Study the codebase

[hydroxycarbamide]

Unless I truly know each part of this software, I won't deploy it to Flathub.

In order to maintain the project properly but I need to:

• Understand how to get the CSS variables used in libAdwaita
• Assess if I actually have the time to maintain it
• Learn Gnome and Flatpak workflow and guidelines
• Learn ADW and Gnome Builder
• Analyse and restore the CI
• Understand the purpose of the Submodules (plugins, gnome-sass etc.)
• Try to fix a few bugs from the original issue list

I might also:

• Make improvements to the CI so I can be as lazy as possible
• Find a solution about yapsy and the plugin manager. Fork it, find an alternative or kill it? I don't want to maintain it though

[eylles]

hi, i'm happy to see there's someone else taking a jab at this, i created my own fork at https://github.com/eylles/Gradience back in october the 28th but unforutnately have not been able to really study the code in depth.

what i wanted to do with gradience wasn't simple maintenance but instead refactor it some, extract a good chunk of the function into a library and package it to pypi, then operate the code that deal with the gui to simply use that library on the backend and the build would simply depend on the library, that has not really materialized as i wanted.

Instead i choose to take a path of lesser resistance that would help me better understand what gradience needs to do, which was to begin implementing such library on my own, on the meantime i managed to write a thin and rather naive script that takes a gradience .json and outputs the correct css to the gtk3 and gtk4 dirs at https://github.com/eylles/pywal16-libadwaita/blob/master/scripts/gradience2css.py tho in the coming days i'll be working to polish the scripts from the repo into a more polished state so that i can import them onto a proper main.py and have the first version of something that can be used and packaged.

[hydroxycarbamide]

Hi @eylles , indeed it would be quite interesting to have a separate package for the backend of Gradience in order to avoid depending too much on the gui libraries. I was myself a little disappointed in the repositories when I first saw that it offered a cli but did not package its backend for its reusage.

I might try to package it but I would like to take over the repository properly and takes over the original org if possible. The obvious issue is that I am not part of the Gradience Team (the one from the original repository) and did not want to release "unofficial" and half-baked project.

When I forked the project, I wasn't quite sure I was ready yet to actually take over the original repository as my time management is quite messed up. And moreover I did not have a good grasp on how to handle Libadwaita colors variable updates (to be fair, I still do not know but I guess I can always try to find out live).

Right now, only secondary features a left not working (gnome sass and the plugin manager) and I do not want to maintain these but I want to release a proper version right now rather having nothing (or having only nightly builds).

[eylles]

yeh, that is reasonable, the plugins don't have a documented API, at least as far as i know, so if you want to put out a release that more or less works i think it would be okay to not include the plugins.

as for the material color utilities, i was looking into as i want to port the monet colorscheme generation backend to pywal16, as one of the scripts i got on the scripts i got on the pywal16-libadwaita repo is for generating gradience json preset from pywal colorschemes, which i wanted to expand upon and maybe for this backend library i'm building make pywal16 into a dependency to use it as a library to handle colorscheme generation from images as that could be a potentially dramatic increase of color generation backends on the side of gradience.

which speaking of, altho i'm currently in the development cycle of pywal16 3.8.0 i've delayed it some as i first want to consolidate the script i got into a library that when installed, pywal can detect and call for generating gtk2, gtk3, gtk4 and qt (through kvantum) styling with the current colorscheme, for which reason i want to get the library with a thin shell entry point out ASAP, once i got a release of that out i will be able to release 3.8.0 and begin porting the gradience plugins to be used on my library, along some other stuff i've been working to recolor icons and cursors from gradience presets.