platforms (nightmare) by just listing a bunch of lv2:binary properties and hosts can just try them all until one works. | 17:34 |
| drobilla | I still wonder if this is actually okay despite how truly terrible of an idea dlopen(thing_for_wrong_platform) seems | 17:34 |
| *** gianMOD has quit IRC | 17:40 | |
| *** gianMOD has joined #lv2 | 17:51 | |
| *** drobilla` has joined #lv2 | 17:57 | |
| *** drobilla has quit IRC | 17:58 | |
| *** gianMOD has quit IRC | 18:04 | |
| rgareus | drobilla`: `uname -m` as binary property would work. that's pretty standard and windows has something equivalent | 18:11 |
| *** gianMOD has joined #lv2 | 18:11 | |
| falktx | libmagic can tell the arch of a binary | 18:15 |
| * falktx wishes linux had universal binaries | 18:15 | |
| rgareus | well, it would be nice to have a universal universal LV2. linux, OSX, Windows, maybe *BSD, i386/x86_86,armhf | 18:16 |
| rgareus | except size. | 18:16 |
| rgareus | if you include 9 static libQTs :) | 18:16 |
| *** ensonic has joined #lv2 | 18:24 | |
| *** drobilla` is now known as drobilla | 19:03 | |
| *** oofus_lt has joined #lv2 | 19:18 | |
| *** ensonic has quit IRC | 19:39 | |
| *** rncbc has joined #lv2 | 19:44 | |
| *** ColaEuphoria has quit IRC | 21:05 | |
| *** oofus_lt_ has joined #lv2 | 21:59 | |
| *** oofus_lt has quit IRC | 22:03 | |
| *** oofus_lt_ has quit IRC | 22:08 | |
| *** oofus_lt_ has joined #lv2 | 22:15 | |
| *** oofus_lt_ has quit IRC | 22:16 | |
| *** oofus_lt has joined #lv2 | 22:17 | |
| *** Daveydave has joined #lv2 | 22:19 | |
| Daveydave | hellooo | 22:19 |
| *** wumpus has quit IRC | 22:19 | |
| *** wumpus has joined #lv2 | 22:21 | |
| *** Daveydave has quit IRC | 22:21 | |
| *** oofus_lt_ has joined #lv2 | 22:24 | |
| drobilla | rgareus: I suppose, but that's just arch, we need system, too | 22:26 |
| drobilla | Which the ol' system "triple" thing does, but it's always struck me as sketchy at best | 22:27 |
| *** oofus_lt has quit IRC | 22:27 | |
| drobilla | I don't know. Never had any incentive to really look into it that deeply. Probably not that hard, lilv can just do it automagically and nobody really needs to care | 22:27 |
| *** oofus_lt_ has quit IRC | 22:28 | |
| *** oofus_lt_ has joined #lv2 | 22:28 | |
| * drobilla doesn't distribute binaries, so easy candidate for the Not My Problem drawer :) | 22:29 | |
| dsheeler | If I need an array the size of the jack period, when would i allocate it? | 22:36 |
| *** oofus_lt has joined #lv2 | 22:38 | |
| *** oofus_lt_ has quit IRC | 22:42 | |
| *** oofus_lt has quit IRC | 22:47 | |
| falktx | activate | 23:06 |
| falktx | or instantiate | 23:06 |
| rgareus | dsheeler: back to the drawing board. the buffersize can change at any time | 23:13 |
| rgareus | dsheeler: there are extensions to (A) get informed about about changes and (B) tell a host about min/max. but generally that's best avoided | 23:14 |
| *** unclechu has quit IRC | 23:19 | |
| dsheeler | Ok. Wondered if there was stuff like you mention rgareus, but i'll avoid... so pre-allocating for various buffersizes is a way to go? | 23:23 |
| rgareus | dsheeler: assuming a max of 8192 will get you through anything | 23:25 |
| rgareus | with jack you can call jack_bufsize <num> at any time to change it. jack's max is 8192 | 23:26 |
| rgareus | Ardour does switch to 8K when exporting. but will split the processing into chunks whenever there's an automation event. | 23:26 |
| rgareus | so a plugin could see run(87) ; new control-data ; run(8105) | 23:27 |
| drobilla | Get the max at instantiation time, allocate that | 23:27 |
| dsheeler | rgareus: that's interesting so run might not be running on the buffersize, anyway | 23:31 |
| rgareus | dsheeler: correct | 23:31 |
| rgareus | dsheeler: a plugin can tell a host min = max = current. | 23:31 |
| rgareus | in which case ardour won't load the plugin :) | 23:31 |
| dsheeler | hehe | 23:32 |
| rgareus | jalv will, I think | 23:32 |
| rgareus | I've proposed a lv2 extension "noSampleAccurateCtrl" for that. | 23:33 |
| rgareus | nobody else had need for it yet, and it was pretty much ignored. | 23:33 |
| dsheeler | ok. thanks for the info folks | 23:36 |
| falktx | rgareus: hey. I found a need for it | 23:37 |
| falktx | rgareus: well, carla | 23:37 |
| * falktx feels like carla gets ignored too many times | 23:37 | |
| rgareus | falktx: does carla have automation? | 23:37 |
| falktx | midi cc | 23:39 |
| falktx | it splits the events for lv2 plugins | 23:39 |
| rgareus | aah, indeed | 23:40 |
| falktx | also, carla itself should have that property | 23:40 |
| falktx | makes no sense for carla process to be splitted right now as it has no params | 23:40 |
| rgareus | does carla expose any control ports ? | 23:41 |
| falktx | not atm | 23:41 |
| rgareus | so then there's no need for it ATM | 23:41 |
| falktx | ardour doesn't split the buffer during midi events? or tempo change events? | 23:41 |
| rgareus | midi-events: no. | 23:41 |
| rgareus | tempo-changes: not yet | 23:42 |
| rgareus | only control-port changes and only for automation playback. | 23:42 |
| falktx | and loop points I guess | 23:42 |
| rgareus | yes, I was just going to say that | 23:42 |
| rgareus | though that's a bit different. automation -> per individual plugin. Loop: all processing all plugins | 23:43 |
| rgareus | from the plugin's POV it's the same | 23:43 |
| rgareus | midi events are just pushed timestamped into the atom buffer. no need to split | 23:44 |
| rgareus | midi control (live) we assume that 1 cycle is more than accurate enough. | 23:44 |
| falktx | same for time | 23:44 |
| falktx | time events can be timestamped | 23:44 |
| falktx | ...that sounds weird to say | 23:45 |
| rgareus | heh, musical time events happen at a given wall-clock time :) | 23:45 |
| *** Spark[02] has joined #lv2 | 23:47 | |
| *** Spark[01] has quit IRC | 23:52 | |
| *** edogawa has quit IRC | 23:53 | |
Generated by irclog2html.py 2.13.0 by Marius Gedminas - find it at mg.pov.lt!