| drobilla | falktx: "host should use be" ... | 00:00 |
|---|---|---|
| falktx | err, meant using | 00:00 |
| drobilla | falktx: I get this message using the show interface | 00:00 |
| falktx | be using | 00:00 |
| drobilla | falktx: It seems to create a window, but I don't actually see it | 00:00 |
| drobilla | maybe broken size or something | 00:00 |
| falktx | did you call show->show() ? | 00:00 |
| drobilla | yeah, tiling broken. if I float it, I get a window that's the right size anyway | 00:01 |
| drobilla | let me implement idle, then it should hopefully work | 00:01 |
| falktx | I don't think the window can be shown if idle is not called | 00:02 |
| drobilla | Yeah, that's it. | 00:06 |
| drobilla | Now to figure out how to make this horrid mess actually feasible code... | 00:06 |
| * falktx takes the chance to implement a custom carla option | 00:07 | |
| drobilla | (jack session is making this annoying) | 00:10 |
| drobilla | Maybe I'll implement the most half-assed command line interface possible | 00:13 |
| drobilla | Ultimately it would be nice to have that for all versions of jalv, but this will do for now | 00:14 |
| falktx | don't rush it. there's time after lac | 00:14 |
| drobilla | Hard to do with idle anyway, so nevermind | 00:16 |
| drobilla | (no threads) | 00:16 |
| drobilla | falktx: http://dev.drobilla.net/changeset/5377 | 00:44 |
| drobilla | falktx: jalv.console -s | 00:44 |
| drobilla | I need to add a timedwait to my semaphore wrapper, but can't test on other platforms right now, so usleep kludge... | 00:45 |
| falktx | nice | 00:49 |
| falktx | drobilla: will any example plugin have this? | 00:50 |
| drobilla | falktx: sampler | 00:50 |
| drobilla | Haven't committed, it doesn't do anything on window close yet | 00:50 |
| drobilla | Think I might just not care | 00:50 |
| falktx | so I guess it inits gtk if needed, correct? | 00:50 |
| drobilla | falktx: yeah. it just calls it, you can call gtk_init many times | 00:51 |
| drobilla | then idle calls gtk_main_iteration | 00:51 |
| drobilla | not sure what the latter will do in an app that already has glib running | 00:51 |
| drobilla | Also think I might just not care :) | 00:51 |
| falktx | there should be a way to detect if gtk is already setup | 00:51 |
| falktx | and then only call iteration if not | 00:52 |
| falktx | but this would break 2nd instances, so better not do it | 00:52 |
| drobilla | As far as I can tell there isn't | 00:54 |
| falktx | ah, we should only call gtk_main_iteration if ui widget was null | 00:55 |
| drobilla | I could make it only return idle if it's actually been shown via show/hide, but I don' tknow how things will tolerate it returning null for extension_data that should be there | 00:55 |
| drobilla | You can call gtk_main_iteration recursively anyway, it shouldn't actually cause any problems | 00:55 |
| falktx | I need to try and put this in calf | 00:56 |
| falktx | and want that analyzer in carla :) | 00:56 |
| falktx | drobilla: can we have a option for ui title? | 00:56 |
| drobilla | I guess. | 00:57 |
| falktx | LV2_OPTIONS_INSTANCE, type atom-string, uri ui:title can work | 00:57 |
| falktx | drobilla: I'm using a custom option to set the parent Id (not embed, but for modal dialog) | 00:58 |
| falktx | not sure if you think that's useful | 00:58 |
| drobilla | falktx: What kind of ID? | 00:59 |
| falktx | x11 win Id | 00:59 |
| falktx | it makes plugin UIs always show on top of the host | 00:59 |
| drobilla | falktx: Why not just use ui:parent? | 00:59 |
| falktx | cause I can't embed it | 01:00 |
| falktx | ie, show/hide stuff | 01:00 |
| drobilla | so? | 01:00 |
| drobilla | oh, it's a feature. n/m | 01:00 |
| falktx | I don't need embed, don't use it in carla | 01:00 |
| falktx | I'm fine with it being a custom internal thing in carla. it would get messy with x11 vs other Ids anyway | 01:01 |
| drobilla | If you expect every UI in existence to implement this interface, you'll be waiting forever, but that's your call | 01:01 |
| drobilla | A generic wrapper/lib that will popup an <anything> UI is (still) the best solution | 01:02 |
| falktx | I'm hoping to patch at least calf and ir.lv2 for this | 01:02 |
| falktx | those are the problematic ones cause they need instance-data | 01:02 |
| drobilla | 99999999999999999999999999999999999999 implementations of pop-up-a-window-and-call-main-interface | 01:03 |
| drobilla | 50% of them buggy no doubt | 01:03 |
| drobilla | A stupid idea in a sane interface is at least an improvement, though | 01:04 |
| drobilla | Maybe some day I will add a just_show_a_window_somehow() to suil that works in every case, but not now | 01:06 |
| falktx | drobilla: if/when you make eg-sampler work with this, let me know | 01:06 |
| drobilla | falktx: r914 | 01:08 |
| drobilla | falktx: I don't get it though, what UIs does carla currently support before this? external-ui only? | 01:09 |
| falktx | the gtk2 ones were only using bridges | 01:09 |
| falktx | so no ir.lv2 and regular calf | 01:09 |
| drobilla | "bridges" being? | 01:09 |
| falktx | an external gtk2 app | 01:10 |
| drobilla | I don't get why you want to patch hundreds of plugins with the same thing instead of just doing that thing in once place | 01:11 |
| falktx | what thing? | 01:12 |
| drobilla | pop up a (gtk or whatever) window and drive the loop | 01:12 |
| falktx | carla backend uses no toolkits right now | 01:13 |
| falktx | and I'd like to keep it that way | 01:13 |
| drobilla | Well, it's loading UIs that use toolkits... | 01:13 |
| falktx | UIs will be bridged if possible, in a separate process | 01:14 |
| drobilla | Same idea basically, just lib instead of process, then you only have to do it once and it will work for any gtk/whatever UI | 01:15 |
| drobilla | I guess if it's only for instance-access tings it's not so terrible... | 01:16 |
| drobilla | Doing the same thing in countless places instead of one is about the most objective "terrible software engineering" smell there is, is all | 01:16 |
| drobilla | It's the kind of thing that makes me have to deal with an Internet full of "LV2 is shiiiiiiiiiiiiiit" | 01:17 |
| drobilla | I dislike this :)_ | 01:17 |
| drobilla | But whatever, I don't have the energy to care enough | 01:19 |
| drobilla | Every plugin ever having to implementing this fallback is very definitely the wrong solution though | 01:19 |
| falktx | hopefully new plugins will not use instance-access | 01:20 |
| falktx | or 99% of them | 01:20 |
| drobilla | True | 01:21 |
| *** timbyr has quit IRC | 02:12 | |
| *** timbyr has joined #lv2 | 02:14 | |
| drobilla | Alright, well, that's it for me today. Maybe I can do some packaging tomorrow | 02:19 |
| drobilla | Kinda hasty, but I don't thiiiiiiink this stuff can cause too many problems | 02:19 |
| drobilla | </famousLastWords> | 02:19 |
| falktx | beaten twice by deprecated stuff in calf... :( | 03:23 |
| falktx | first, they need ui:makeResident or else I get weird crashes on dlclose | 03:24 |
| falktx | then the detection of that was failing | 03:24 |
| falktx | it still had ui:requiredFeature!! arrr sigh | 03:25 |
| falktx | it's supposed to be lv2:requiredFeature damn it | 03:25 |
| falktx | patch here: http://kxstudio.sourceforge.net/Paste/repo/2uWnY | 03:44 |
| *** falktx has quit IRC | 04:50 | |
| *** edogawa has joined #lv2 | 06:45 | |
| *** ricardocrudo has joined #lv2 | 12:21 | |
| *** ricardocrudo has quit IRC | 12:56 | |
| *** ricardocrudo has joined #lv2 | 13:33 | |
| *** ricardocrudo has quit IRC | 13:41 | |
| *** bgribble has joined #lv2 | 14:43 | |
| *** bgribble has quit IRC | 15:05 | |
| drobilla | last I tried calf was accepting patches, but I'm not up to date | 16:31 |
| drobilla | Also still uses event. I need to think about how to deal with that, since something like Calf isn't about to change URIs, but it's teeeeeeeeeechnically not okay to completely change a port type on the same plugin.... | 16:31 |
| *** bgribble has joined #lv2 | 17:15 | |
| *** rncbc has joined #lv2 | 17:24 | |
| *** mlpug has joined #lv2 | 18:08 | |
| *** edogawa_ has joined #lv2 | 18:12 | |
| *** edogawa has quit IRC | 18:15 | |
| *** falktx has joined #lv2 | 18:43 | |
| Anchakor | apparently some recent change in lv2 svn broke tal noisemaker lv2: http://codepad.org/5Ot9mvQS | 19:26 |
| drobilla | Anchakor: Unlikely. Sounds like botched builds to me | 19:31 |
| drobilla | "Package manager? What's that?" | 19:31 |
| drobilla | If I had a nickel for every time one of those sorts of people just fucked up their own system and tried to blame some code change on it ;) | 19:31 |
| *** ricardocrudo has joined #lv2 | 19:32 | |
| drobilla | If it goes away when lv2 is reverted, then maybe, but if nothing was rebuilt in the first place... | 19:32 |
| Anchakor | I told him to rebuild noisemaker | 19:33 |
| Anchakor | but considering lv2 is pretty ABI stable, that shouldn't be necessary I think | 19:35 |
| drobilla | I need to review the diff, but the only really significant thing that's changed is atom stuff | 19:36 |
| drobilla | and even there not in an ABI breakey way | 19:36 |
| drobilla | But if something is broken about LV2 I certainly want to know ASAP | 19:36 |
| *** mlpug has quit IRC | 19:38 | |
| *** rncbc is now known as rncbc|AFK | 20:04 | |
| *** bgribble has quit IRC | 20:34 | |
| *** ricardocrudo has quit IRC | 21:22 | |
| *** edogawa_ has quit IRC | 22:03 | |
| *** rncbc|AFK is now known as rncbc | 22:49 | |
| *** rncbc has quit IRC | 22:53 | |
Generated by irclog2html.py 2.13.0 by Marius Gedminas - find it at mg.pov.lt!