| *** abique|nephalem has joined #lv2 | 07:48 | |
| abique|nephalem | Hi, there is a mistake in lv2plug.in/gmpi.html, "25, Met - Events are simply 'sent' synchrnonously in a POD buffer like any other port." => synchronously | 07:49 |
|---|---|---|
| abique|nephalem | Also is the GMP 17 still not implemented? u-he Diva has a quality setting for realtime and an other one for offline rendering. | 07:56 |
| *** mlpug has joined #lv2 | 08:03 | |
| abique|nephalem | What is the difference between lilv and slv2? | 08:28 |
| abique|nephalem | coming back in a few minutes | 08:35 |
| *** abique|nephalem has quit IRC | 08:35 | |
| *** mlpug has quit IRC | 08:42 | |
| *** ddom has joined #lv2 | 08:57 | |
| *** abique|nephalem has joined #lv2 | 09:20 | |
| abique|nephalem | Hi again :) | 09:20 |
| *** abique|nephalem has quit IRC | 10:14 | |
| *** abique|nephalem has joined #lv2 | 10:27 | |
| *** abique|nephalem has quit IRC | 10:42 | |
| *** edogawa has joined #lv2 | 11:49 | |
| *** abique has joined #lv2 | 11:54 | |
| *** bgribble has joined #lv2 | 13:30 | |
| *** unclechu has joined #lv2 | 14:51 | |
| *** bgribble has quit IRC | 15:56 | |
| *** ddom has quit IRC | 15:58 | |
| *** zth has joined #lv2 | 17:03 | |
| *** mlpug has joined #lv2 | 17:24 | |
| *** falktx has joined #lv2 | 19:13 | |
| drobilla | Hrm... sord semantics break, but not an actual ABI break... to bump major version or not... :/ | 19:15 |
| drobilla | abique: fixed, thanks | 19:16 |
| drobilla | abique: It would be trivial to do (probably with an option), but AFAIK nobody actually is and we don't have an appropriate property for it | 19:17 |
| drobilla | abique: slv2 is extremely ancient. Don't use it. | 19:17 |
| falktx | drobilla: do you get notifications about lv2plug.in bugs? | 19:26 |
| *** NickSB has quit IRC | 19:33 | |
| *** zth has quit IRC | 19:36 | |
| *** mlpug has quit IRC | 19:47 | |
| *** falktx has quit IRC | 20:04 | |
| *** falktx has joined #lv2 | 20:08 | |
| drobilla | falktx: yes. | 20:09 |
| falktx | ok, just wondering. | 20:12 |
| *** unclechu has quit IRC | 20:15 | |
| drobilla | Also notification about my own comments, which is a bit of an annoying misfeature in Trac | 20:22 |
| *** abique|home has joined #lv2 | 20:36 | |
| abique|home | Hi, is there a program which can generate a .ttl out of a dyn-manifest plugin? Thanks! | 20:45 |
| *** falktx_ has joined #lv2 | 20:46 | |
| *** falktx has quit IRC | 20:47 | |
| drobilla | abique|home: lv2info -p -m | 21:03 |
| abique|home | drobilla, thanks! | 21:04 |
| *** edogawa has quit IRC | 21:10 | |
| *** edogawa has joined #lv2 | 21:13 | |
| falktx_ | rgareus: drobilla: is this http://dev.drobilla.net/changeset/307/sord ready for usage? should I update my local sord build? | 21:44 |
| rgareus | falktx_: AFAICT, it's ready for usage, but not quite ready for distribution (liblilv update in the queue, you may want to wait for that) | 21:46 |
| falktx_ | I don't plan to update the reposity lilv, but the local build used in carla | 21:47 |
| rgareus | falktx_: changeset 308 FTW. | 21:48 |
| rgareus | 307 had some issues on 32bit systems | 21:48 |
| falktx_ | not seeing it here http://dev.drobilla.net/changeset/308/sord | 21:49 |
| * falktx_ wants to grab a raw diff | 21:49 | |
| *** edogawa has quit IRC | 21:49 | |
| rgareus | falktx_: no idea. I just did a git svn rebase | 21:49 |
| falktx_ | trac should update automatically | 21:50 |
| falktx_ | rgareus: is this minutes fresh perhaps? | 21:50 |
| rgareus | svn is at 309 http://svn.drobilla.net/sord/ | 21:50 |
| rgareus | though trac indeed only list up to 307 at http://dev.drobilla.net/browser :( | 21:51 |
| rgareus | drobilla: ^^ | 21:51 |
| rgareus | falktx_: no, it's at least a day old | 21:52 |
| rgareus | rev 308 is from Sep 23 23:03:12 2014 +0000 | 21:53 |
| falktx_ | hmm | 21:56 |
| falktx_ | python3: sord-0.12.2/src/zix/btree.c:626: zix_btree_get: Assertion `frame->index < frame->node->n_vals' failed. | 21:56 |
| falktx_ | happens with amsynth | 21:57 |
| falktx_ | rgareus: can you confirm? (amsynth-lv2) | 21:57 |
| rgareus | falktx_: hang on a sec. | 21:58 |
| rgareus | falktx_: I did try various LV2s but not amsynth | 21:59 |
| falktx_ | fabla does it too | 21:59 |
| rgareus | falktx_: does this happen on scan or when you instantiate the Lv2? | 21:59 |
| falktx_ | when loading the plugin | 21:59 |
| falktx_ | sorcer is ok, fabla asserts | 22:00 |
| falktx_ | rgareus: this assumes a debug build of course, otherwise assert() does nothing :) | 22:00 |
| rgareus | falktx_: I do have a debug-build of the complete lad stack (incl. sord) | 22:01 |
| falktx_ | tested a bunch of others, amsynth and fabla for now cause the assertion | 22:02 |
| rgareus | indeed ardour-3.5.3041: ../sord/src/zix/btree.c:626: zix_btree_get: Assertion `frame->index < frame->node->n_vals' failed. | 22:02 |
| falktx_ | both on the same line | 22:02 |
| rgareus | with amsynth-lv2 | 22:02 |
| falktx_ | try fabla | 22:02 |
| *** NickSB has joined #lv2 | 22:03 | |
| rgareus | falktx_: fabla works | 22:03 |
| rgareus | might be an issue as to where in the tree the information is stored. which is different depending on the # of plugins installed | 22:04 |
| falktx_ | later, cya | 22:05 |
| falktx_ | and thanks | 22:05 |
| *** falktx_ has quit IRC | 22:05 | |
| abique|home | Hey, is there a recommanded lv2 host? | 22:21 |
| abique|home | to test in development plugins | 22:22 |
| rgareus | abique|home: no | 22:22 |
| rgareus | abique|home: jalv is pretty good for testing | 22:22 |
| abique|home | yeah | 22:22 |
| abique|home | thanks :D | 22:22 |
| rgareus | abique|home: since most (if not all hosts) use liblilv to do all the heavy lifting, one is as good as the other | 22:22 |
| abique|home | ok | 22:23 |
| rgareus | abique|home: the main difference between hosts is GUI handling. (due to toolkit mixing) | 22:23 |
| rgareus | e.g a QT5 GUI in a QT4 host won't fly. | 22:23 |
| rgareus | same with gtk3 and gtk2 | 22:23 |
| abique|home | I only use xlib and cairo | 22:23 |
| rgareus | abique|home: you'll be fine then. use jalv.gtk <your uri> | 22:24 |
| abique|home | and I have to put my plugin in ~/.lv2/ right? | 22:24 |
| abique|home | lv2info -p -m | 22:31 |
| abique|home | sorry, lv2info -p -m segfaults on archlinux | 22:31 |
| abique|home | I have no ~/.lv2 and only lv2 plugins avalaible from the repository | 22:32 |
| rgareus | ~/.lv2/<plugin-dir>/ is fine | 22:49 |
| rgareus | abique|home: try lv2ls first | 22:49 |
| abique|home | Yes I can see my plugin now | 22:49 |
| abique|home | by the way do you have an exemple of a dyn-manifest plugin? | 22:50 |
| abique|home | I find this turtle thing a bit comlicated | 22:50 |
| rgareus | as for the segfault, maybe that's because of wrong args: lv2info -P some_file_name -m some_other_filename <URI> | 22:50 |
| * rgareus has no idea what dyn-manifest is. | 22:51 | |
| * rgareus likes plain RDF | 22:51 | |
| abique|home | rgareus, lv2plug.in/ns/ext/dynmanifest/ | 22:51 |
| abique|home | I'm trying to port u-he synthesizers to lv2 | 22:52 |
| abique|home | so they have lots of parameters | 22:52 |
| abique|home | and plugins | 22:52 |
| abique|home | I have all the metadata in the code | 22:53 |
| abique|home | and I can't write the rdf by hand | 22:53 |
| rgareus | abique|home: maybe a simple script can help | 22:54 |
| abique|home | that could be the lv2info's job ;-) | 22:54 |
| rgareus | abique|home: I don't think so, it's a chicken & egg. lv2info parses the RDF in the first place | 22:55 |
| rgareus | lv2info does not instantiate the plugin | 22:55 |
| rgareus | it just reads RDF | 22:55 |
| abique|home | I think I found my way through dyn manifest :) | 22:58 |
| *** ricardocrudo has quit IRC | 23:02 | |
| drobilla | abique|home: dyn manifest just parses ttl in code instead of from a file. The plugin still has to generate ttl | 23:08 |
| drobilla | I don't particularly like this way of doing dynamic data, particularly since we have plugins, but in any case, dyn-manifest isn't a way around ttl | 23:08 |
| drobilla | er, we have atoms* | 23:08 |
| abique|home | drobilla, ok so what do you recommand? | 23:09 |
| drobilla | rgareus: Current sord and lilv in svn have issues. Unfortunately it's going to be a hard dependency, version change, upgrade path. | 23:09 |
| drobilla | rgareus: I have it all fixed but haven't committed yet | 23:09 |
| drobilla | abique|home: What format are you porting from? | 23:09 |
| abique|home | drobilla, I'm porting u-he plugins | 23:10 |
| drobilla | Whatever those are. | 23:10 |
| abique|home | they have an internal plugin descriptor format, with parameters | 23:10 |
| drobilla | Well, if it's something supported by naspro, I do that. Wrap, use lv2info to dump data, clean up by hand. | 23:11 |
| drobilla | If not, you'll have to write the data somehow yourself. | 23:11 |
| abique|home | I can't write it myself | 23:11 |
| abique|home | neither clean it by hand | 23:12 |
| drobilla | er, okay? | 23:12 |
| abique|home | I need to generate it for sure | 23:12 |
| drobilla | That would fall under "somehow". | 23:12 |
| abique|home | so dynmanifest is not my solution then? | 23:13 |
| drobilla | No. | 23:13 |
| drobilla | You'd need to solve the exact same problem (generating ttl) anyway. | 23:13 |
| abique|home | that's fine to generate it in dyn-manifest | 23:14 |
| drobilla | ... why is that fine but writing it to a file is not? | 23:14 |
| rgareus | drobilla: did you see the ../sord/src/zix/btree.c:626: zix_btree_get: Assertion `frame->index < frame->node->n_vals' failed. issue? | 23:14 |
| drobilla | rgareus: As I have mentioned to you about 3 times in two different channels, yes, I know. It does not work. ;) | 23:15 |
| drobilla | rgareus: New sord is not a compatible change. | 23:15 |
| drobilla | rgareus: Old lilv inserts while iterating. Can't do that with a B-tree. | 23:15 |
| abique|home | drobilla, what I do not want is to write one more program which has to load my plugin and generate the file. If I can do it through dyn-manifest, then use lv2info to get the manifest it is fine | 23:15 |
| rgareus | drobilla: aah. and current svn lilv is still 'too old' | 23:15 |
| drobilla | abique|home: Fair enough, I guess. Seems like a whole lot more effort than skipping the middle man, to me. | 23:16 |
| abique|home | I do not understand your sentence | 23:16 |
| drobilla | rgareus: Yes. As is current svn sord. I didn't foresee this, only though of removing | 23:17 |
| drobilla | abique|home: Implementing dyn-manifest is more work than not implementing dyn-manifest | 23:17 |
| drobilla | abique|home: But implement dyn-manifest if you want. I just don't see the point if you're essentially just going to throw it away once you get the ttl anyway | 23:18 |
| drobilla | rgareus: Will bump the version and go through everything and whatnot tonight. Just got home. | 23:18 |
| drobilla | rgareus: I was kind of hoping to keep ABI stable until more significant changes, but oh well. | 23:19 |
| abique|home | so you suggest to build my own entry point, a simple program which loads the .so and then call the ttl generation? | 23:19 |
| drobilla | abique|home: Can't you just load the plugin in code? Are these proprietary or something? | 23:20 |
| abique|home | what do you mean by load? Yes they're proprietary. | 23:21 |
| drobilla | Oh. You might want to have mentioned that... | 23:21 |
| drobilla | They're proprietary but don't implement any standard plugin API? | 23:21 |
| * drobilla doesn't get how this is something that can be useful | 23:22 | |
| abique|home | they implement vst, aax, au, ... | 23:22 |
| drobilla | FFS | 23:22 |
| drobilla | naspro wraps VST. | 23:22 |
| drobilla | I've never used that bit and only done this for LADSPA but it shouldn't make any difference. | 23:23 |
| drobilla | There will be at least a little bit of fixing up to do, though, since the plugin URIs much change at the very least | 23:24 |
| abique|home | I downloaded the version 0.5.1 from http://naspro.sourceforge.net/plugins.html but grep -i vst is empty | 23:26 |
| abique|home | Yet there is enough in the tarball | 23:27 |
| drobilla | I dunno. Site explicitly mentions VST. | 23:28 |
| drobilla | Maybe it "supports" VST in some sense but there is no VST=>LV2 wrapper. | 23:29 |
| drobilla | Since you're working with proprietary VST plugins you essentially would be implementing a VST=>LV2 wrapper by implementing dyn-manifest for these. | 23:29 |
| drobilla | Quite literally, since you have no power to do anything 'custom' | 23:32 |
| drobilla | i.e. unless I'm missing something, this is not a u-he specific task | 23:32 |
| drobilla | abique|home: You might find "lvz" in the mda.lv2 source useful. It essentially is already a (slightly half-assed) VST=>LV2 wrapper | 23:34 |
| drobilla | sourced-based, though. no dll loading | 23:37 |
| drobilla | I think that's the class you instantiate from a VST binary, though | 23:38 |
| drobilla | (But I know very little about VST) | 23:38 |
| abique|home | thanks drobilla, I'll have a look later. I need to go to bed. | 23:42 |
| abique|home | good night | 23:42 |
| *** abique|home has quit IRC | 23:42 | |
| drobilla | rgareus: Committed w/o version bump for now. Issue should be fixed | 23:46 |
| rgareus | drobilla: compiling... | 23:49 |
Generated by irclog2html.py 2.13.0 by Marius Gedminas - find it at mg.pov.lt!