| *** gianMOD has joined #lv2 | 00:17 | |
| *** gianMOD has quit IRC | 00:44 | |
| *** ricardocrudo has quit IRC | 00:46 | |
| *** artfwo has quit IRC | 02:30 | |
| *** fundamental has left #lv2 | 04:10 | |
| *** falktx has joined #lv2 | 04:11 | |
| *** falktx` has quit IRC | 04:16 | |
| *** sigma6 has joined #lv2 | 07:53 | |
| *** edogawa has joined #lv2 | 08:04 | |
| *** ColaEuphoria has joined #lv2 | 08:19 | |
| ColaEuphoria | http://i.imgur.com/yUmibTb.png | 08:20 |
|---|---|---|
| ColaEuphoria | ehhh | 08:20 |
| ColaEuphoria | This is starting to become a problem. | 08:20 |
| ColaEuphoria | I haven't even implemented envelopes yet. | 08:20 |
| ColaEuphoria | Is there any way to serialize matrices and ports that act on them? | 08:20 |
| *** ventosus has joined #lv2 | 09:00 | |
| *** falktx|work has joined #lv2 | 09:06 | |
| *** dsheeler has quit IRC | 09:18 | |
| *** ricardocrudo has joined #lv2 | 09:18 | |
| ventosus | ColaEuphoria: matrices no, but http://lv2plug.in/ns/ext/port-groups/ may at least help structuring ports together | 09:27 |
| ventosus | not all hosts support them, though | 09:27 |
| ColaEuphoria | I see. | 09:29 |
| ColaEuphoria | Perhaps I'll start working on a "port layout" feature entirely in ttl and mail you guys and see how it goes | 09:32 |
| ColaEuphoria | Although I have finals this week so it may take a few days before I can get started. | 09:32 |
| ventosus | ColaEuphoria: there is always the possibility to run your own gui for such special things, of course https://github.com/OpenMusicKontrollers/midi_matrix.lv2 | 09:35 |
| falktx|work | ssj71: btw, those CURIE errors about missing ui prefix are quite serious now | 09:35 |
| falktx|work | ssj71: with the latest lilv from git plugins with such errors are not loadable | 09:35 |
| falktx|work | got one in casynth.ttl | 09:35 |
| ColaEuphoria | ventosus: I could make an external gui, although I also see potential in serialization | 09:37 |
| ColaEuphoria | it could allow for people to make their own "preset layouts" | 09:38 |
| ColaEuphoria | so that users aren't at the mercy for the maintainers to make an aesthetically pleasing GUI | 09:38 |
| *** dsheeler has joined #lv2 | 09:38 | |
| falktx|work | the mod wizard kinda does that already | 09:39 |
| *** ssj71 has quit IRC | 09:49 | |
| *** gianMOD has joined #lv2 | 10:31 | |
| *** brummer has joined #lv2 | 10:34 | |
| *** gianMOD has quit IRC | 10:42 | |
| *** gianMOD has joined #lv2 | 11:16 | |
| *** gianMOD has quit IRC | 11:27 | |
| *** brummer has quit IRC | 12:26 | |
| *** artfwo has joined #lv2 | 12:38 | |
| *** ricardocrudo has quit IRC | 13:37 | |
| *** NickSB2_ has quit IRC | 14:16 | |
| rgareus | drobilla: I'm already looking forward to rolling out new liblilv http://pastebin.com/dL5PWKgB | 14:26 |
| *** artfwo has quit IRC | 14:43 | |
| drobilla | rgareus: Hrm. "error reading file" errors but no parse error printed.. | 15:07 |
| drobilla | oh, failed to open that one. n/m | 15:08 |
| rgareus | well, 600-700 lines due to LV2 plugins.. dwarving the actual backtrace. | 15:09 |
| falktx|work | who reported that? | 15:21 |
| falktx|work | quite some old plugins there | 15:21 |
| *** rncbc has joined #lv2 | 15:25 | |
| *** rncbc has quit IRC | 16:35 | |
| *** rncbc has joined #lv2 | 16:36 | |
| *** ssj71 has joined #lv2 | 16:36 | |
| ssj71 | falktx: why doesn't it show up with sord_validate? | 16:36 |
| *** son0p has quit IRC | 16:39 | |
| falktx|work | a bug in sord_validate I guess | 16:40 |
| ssj71 | casynth.ttl uses prefix lv2ui did you patch it and use just ui? | 16:40 |
| falktx|work | I didn't touch that file | 16:41 |
| *** son0p has joined #lv2 | 16:41 | |
| ssj71 | hrm. when I open it in jalv it gives me no CURIE error | 16:41 |
| ssj71 | I'm really baffled | 16:42 |
| falktx|work | you don't have latest lilv git? | 16:42 |
| falktx|work | if you use the kx repos I haven't rebuilt jalv against it yet | 16:42 |
| ssj71 | then no | 16:43 |
| ssj71 | and I haven't updated for a week or so | 16:43 |
| falktx|work | the new lilv is more picky :) | 16:43 |
| falktx|work | I'm waiting for next week to start pushing all the updates | 16:43 |
| ssj71 | ya there's plenty on your plate | 16:44 |
| ssj71 | I'm updating. will anything show me the error? sord_validate? carla? | 16:44 |
| falktx|work | mod-host | 16:44 |
| ssj71 | rats. I need to close firefox (web-irc client) biab | 16:46 |
| *** son0p has quit IRC | 16:46 | |
| *** ssj71 has quit IRC | 16:46 | |
| *** son0p has joined #lv2 | 16:47 | |
| *** ssj71 has joined #lv2 | 16:47 | |
| *** artfwo has joined #lv2 | 17:13 | |
| *** deva has joined #lv2 | 17:13 | |
| *** brummer has joined #lv2 | 17:20 | |
| *** ssj71 has quit IRC | 17:20 | |
| drobilla | serdi will show parse errors | 17:27 |
| falktx|work | oh hmm | 17:28 |
| falktx|work | error: failed to expand CURIE `synthv1_lv2:103_LFO1' | 17:28 |
| falktx|work | error: /usr/lib/lv2/drumkv1.lv2/drumkv1.ttl:357:35: expected `]', not `;' | 17:28 |
| *** ssj71 has joined #lv2 | 17:29 | |
| falktx|work | rncbc: you might want to look at that^ | 17:30 |
| falktx|work | using 0.7.2 here | 17:30 |
| rncbc | falktx|work: let'see | 17:32 |
| drobilla | (It'll also pretty print, so pipe through serdi is my usual first pass; or > /dev/null just for the errors) | 17:33 |
| rncbc | falktx|work: i don't find anything here, may ou look in your drumkv1.ttl ? | 17:35 |
| rncbc | falktx|work: using all drobilla.nat/lad git stack from... today! | 17:36 |
| falktx|work | you use synthv1_lv2 prefix on drumkv1, that's the error | 17:36 |
| rncbc | falktx|work: where? | 17:37 |
| falktx|work | line 357 | 17:37 |
| falktx|work | also this | 17:37 |
| falktx|work | error: failed to expand CURIE `synthv1_lv2:103_LFO1' | 17:37 |
| falktx|work | error: /usr/lib/lv2/samplv1.lv2/samplv1.ttl:367:35: expected `]', not `;' | 17:37 |
| rncbc | falktx|work: ah | 17:37 |
| falktx|work | same with samplv1 ttl, uses synthv1_lv2 prefix | 17:38 |
| rncbc | falktx|work: you're right | 17:38 |
| falktx|work | thing is, with the new lilv I can no longer load this plugin :P | 17:38 |
| falktx|work | good luck I haven't rebuilt everything in kxstudio yet :) | 17:38 |
| *** sigma6 has quit IRC | 17:40 | |
| rncbc | falktx|work: good catch | 17:41 |
| rncbc | falktx|work: fix is str8fwd! just patch those .ttl's! | 17:44 |
| falktx|work | see the nickname. busy building other stuff :P | 17:44 |
| rncbc | falktx|work: i never said it has to be ptached this year | 17:45 |
| rncbc | :) | 17:45 |
| rncbc | falktx|work: wait, you're deploying tht new lilv? then why not making the fix as you go anyway? | 17:47 |
| falktx|work | I will. the v1 series are not on mod right now | 17:48 |
| falktx|work | for kxstudio I only updated it because of mod stuff | 17:48 |
| drobilla | rgareus: Should we maybe add a preferredBlockLength property while we're at it? | 17:51 |
| rncbc | falktx|work: ye know, the v1s are not quite elligigle for mod as they use the outrageous instance-access bwahahah | 17:53 |
| falktx|work | yep | 17:53 |
| falktx|work | although synthv1 might work | 17:53 |
| drobilla | FWIW the previous "tolerant" (but not really) approach lilv took almost certainly resulted in botched half-loaded data for plugins | 17:53 |
| *** edogawa has quit IRC | 17:59 | |
| rgareus | drobilla: I wonder which host would implement that. ardour could theoretically IFF preferredBlockLength is smaller than the current block-size | 17:59 |
| rgareus | drobilla: but most hosts are really bound to the user-configured block-length for JACK (or whatever backend) | 18:00 |
| rgareus | drobilla: Oh, and I forgot to tell you, the colorful jalv --dump output is awesome. | 18:01 |
| *** ssj71 has quit IRC | 18:03 | |
| *** ventosus has quit IRC | 18:06 | |
| rgareus | on http://lv2plug.in/pages/developing.html just below "What not to do" .. can we add | 18:06 |
| rgareus | "There is zero-tolerance for violating the specs. All violations will be prosecuted." ;) | 18:06 |
| rgareus | signed.. the LV2 High Inquistor. | 18:07 |
| *** ventosus has joined #lv2 | 18:07 | |
| * rncbc goind for a reboot, new patched kernel 4.1.15.rt15... | 18:10 | |
| *** rncbc has quit IRC | 18:10 | |
| *** ssj71 has joined #lv2 | 18:10 | |
| *** rncbc has joined #lv2 | 18:12 | |
| rgareus | rncbc: that was quick | 18:12 |
| rncbc | rgareus: this is a i7 :) | 18:12 |
| rncbc | rgareus: and also a ssd | 18:13 |
| rgareus | rncbc: 4.1.15-rt15 works well here (also i7) but sucks a lot of power (nearly twice as much as 4.2.6, cutting the battery lifetime in two) | 18:13 |
| rncbc | rgareus: weel this is a custome kconfig... | 18:14 |
| rgareus | 8 or 9 Watts idle with 4.1.15-rt15 , 3-4 Watts on 4.2.6 according to powertop | 18:14 |
| rgareus | on a Thinpad X250 | 18:14 |
| rgareus | Thinkpad, even | 18:14 |
| rgareus | rncbc: is your setup on a laptop? | 18:16 |
| rncbc | rgareus: yep, toshiba z30 | 18:16 |
| rgareus | rncbc: so you're potentially concearnd about power-consumption as well | 18:17 |
| rncbc | rgareus: yes, but i never got passed the 5h mark anyway | 18:18 |
| rncbc | rgareus: the thing was marketed as a 8h batt life | 18:18 |
| rgareus | I was surprised recently. the X250 held out a cross atlatic flight and had battery left. spec'ed for 24h. I get maybe 12-14h out of it. (depends how much I compile) | 18:19 |
| rgareus | but that was with a near brand-new battery. | 18:20 |
| rncbc | rgareus: depends also if you jkeep jackd up all the time :) | 18:20 |
| rgareus | true | 18:20 |
| rgareus | I used to do that on my old laptop (start it with X). | 18:20 |
| rncbc | rgareus: running jackd -p64 sucks or should i say fries the batt out of hell? :) | 18:22 |
| * rgareus 's current strategy is to reboot into rt-kernel and only then use jack when doing serious audio-work. | 18:23 | |
| rgareus | and in those cases a power-outlet is usually closeby, as well | 18:23 |
| ssj71 | archlinux users are reporting the my new lushlife UI doesn't draw in ardour 4.4.0. My other GUIs show fine, and I can't reproduce with kxstudio ardour 4.4.143. Anyone have a suggestion on how I can try to debug this?? | 18:26 |
| rncbc | rgareus: if i don't do like several make -j4 kernel builds, like i do, or run jackd, this laptop might just make it catalogged 8h battery life for one :) | 18:28 |
| rncbc | for once | 18:28 |
| rgareus | ssj71: any C++ in the GUI? maybe a gcc4/gcc5 issue? what gui toolkit do you use? | 18:28 |
| ssj71 | ntk. it is c++ | 18:29 |
| rncbc | but, you never know... i keep the good prective to recharge whenever battery is just below 51%... | 18:29 |
| rncbc | practice* | 18:30 |
| rgareus | ssj71: does ntk still has this "return pointer to X-window" issue? (like all of harry's plugins which return a pointer to the fltk-window) | 18:30 |
| ssj71 | I thought I resolved that, It returns the XID | 18:31 |
| rgareus | ok. | 18:31 |
| ssj71 | https://github.com/ssj71/infamousPlugins/blob/master/src/lushlife/lushlife_ui_main.cxx#L65 | 18:31 |
| rgareus | ssj71: well, a first test is: does the GUI .so load. (no missing symbols etc). IIRC that's reported on stderr. | 18:32 |
| * rncbc now reboots into vanilla 4.3.3 ... | 18:33 | |
| *** rncbc has quit IRC | 18:33 | |
| ssj71 | I know a window is created. I'll ask if they can test running ardour in the termincal | 18:33 |
| rgareus | ssj71: and there will be a more "userfriendly" error in Ardour's Menu > Window > Log. | 18:33 |
| rgareus | ssj71: maybe it's a simple as the .ttl file missing lv2:extensionData ui:idleInterface | 18:34 |
| rgareus | but then again it should also not work on your box | 18:34 |
| ssj71 | right | 18:34 |
| *** rncbc has joined #lv2 | 18:34 | |
| ssj71 | well, I'll try to get more feedback | 18:35 |
| *** NickSB2 has joined #lv2 | 18:36 | |
| rncbc | falktx: samplv1, drumkv1 lv2 .ttl typo fixed upstream v0.7.2.1 | 18:45 |
| *** rncbc has quit IRC | 18:46 | |
| falktx|work | cool | 18:47 |
| *** rncbc has joined #lv2 | 18:48 | |
| *** rncbc_ has joined #lv2 | 18:49 | |
| rncbc | falktx: thanks | 18:50 |
| *** son0p has quit IRC | 18:59 | |
| ventosus | ssj71: archer here, can confirm no UI in Ardour only. commenting ''self->pitchfreq0->do_callback" and "self->pitchfreq1->do_callback" in port_event shows an UI | 19:31 |
| ventosus | ssj71: there may be a feedback loop in there somewhere, ardour sends increasing values to these two control ports (until inifinity) | 19:36 |
| ssj71 | yes, must be a feedback loop. I;ll look into that. Thats very helpful thanks | 19:41 |
| *** falktx|work has quit IRC | 19:55 | |
| rgareus | ventosus, ssj71: might be worth checking recent ardour. | 19:58 |
| rgareus | the port-event logic has been updated. | 19:59 |
| *** brummer has quit IRC | 19:59 | |
| rgareus | ardour now initially sends a value to all GUI ports when the gui is [re]-opened (it didn't do that before) | 19:59 |
| rgareus | and after that port_events only when the value changes. | 20:00 |
| rgareus | ardour also notifies the GUI of changes made by itself. (to enable automation touch and playback modes. e.g gui sets value, ardour says "no, automation is played back -> use value X". | 20:01 |
| rgareus | ^^^ "and after that port_events only when the value changes." -> "and after that *emit* port_events only when the value changes." | 20:02 |
| rgareus | KXStudio ships Ardour 4.4.140 (or something) that already has the reworked port-event mechanism | 20:04 |
| ventosus | rgareus: ArchLinux is at Ardour4.4.0 (built using 4.4 and GCC version 5.3.0) | 20:07 |
| rgareus | the "old" way: plugins will have to keep track of changes themselves. | 20:08 |
| rgareus | ventosus: the loopback you describe is reasonable. | 20:08 |
| *** ssj71 has quit IRC | 20:15 | |
| *** ssj71 has joined #lv2 | 20:22 | |
| *** uncle-j_j has joined #lv2 | 20:22 | |
| *** ssj71 has quit IRC | 20:23 | |
| *** ssj71 has joined #lv2 | 20:23 | |
| ssj71 | ventosus: sorry, lost my backlog, which lines did you comment? | 20:24 |
| ventosus | ssj71: 'self->pitchfreq0->do_callback" and "self->pitchfreq1->do_callback" | 20:25 |
| ssj71 | ventosus: thanks | 20:26 |
| ssj71 | I'm guessing its because the callbacks resend the value through the write function | 20:27 |
| ssj71 | ok, ya, I see some issues. | 20:32 |
| *** ssj71 has quit IRC | 20:36 | |
| *** coderkevin has quit IRC | 20:44 | |
| *** ssj71 has joined #lv2 | 20:46 | |
| ssj71 | ventosus: I've just pushed a change that I think should make the loop stop. any chance you could build from master and test? | 20:57 |
| falktx | can't you just valgrind the thing? | 20:59 |
| ssj71 | falktx: good idea. I'll try | 21:02 |
| ssj71 | valgrind or callgrind? | 21:02 |
| falktx | valgrind | 21:03 |
| ventosus | ssj71: still no UI | 21:03 |
| ventosus | ssj71, rgareus: "new" Ardour (4.4.489) indeed shows the UI just fine | 21:07 |
| ventosus | even the unpatched one | 21:07 |
| ssj71 | interesting | 21:08 |
| rgareus | falktx: how would valgrind help here? | 21:14 |
| falktx | there is a possible null pointer usage here | 21:14 |
| rgareus | falktx: aah, ok. I thought it's an endless feedback loop (port events) | 21:15 |
| ssj71 | I think it is. If I send a value through the write function, does ardour send it back to the UI as a port event too? | 21:16 |
| ventosus | the loop ends pretty quickly (7-8 iterations), when the values for ports (12, 21) reach INF ;-) | 21:17 |
| ssj71 | ventosus: but only with ardour 4.4.0? | 21:18 |
| ssj71 | the issue is that the widgets store a linear value but actually send a quadratic to the dsp (to approximate log scale) | 21:19 |
| ssj71 | port events were setting the log scale value which then got squared again (causing it to grow and grow) | 21:20 |
| ssj71 | my change shouldn't actualyl end the loop, but it should prevent the diverging | 21:20 |
| ventosus | ssj71: yes only for 4.4.0, 4.4.489 does not show increasing values for those ports | 21:21 |
| ssj71 | ventosus: did you test if latest master with 4.4.0 still diverges? | 21:24 |
| ventosus | ssj71: I'm at commit a14729d6a1a65e3e9dc0f11051dd28b5f050b568 | 21:27 |
| ventosus | 4.4.0 still diverges, yes | 21:27 |
| ssj71 | ok | 21:28 |
| ssj71 | hm | 21:28 |
| ssj71 | ack! I think I didn't actually commit the change | 21:38 |
| ssj71 | ventosus: I'm sorry didn't actually make the change. If you could test commit ec4fd66 (latest master) it would be awesome | 21:39 |
| ssj71 | test with 4.4.0 please | 21:39 |
| *** deva has quit IRC | 21:43 | |
| ventosus | ssj71: fixed in 4.4.0, and still fine in 4.4.489 ;-) | 21:49 |
| ssj71 | WOOT! | 21:49 |
| ssj71 | thanks a ton ventosus! | 21:50 |
| rgareus | ssj71: can you sumarize what the issue was and how you fixed it? | 21:50 |
| ventosus | np | 21:50 |
| ssj71 | rgareus: port event calls the callback of the widget so that it redraws the data and affects other widgets properly | 21:51 |
| ssj71 | rgareus: in said callback, the data is scaled in a log approximation to provide better resolution | 21:51 |
| ssj71 | the scaled data is sent through the write function back to the dsp, and I believe ardour would then see the dsp value had changed and send another port event | 21:52 |
| rgareus | aah 'rounding' (log approx) -> value differs.. | 21:52 |
| ssj71 | rgareus: the solution was the first reverse the scaling of the value, then call the callback | 21:52 |
| ssj71 | so it was much worse than rounding error, it was exponential | 21:53 |
| ssj71 | its a quadratic scale, so callback(val); write_function(val*val); callback(val*val) etc | 21:54 |
| ssj71 | solution is to change it to callback(sqrt(val)) | 21:54 |
| rgareus | ssj71: can't you just block signal emission (if the widget is updated via port_event() ? | 21:57 |
| rgareus | in gtk/sigc++ it'd be sth like widget_connection.block(); widget.set_value(new_value); widget_connection.unblock(); | 21:58 |
| ssj71 | yes, but it will take a little architecture change that I'd rather not do if not necessary. Currently it doesn't feed back. | 21:58 |
| rgareus | I use a similar approach in in robtk to suspend signal emission. | 21:58 |
| ssj71 | I actually deliberately call the callback, its not called from the signal | 21:59 |
| rgareus | ssj71: in most cases you'll be fine but with double or float val == sqrt (val * val) is not a give. | 21:59 |
| rgareus | n | 21:59 |
| ssj71 | rgareus: I know, but it seems to be close enough for engineering | 22:00 |
| rgareus | yeah, and even if it mismatches once in a while it'll probably stop looping at some point. | 22:00 |
| ssj71 | right. I tested a bit in jalv and never saw a loop | 22:01 |
| rgareus | and in 1/2 a years time when everyone moved to Ardour 4.5 (or newer) it's water under the bridge anyway | 22:02 |
| ssj71 | rgareus: a4.5 is that far out? that adds tabbed mode right? | 22:02 |
| rgareus | probably not... | 22:03 |
| rgareus | there were another huge amount of fixes/updates. since 4.4 | 22:03 |
| rgareus | las was supposed to tag 4.5 yesterday... but then some personal issues came in-between. | 22:04 |
| ssj71 | ah. so 4.5 is close, but tabbed mode isn't landed. It'll be ready when its ready | 22:04 |
| rgareus | yep | 22:04 |
| rgareus | well, tabbed is ready to be merged. | 22:04 |
| rgareus | tabbed is also held back by Mixbus (upcoming 3.1 includes control-surface support), las is finishing that. Mixbus 3.1 will come w/o tabbed. | 22:05 |
| *** ventosus has left #lv2 | 22:13 | |
| ssj71 | ah too bad. I'm not sure I will even use tabbed mode, but its a flashy feature that I'm sure many will find useful | 22:20 |
| falktx | ssj71: should I update infamous-plugins package in kx? | 22:41 |
| ssj71 | falktx: there is one issue still outstanding that I cannot reproduce. bluebell reports that opening the GUI in qtractor a second time crashes qtractor | 22:42 |
| falktx | dangling pointer I guess | 22:42 |
| ssj71 | falktx: I need more feedback, but he's asleep | 22:42 |
| falktx | fl_open_display being called twice? | 22:43 |
| falktx | hmm I don't think I should look at that code too much... | 22:44 |
| ssj71 | too ugly? | 22:45 |
| falktx | you're checking if a pointer returned by 'new ...' is null | 22:46 |
| falktx | that never happens unless you specifically tell the compiler to disable some stuff | 22:46 |
| ssj71 | I guess I'm being too safe :) | 22:46 |
| falktx | if 'new' fails it will throw | 22:47 |
| falktx | returned pointer is *always* valid | 22:47 |
| falktx | also "self->ui = self->show();" feels very wrong to me... | 22:47 |
| ssj71 | this was copied from harry | 22:47 |
| falktx | harry's code is very dirty... | 22:47 |
| ssj71 | somebody should have told me that before I copied it all | 22:48 |
| falktx | that "self->ui = ..." + "delete self->ui;" feels crashy | 22:48 |
| falktx | what kind of pointer does self->show() return? | 22:48 |
| ssj71 | an Fl_Window | 22:49 |
| falktx | are you allowed to delete that? | 22:49 |
| ssj71 | don't know. Its just a window to me | 22:50 |
| falktx | well, it's very important | 22:50 |
| falktx | no idea why a 'show' method would allocate a new instance of an UI. doesn't feel right | 22:51 |
| ssj71 | I understand that, the documentation doesn't forbid it anywhere I can see | 22:51 |
| ssj71 | show is probably the wrong name | 22:51 |
| falktx | if show does not allocate a new thing, then you can't delete it | 22:51 |
| falktx | yes | 22:52 |
| falktx | I can't really read that fltk/ntk fl code | 22:52 |
| falktx | in other news, the max-gen to real linux plugins is working very well | 22:54 |
| ssj71 | falktx: here's the generated code: http://pastebin.com/0LRSZfdW | 22:54 |
| ssj71 | line 1202 | 22:54 |
| *** edogawa has joined #lv2 | 22:55 | |
| falktx | haha, so it is creating a new window | 22:55 |
| falktx | Fl_Double_Window* o = ui = new Fl_Double_Window(630, 400, "the infamous lush life"); | 22:55 |
| falktx | so the code is fine | 22:55 |
| falktx | just the function name that is terrible :P | 22:55 |
| ssj71 | yep. Inherited from openAV productions :) | 22:57 |
| ssj71 | I hadn't worked with any toolkits much, I thought artyfx were a simple example to copy | 22:57 |
| falktx | they are. but made in a hacky way | 22:58 |
| falktx | I was the one that introduced the fltk embed code to male, some years ago. which allowed ntk to work as a real plugin ui toolkit | 22:58 |
| ssj71 | I didn't know that | 22:58 |
| falktx | based on some fltk browser thing | 22:58 |
| falktx | they reworked it a bit | 22:59 |
| falktx | then later I added another hack to make ntk menus work when embed | 22:59 |
| falktx | lots of hacks in there ;) | 22:59 |
| falktx | I mean, just look at it https://github.com/original-male/ntk/commit/5719b0044d9f267de5391fab006370cc7f4e70bd | 22:59 |
| ssj71 | you think the current AVTK is much better? | 22:59 |
| falktx | haven't seen it much | 22:59 |
| falktx | but it should be yes. based on pugl | 23:00 |
| ssj71 | I want to switch, but its going to be another big time investment, and doesn't feel like a big gain | 23:00 |
| ssj71 | I should say I want it to BE switched, but I don't really want to do it :) | 23:01 |
| * falktx proposes DPF for a sane long term investment | 23:01 | |
| falktx | sometime later you might want to release cross-platform | 23:02 |
| ssj71 | problem is my obsessive goal is to have scalable UIs made in inkscape. Not sure DPF gets me there. Also I've become pretty comfortable with LV2, so another API seems like more growing pains | 23:02 |
| falktx | even if not win/mac, but perhaps bsd or haiku | 23:03 |
| ssj71 | yes, when it comes to that I'll have to move to something. | 23:03 |
| falktx | there's a project for svg parsing, made by the same author as nanovg | 23:03 |
| falktx | so I imagine it's possible to render the svg stuff directly into opengl | 23:03 |
| ssj71 | I've got a pretty good workflow now converting the svgs to cairo | 23:04 |
| falktx | it already is for icons btw. using the now common icon-as-font-character trick | 23:04 |
| ssj71 | so AVTK is probably path of least resistance | 23:04 |
| falktx | yes | 23:04 |
| ssj71 | IDK. I'll look at that more when I feel more need to be cross platform. In some ways I WANT to have linux only plugins | 23:05 |
| ssj71 | make those other platforms jealous ;) | 23:05 |
| falktx | someday I might switch to haiku | 23:07 |
| falktx | linux is getting a bit too mainstream | 23:07 |
| falktx | everything working, nothing to fix... boring... | 23:07 |
| ssj71 | lol :) | 23:07 |
| ssj71 | I'm pretty good with it as long as it stays open source | 23:08 |
| ssj71 | once all the proprietary companies start moving over I'll look for something else | 23:08 |
| ssj71 | also if you think everything is working, get some brand new mac hardware. That will give you plenty to fix. | 23:09 |
| *** nordstrom has quit IRC | 23:10 | |
| *** nordstrom has joined #lv2 | 23:11 | |
| *** nordstrom has quit IRC | 23:12 | |
| rgareus | ssj71: is AVTK already released? I thought Harry's still working on it. | 23:13 |
| falktx | it's "released" on git | 23:13 |
| ssj71 | rgareus: I don't think its officially released, but it hasn't had a commit for months | 23:13 |
| rgareus | I thought avtk is pugl (GL) not FLTK | 23:13 |
| *** nordstrom has joined #lv2 | 23:14 | |
| ssj71 | the original avtk that sorcer and 1st gen artyfx were ntk. I forked that for my GUIs | 23:15 |
| rgareus | http://openavproductions.com/avtk/ | 23:15 |
| ssj71 | not too long after I did that he scrapped ntk and went for pugl | 23:15 |
| rgareus | ssj71: sorcer etc are NTK - not avtk | 23:15 |
| rgareus | avtk is pure pugl + cairo (a complete departure from fltk/ntk). | 23:16 |
| ssj71 | rgareus: https://github.com/harryhaaren/openAV-Sorcer/tree/master/gui/avtk | 23:16 |
| ssj71 | the folder is called avtk | 23:16 |
| ssj71 | the namespace is avtk. They were just ntk widgets | 23:17 |
| rgareus | ok, so harry uses the same name for 2 different projects. | 23:17 |
| ssj71 | yes, it should have been avtk 2.0 or something | 23:18 |
| ssj71 | but I think since he never actually released avtk he didn't bother | 23:18 |
| *** nordstrom has quit IRC | 23:18 | |
| rgareus | I have not seem him around much, either. | 23:19 |
| rgareus | last I heard he's got a new job (doing some sysadmin/network stuff??) | 23:20 |
| ssj71 | ya something like that. jobs are work killers ;) | 23:20 |
| rgareus | he should have stayed on the road with the Kellies :) | 23:21 |
| *** nordstrom has joined #lv2 | 23:21 | |
| *** nordstrom has quit IRC | 23:22 | |
| *** nordstrom has joined #lv2 | 23:22 | |
| *** nordstrom has quit IRC | 23:22 | |
| *** uncle-j_j has quit IRC | 23:23 | |
| *** nordstrom has joined #lv2 | 23:24 | |
| falktx | harry will have some free time starting next week, or so I've heard.. | 23:26 |
| falktx | I know of a specific hardware thing he will be beta-testing... :D | 23:26 |
| falktx | gnight | 23:26 |
| *** rncbc has quit IRC | 23:28 | |
| *** edogawa has quit IRC | 23:34 | |
| *** nordstrom has quit IRC | 23:40 | |
| *** nordstrom has joined #lv2 | 23:42 | |
| *** nordstrom has quit IRC | 23:46 | |
| *** nordstrom has joined #lv2 | 23:49 | |
Generated by irclog2html.py 2.13.0 by Marius Gedminas - find it at mg.pov.lt!
