anklang
anklang copied to clipboard
tim-janik
libgail.so: undefined symbol: g_once_init_enter_pointer
I am loading Vital plugin, then click on Toggle GUI, the app crash and I get this error:
/tmp/.mount_anklanb3qO8M/anklang-0-3/electron/resources/app/../../../lib/AnklangSynthEngine-fma: symbol lookup error: /usr/lib/x86_64-linux-gnu/gtk-2.0/modules/libgail.so: undefined symbol: g_once_init_enter_pointer
AnklangSynthEngine-fma: exit: 127
I try to install libgail, but it does still crash:
sudo apt-get install libgail-3-dev libgail-dev libgail18t64 libgail-3-0t64
This an amazing project!! I was starting over the weekend to build something similar with Juce framework https://www.reddit.com/r/JUCE/comments/1ilpaby/multi_track_sequencer_for_vital/ but it was really a pain and decided to not to continue. Now that I see your project, I really like your stack, using Electron to make the UI is really good idea. As I am professional web developer, maybe I could follow your concept and use Electron ;-) or simply use Anklang as it seems to be more or less what I wanted to build anyway :-)
I am loading Vital plugin, then click on
Toggle GUI, the app crash and I get this error:/tmp/.mount_anklanb3qO8M/anklang-0-3/electron/resources/app/../../../lib/AnklangSynthEngine-fma: symbol lookup error: /usr/lib/x86_64-linux-gnu/gtk-2.0/modules/libgail.so: undefined symbol: g_once_init_enter_pointer AnklangSynthEngine-fma: exit: 127I try to install
libgail, but it does still crash:sudo apt-get install libgail-3-dev libgail-dev libgail18t64 libgail-3-0t64
Please provide a link to the plugin version you were testing, I can look into reproducing it.
This an amazing project!! I was starting over the weekend to build something similar with Juce framework https://www.reddit.com/r/JUCE/comments/1ilpaby/multi_track_sequencer_for_vital/ but it was really a pain and decided to not to continue. Now that I see your project, I really like your stack, using Electron to make the UI is really good idea.
The project suffers from too little resources, in particular building a UI and a sound engine in parallel is taking too long to really be a fun project. That is why on my home machine, I have been looking into/hacking on moving the sound engine over to tracktion_engine which is coincidentally Juce (AGPL) based. This talk provides a lot of background for the motivation: Why You Shouldn’t Write a DAW - David Rowland - ADC23
As I am professional web developer, maybe I could follow your concept and use Electron ;-) or simply use Anklang as it seems to be more or less what I wanted to build anyway :-)
Let me know what you would like to work on in a DAW and what you'd want to achieve, I would be happy to exchange ideas and explore what overlaps there are. FYI, we have been having weekly meetings until last summer that halted for a while and that we are just now picking up again: Regular-Meetings And you can reach us via IRC #Anklang or Discord #jj-fzf.
The plugin i was testing was Vital 1.5.5:
- https://vital.audio/
- https://github.com/mtytel/vital
Lot of people speak about the tracktion engine, but the licence is a bit pushing me away... Originally when I made my little prototype with Juce, I just wanted to try out something quick using the amazing Vital synth engine. This is just because I never been happy about existing DAW that are way too complicated an mainly focus to build song, and I was thinking to make something simple more focus on live performance like a groovebox. My final intention was to bring it on a mini computer and build the midi controller for it... I think Anklang could make the job so i hope the project will develop further and stay simple as it is now :-) Unfortunately, when I look into your code, I see that you are using JavaScript and Vue, I am using TypeScript and React... (not sure I would be up to step back to JavaScript and I am a bit lazy to learn another frontend library like Vue). I think I will just go back to my orginal project https://github.com/apiel/zicBox and stick to it :-) But I will keep a close eyes to your project because i think it could turn out to be something really cool.
The plugin i was testing was Vital 1.5.5:
* https://vital.audio/ * https://github.com/mtytel/vitalLot of people speak about the tracktion engine, but the licence is a bit pushing me away...
The Tracktion license is GPLv3 and if you use it with Juce-7, that is also GPLv3. If we move to Tracktion based on Juce-8, that would mean that the AGPL from Juce-8 applies, but that'd be the case with any project using Juce. Since you have been using/considering Juce already, what other licensing concerns do you have?
Originally when I made my little prototype with Juce, I just wanted to try out something quick using the amazing Vital synth engine. This is just because I never been happy about existing DAW that are way too complicated an mainly focus to build song, and I was thinking to make something simple more focus on live performance like a groovebox. My final intention was to bring it on a mini computer and build the midi controller for it... I think Anklang could make the job so i hope the project will develop further and stay simple as it is now :-) Unfortunately, when I look into your code, I see that you are using JavaScript and Vue, I am using TypeScript and React...
We have been moving away from Vue ever since Vue3 went with hooks instead of classes/components. At this point, only the preference dialog and the shell are left in Vue and once those are ported, Vue can finally be removed entirely from the codebase.
(not sure I would be up to step back to JavaScript and I am a bit lazy to learn another frontend library like Vue). I think I will just go back to my orginal project https://github.com/apiel/zicBox and stick to it :-) But I will keep a close eyes to your project because i think it could turn out to be something really cool.
Locally, I have been using Typescript more than JS lately, and started to check with tsc --erasableSyntaxOnly and run scripts as node --experimental-strip-types script.ts instead of node script.js. It'll just take a while since everything is Typescript, luckily the move can be done incrementally.
As for the framework, after Vue3 moved away from classes, I looked into several options again and ended up using Lit, because that lets me stay as close to native Web components as possible. I'd rather not cause several future framework migrations if that can be avoided and native web components are the only thing that is for sure going to be around in 20 years.
React looks like a close second though, because it's use is widespread these days. But then it comes with a bunch of stuff (router, SSR) that we cannot make use of. So I plan to give it a test shot for one or two isolated components (that's also how I tried Lit initially), but if it is not a significant improvement over native web components, migrating yet again is probably not worth it. OTOH, another project I was contributing to (llama.cpp) just migrated to React, so I'll probably have to learn that anyway ;-]
zicBox looks nice, and interesting idea to include the wiki as a submodule ;-)
If you aim to go to TypeScript maybe you could use some tools to do the conversion for you, either with AI or maybe something like https://js2ts.com/ . Using vanilla js with lit for web component is maybe a smart move, unfortunately not much people go this direction today, even if I personally think that it should be way more the approach to go in most of the case. I never been so fanatic of React, especially because of the state management, but it has been the industry standard for long now, and of course I had to adapt to it. The only thing I really like in React is JSX but lithtml seems pretty nice as well... All depends what you want to do, if you want attract other developer, you will have much more chance if you use React, simply because it is way more popular. And maybe you should also try to combine it with UI library like Ant design or whatever other popular UI library. If you find the right combo, it could speed up your development and attract more people to help you.