csd4ni3l

Day 3 (very late devlog lmao)
I fixed the html & css cache, and i now use base64 for cache filenames. The app now automatically updates tab titles, stores the old url & reverts it to the current tab’s url when needed. I made the browser work by requesting the url given on enter, and i added the css/html parse and receive. I fixed the Host header not changing for css urls (which wasnt fixed in the previous Python version). and i am now using better & less complicatedJSON serialization for css cache. Other than that, I improved the renderer’s boilerplate code, which will be the hardest part. I havent worked for a bit of time on this.

I added a feature table to the README, added youtube downloader support with yt-dlp, added ffmpeg & yt-dlp auto-download for Windows and warns for all platforms if missing, and a warn on Windows when VB Cable is missing, made the bottom panel fixed size so content doesn’t go up and down when playing new sound effects, added more apps to Linux excludes, added support for multiple virtual outputs by using buttons(on/off) on Linux so you can select the virtual mic for multiple apps at the same time, and removed the ffmpeg help output in the cmd. I last minute also fixed a bug where it never set permissions for yt-dlp so it would crash sometimes. I still need to work on the UI though.

Day 1 & 2
Basically I started working on moving the whole code to Rust, by creating a new git branch, moving all code to a python folder and then initializing a Bevy project. The main UI was pretty easy, but then came the HTML parser and Connection, which was a lot harder to port. mutability, references, borrowing makes it so much harder in Rust than in Python. Of course, it’s not impossible but its more complicated. I hope it’s gonna be worth it for the learning & the performance. Also, i figured out that bevy_egui might not work well for this, so i might have to use something else, but we will see… I am currently working on the html & css cache and porting the full connection & html parsing mechanisms, and then comes rendering, which is probably gonna be the hardest part.

Welp, again a long time cause idk im not having the motivation to do much. I basically only need Windows support now though and to test it. I formatted the code, added resume all & pause all buttons, and Linux works without any issues now, meaning you can speak and others will hear you, I also improved current directory highlight, Windows might work now as well, but haven’t tried, fixed a bunch of Rust issues, i updated the README as well to better reflect the current situation, available apps on Linux now update automatically, and show a no apps label if none found. Windows & MacOS show a not supported label.

Well i forgot to devlog frequently again. So, i added Github Actions so its runnable now! And its cached so really fast. I had to fix a bunch of issues and add libraries for Linux many times so thats why the repo has so many commits. I separated Windows and Linux stuff into different files and added virtual mic set support for Linux (meaning it will automatically switch the selected app to the virtual mic.). I also removed unused Bevy features to reduce size of the EXE from 80mb to 60mb!! I tried adding both virtual mic & normal sound play support but nothing yet.

Well there was the Christmas break and i took a HUGEE break from coding and forgot to devlog, but i didnt do that much. I made the base UI using bevy_egui, made it work (but its hacky and only on Linux) and i need to add a lot more stuff still. Rust is cool but borrowing and pointers are such a hassle. In the attached image i didnt want to leak any of my files but yeah it shows only the supported file formats.

First of all, i started learning Rust yesterday, and today Bevy. I didnt do much yet but its on Github now and the base libraries are added with some boilerplate code..