You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Want to make a general issue to publicly list some plans and thoughts we/I have regarding the future of userstyles. The following list will be updated somewhat frequently as I get through the list :)
Utilize Deno 2 (fully releasing later this month!) for formatting of CSS and YAML files as well, possibly others - basically remove Prettier from the equation.
Scripts-related
Fix codeowner perms by fetching and specifying userstyles staff team members individually (request the individuals not the team so that all have to approve aka me and bell).
Automatically ping maintainers on issues for their userstyles.
Some issues with performance and linting not being completed for certain conditions that don't necessarily need to exist (e.g. you can't lint the CSS of a userstyle if you haven't added the metadata for it iirc? or similar).
Restructure the script execution to run in parallel, and maybe use a more efficient way of compiling LESS (sass has this neat embedded Dart binary package for super fast compilation, and you can create a compiler instance to reuse - LESS might have similar?).
Documentation
Document the CSS color filter generator technique for theming single color images #1089: CSS filters are somewhat used already in the codebase, would like to standardize, document, and cleanup existing implementations. My thinking is we would have an additional "palette" block for filters as to not bloat all of the userstyles, and this would mean we can make sure all the generated filters are as precise as possible.
Additionally add some documentation on PR reviewing. What to look for and where, tips and tricks for finding unthemed areas. Reasoning behind "websites fully themed in init PR" could be explained better. Thoughts on code management and organization could also be mentioned here.
Establish criteria for userstyle acceptance: some usage metrics, or some amount of pages that benefit significantly from having a userstyle (I don't think we really want to theme single page websites like home pages).
Simplify the template and userstyles by making some small changes related to accentColor/accent/accent-color.
Can also simplify prefers-color-scheme media queries with sub selectors by putting the media query inside one single selector (was not supported a few months back? maybe?): https://codepen.io/uncenter/pen/ExqbRKw.
Consolidation
Bootstrap sites: we have quite a few of these already, I believe 2+ current PRs also theme Bootstrap websites. Would like to look into creating a port for it so we can reuse the code!
Ace Editor: Split it it's own port? I believe some? one? userstyle themes it and it might be useful by itself.
Tailwind: TailwindCSS v4 is quickly approaching us and I think it means that sites using Tailwind that upgrade will use CSS variables! 🎉 Hopefully this will work as expected and be on by default, making our lives easier.
Want to make a general issue to publicly list some plans and thoughts we/I have regarding the future of userstyles. The following list will be updated somewhat frequently as I get through the list :)
Short term
Medium term
Scripts-related
Documentation
Long term
accentColor
/accent
/accent-color
.Consolidation
Misc
The text was updated successfully, but these errors were encountered: