Probably because of a lack of proper realtime audio due to how the audio input layer was ported, but that's hypothetically fixable by changing the architecture.
I wonder if we'll ever see a move towards "headless" apps that have the business logic written in whatever language(s) shared as libraries and APIs, but each platform would then be able to write its own UI frontend in its native stacks.
In the case of Audacity, that could mean all the audio processing stuff in WASM but the UI could be done as a traditional web app instead of a ported canvas.
Isn't that more like a VNC type thing where the server renders the bitmaps and sends them to the client?
That's not really the same as a native Android app or a Javascript web app or a ios app being able to share the same libraries while having bespoke UIs. If anything X encourages the use of the same UI across different devices, which is the opposite of the way real users ended up using apps (with different interfaces for desktop, tablet, mobile, and watch).
Debian's x11-apps has a collection of what I think of as pretty old-school-ish apps. They all feel monolithic: the app is both a backend and front end.
There is no decoupling, where the same headless app has different possible front ends.
An example of a popular headless app is neovim. You can run it in headless mode, and then connect different gui clients to neovim. A lot of the projects here are good examples of independent clients: https://github.com/topics/neovim-guis
11 comments
[ 2.8 ms ] story [ 40.3 ms ] threadhttps://news.ycombinator.com/item?id=37356497
I believe you can use it to edit audio, but it doesn’t work with a microphone.
In the case of Audacity, that could mean all the audio processing stuff in WASM but the UI could be done as a traditional web app instead of a ported canvas.
Just a thought.
That's not really the same as a native Android app or a Javascript web app or a ios app being able to share the same libraries while having bespoke UIs. If anything X encourages the use of the same UI across different devices, which is the opposite of the way real users ended up using apps (with different interfaces for desktop, tablet, mobile, and watch).
There is no decoupling, where the same headless app has different possible front ends.
An example of a popular headless app is neovim. You can run it in headless mode, and then connect different gui clients to neovim. A lot of the projects here are good examples of independent clients: https://github.com/topics/neovim-guis