| *** ahellquist-mobil has joined #lv2 | 00:02 | |
| *** AndChat|48116 has quit IRC | 00:02 | |
| *** ahellquist-mobil has quit IRC | 00:08 | |
| *** AndChat|48116 has joined #lv2 | 00:08 | |
| *** bgribble has joined #lv2 | 00:47 | |
| *** HarryHaaren has quit IRC | 01:05 | |
| *** ddom has quit IRC | 01:11 | |
| *** falktx has quit IRC | 01:23 | |
| drobilla | Hmmmmmmm. Well, releases out, feel moderately inclined to do Pugl things, but don't have the ability to keep the other platforms in sync | 01:34 |
|---|---|---|
| rgareus | drobilla: I recently hoped to be able to /steal/ from pugl for the xjadeo openGL windows and mac versions. | 02:02 |
| rgareus | drobilla: but all I eventually got out of it was CreateWindowEx() :) | 02:02 |
| rgareus | writing a standalone app is quite different. | 02:03 |
| drobilla | rgareus: How so? | 02:05 |
| rgareus | mainly the event-handling (except on OSX, the GUI is not in the main thread, because the GUI thread can block eg. window-resizes are modal on windows. | 02:08 |
| drobilla | There's two main things I want to do with pugl: 1) Cairo ability 2) Event struct, and move glut compatibility to a header that's driven by the one event callback | 02:09 |
| drobilla | Where the latter is basically jumping off the "yeah, whatever, screw this half glut compatibility thing" but it certainly seems infinitely easier to implement widgets and such that way | 02:09 |
| drobilla | cliff* buf | 02:09 |
| drobilla | but* | 02:09 |
| drobilla | (good grief) | 02:09 |
| rgareus | drobilla: I like (1) a lot | 02:11 |
| rgareus | drobilla: but really it should be part of pugl itself. | 02:11 |
| rgareus | drobilla: it's just a call to cairo_create_surface() or _create_similar_surface(...,GL buffer); a | 02:12 |
| rgareus | in my experience, cairo software surfaces + blit are faster on 99% of all machines compared to cairo's GL backend | 02:13 |
| rgareus | Fons made the same experience when he wrote his flight-simulator. | 02:13 |
| rgareus | drobilla: were you there when he showed that off at LAC? | 02:14 |
| drobilla | rgareus: Yeah, I just mean separate in terms of implementation. pugl/glut_dispatch.h or whatever | 02:15 |
| drobilla | Yeah, just need to rework the API so the init doesn't set up a bunch of GL stuff | 02:15 |
| rgareus | ok | 02:15 |
| drobilla | The window set up will be slightly different, so some kind of separation is needed there | 02:15 |
| drobilla | and a pugl_get_context() that won't actually do anything for GL | 02:16 |
| drobilla | rgareus: Nope, missed that | 02:22 |
| drobilla | I think it was during the Faust talk or something | 02:23 |
| rgareus | drobilla: I like having separate callbacks for different events - but that's easy enough to select() in the plugin. | 02:23 |
| rgareus | switch(), even | 02:23 |
| drobilla | rgareus: Yeah, given one, the other is easy. I'll just rework the existing callback API to be driven by an event struct instead | 02:23 |
| drobilla | 'course this means I need to invent yet-another almost-like-X11 event struct which is annoying | 02:24 |
| drobilla | but also more extensible | 02:24 |
| drobilla | The total lack of extensibility of these callbacks makes me nervous, even if they're from something established, we've already had to embrace and extend them... | 02:24 |
| rgareus | drobilla: I think this is less of an issue with pugl -- unless you want to aim for 'system wide shared pugl lib' on Linux. | 02:26 |
| rgareus | drobilla: for osx + win, plugins are going to statically link against pugl anyway. | 02:26 |
| rgareus | and hence most plugins won't special case for system-wide libpugl on linux. | 02:27 |
| rgareus | so you can change the API and ABI and extend what you want, anytime. | 02:28 |
| drobilla | ooooooooooorrrrrr maybe I'll just build my ErgoDox instead :) | 02:28 |
| drobilla | rgareus: I do plan on releasing it as a proper library | 02:28 |
| drobilla | rgareus: If people want to bundle it, fine, but this copy-and-fork nonsense was a bad idea | 02:28 |
| drobilla | There are reasons why there is The Way we do these things, and they're good ones... | 02:29 |
| rgareus | drobilla: yes. keeping it API compatible will certainly be nice. | 02:29 |
| rgareus | drobilla: but still, in the context of x-platform portable plugins using it as shlib makes little sense. | 02:30 |
| drobilla | No shortage of dev community around here who would much prefer a package just being there on their distro and wouldn't do all the binary static nightmare for other platforms anyway | 02:30 |
| drobilla | rgareus: yep | 02:30 |
| rgareus | sure `pkg-config --libs --static libpugl` | 02:31 |
| drobilla | Except not static because plugins and all :) | 02:31 |
| rgareus | --static with pkg-config only adds dependent libs | 02:31 |
| drobilla | Really it's just more about starting to release it as a proper project | 02:31 |
| rgareus | it would return libpugl.a -lgl | 02:31 |
| rgareus | instead of -lpugl | 02:32 |
| drobilla | Current use cases will probably stay the same | 02:32 |
| rgareus | well, plugins can be statically linked, too. | 02:32 |
| rgareus | pretty much all of falkTX's are | 02:32 |
| rgareus | ldd returns libc.so.6 and libdl.so.2 that's mosly it. All of QT, cairo, and whatever else is inside. | 02:34 |
| rgareus | s/mosly/mostly/ | 02:34 |
| * drobilla wonders what situation the non-portable warnings gcc gives you about that refer to | 02:40 | |
| drobilla | I guess statically linked, but PIC | 02:41 |
| * drobilla looks at the clock | 02:43 | |
| drobilla | Who am I kidding, I'm not doing shit tonight | 02:43 |
| rgareus | 4:43 am. | 02:43 |
| drobilla | Only 22:48 here, but I'm procrastinating and I know how that goes :) | 02:48 |
| * drobilla wonders how Ingen's doing... | 02:49 | |
| rgareus | jalv, ganv, lilv, suil and sratom. wow. quite a release day | 02:52 |
| drobilla | Yeah. still patchage, fomp.lv2, mda.lv2 but I don't know if I'll bother | 02:52 |
| rgareus | drobilla: should jalv exit when jackd terminates? | 02:53 |
| drobilla | rgareus: Um. | 02:53 |
| drobilla | mu. | 02:53 |
| drobilla | It should (ideally) either exit or have a reconnect option | 02:53 |
| rgareus | drobilla: here it did not. but I rarely stop jackd to I did not notice that before. | 02:54 |
| drobilla | rgareus: If you mean "should" as in I currently expect it to, no idea. So, I doubt it. | 02:54 |
| drobilla | Unless exit-via-crash counts :D | 02:55 |
| rgareus | no jack_on_shutdown() handler in jalv | 02:55 |
| drobilla | Jalv is, by definition, perfect | 02:56 |
| drobilla | Until sufficient time has passed since release day :) | 02:56 |
| rgareus | drobilla: jalv just kept going as if nothing happened. | 02:56 |
| rgareus | drobilla: I'll open an issue on your bug-tracker. | 02:57 |
| drobilla | Random note: I was really entertained to hear all the Europeans say "yalv" | 02:57 |
| drobilla | I had never even considered it | 02:57 |
| drobilla | rgareus: Please do. It's getting empty | 02:57 |
| rgareus | jalv "a" as in pot*aaa*to or tom*aaa*to :) | 02:57 |
| drobilla | Most sure-fire way to make drobilla do something eventually, odd how little exploited that is, really :) | 02:58 |
| rgareus | drobilla: I may attach a patch to it as well. so maybe a few days time. | 02:58 |
| rgareus | in that case. | 02:58 |
| rgareus | drobilla: how does that correlate to bug priority? | 03:00 |
| * rgareus chose 'minor' 'feature request' at http://dev.drobilla.net/ticket/978#ticket | 03:00 | |
| drobilla | rgareus: Doesn't really matter. That seems reasonable. | 03:01 |
| rgareus | though I could certainly argue that it's a cruitical bug to ignore jack_on_shutdown(); if that helps :) | 03:02 |
| rgareus | drobilla: sorry for spoiling your release-day. | 03:03 |
| drobilla | meh | 03:04 |
| drobilla | Not enough of an issue to bother me | 03:04 |
| drobilla | Lots of improvements in Jalv this time around anyway | 03:04 |
| drobilla | Shit, forgot about falktx's randomize thing | 03:46 |
| *** bgribble has quit IRC | 04:26 | |
| *** zth has joined #lv2 | 07:09 | |
| *** zth has quit IRC | 09:22 | |
| *** edogawa has joined #lv2 | 09:26 | |
| *** falktx has joined #lv2 | 10:22 | |
| *** ahellquist-mobil has joined #lv2 | 12:46 | |
| *** AndChat|48116 has quit IRC | 12:49 | |
| *** AndChat|48116 has joined #lv2 | 13:08 | |
| *** ahellquist-mobil has quit IRC | 13:11 | |
| *** edogawa_ has joined #lv2 | 13:34 | |
| *** edogawa has quit IRC | 13:37 | |
| *** zth has joined #lv2 | 13:52 | |
| *** HarryHaaren has joined #lv2 | 14:01 | |
| *** edogawa_ is now known as edogawa | 14:24 | |
| *** falktx has quit IRC | 14:28 | |
| *** AndChat|48116 has quit IRC | 14:30 | |
| *** ahellquist-mobil has joined #lv2 | 14:30 | |
| *** AndChat|48116 has joined #lv2 | 14:37 | |
| *** ahellquist-mobil has quit IRC | 14:39 | |
| *** falktx has joined #lv2 | 15:26 | |
| *** ricardocrudo has joined #lv2 | 15:57 | |
| *** ddom has joined #lv2 | 15:57 | |
| *** HarryHaaren has quit IRC | 16:11 | |
| *** ricardocrudo has quit IRC | 16:14 | |
| *** falktx has quit IRC | 16:18 | |
| *** falktx has joined #lv2 | 16:35 | |
| *** edogawa has quit IRC | 19:06 | |
| *** edogawa has joined #lv2 | 19:07 | |
| *** ricardocrudo has joined #lv2 | 19:40 | |
| *** ahellquist-mobil has joined #lv2 | 20:21 | |
| *** falktx_ has joined #lv2 | 20:21 | |
| *** AndChat|48116 has quit IRC | 20:22 | |
| *** falktx has quit IRC | 20:25 | |
| *** ricardocrudo has quit IRC | 20:30 | |
| *** zth has quit IRC | 20:32 | |
| *** unclechu has joined #lv2 | 20:49 | |
| *** AndChat|48116 has joined #lv2 | 21:20 | |
| *** falktx_ has quit IRC | 21:21 | |
| *** ahellquist-mobil has quit IRC | 21:24 | |
| *** falktx has joined #lv2 | 21:43 | |
| *** edogawa has quit IRC | 22:59 | |
| * drobilla flounders WRT what to do with Ingen | 23:03 | |
| drobilla | The task of sitting down and actually investing a bunch of user time to test and figure things out is just too daunting | 23:03 |
| drobilla | Maybe I should just kick out a "prelease" however the hell it is now and hope other people do that | 23:03 |
| drobilla | I need well-defined problems | 23:04 |
| Anchakor | maybe you can ask people who are already using it | 23:31 |
| Anchakor | (not me) | 23:31 |
| drobilla | i.e. almost nobody. That's the problem :) | 23:36 |
| falktx | I would use it if it was a bit more stable... | 23:36 |
| falktx | seems like things change constantly | 23:36 |
| Anchakor | well there are few people, like the guy who made the videos like 0.5-1 year ago | 23:36 |
| drobilla | Aside from this week things haven't changed for months | 23:37 |
| drobilla | But that's the other problem | 23:37 |
| drobilla | Apparently it's not stable | 23:37 |
| drobilla | News to me. | 23:37 |
| falktx | well, I never got a patch to save&load 100% correctly | 23:37 |
| Anchakor | well release 1.0 if it's stable ;) | 23:37 |
| drobilla | Anchakor: It's compatibility changing things that worry me the most. Pretty sure the patch file format is more than good enough (being essentially lv2++), but the stuff in the graph, not so much | 23:38 |
| drobilla | Then again replacing internals with newer versions over time isn't the end of the world | 23:38 |
| Anchakor | as long they sound and work the same as the old ones | 23:39 |
| drobilla | If they work the same then they wouldn't need replacing, now would they? :P | 23:41 |
| drobilla | (Ingen has no sound producing internals. There's like 4 total) | 23:41 |
| Anchakor | well they might have new additional options with backwards compatible defaults :) | 23:41 |
| * drobilla gets an illegal instruction on a fresh compile | 23:43 | |
| drobilla | um..... okay then | 23:43 |
| falktx | stable! :) | 23:44 |
| * falktx runs | 23:44 | |
| drobilla | It was Jack. | 23:44 |
| falktx | set_property ? | 23:45 |
| falktx | I get random crashes on set_property when a previous instance was force-killed | 23:45 |
| falktx | can't reproduce it, seems really random | 23:45 |
| drobilla | yep. Went away when I restarted jackd | 23:47 |
| * drobilla blames bdb, which the metadata implementation shouldn't be based on in the first place | 23:47 | |
| Anchakor | I think most projects are backing from BDB now Oracle changed the license | 23:48 |
| drobilla | Doesn't need a DB anyway, but the responsibility of someone to save that stuff is just a big unanswered question | 23:49 |
| drobilla | Clearly (IMO) the job of a session manager, but then there's that can of worms... | 23:49 |
Generated by irclog2html.py 2.13.0 by Marius Gedminas - find it at mg.pov.lt!