Open ewlsh opened 4 years ago
This revamp takes inspiration from gnome-shell/!404 and will hopefully serve as a testing ground to actually get this feature set upstreamed.
I am working on a "wallpaper tester" that can offer stats for any wallpaper inputted, this should allow me to compile a decent set of parameters to build out the wallpaper detection.
I've ported the app-tweaks and custom theme removal to the current master branch in #126 and merged.
The rest of this will slowly migrate into master too :)
I love DPT but it was admittedly my first major Gnome Shell project and was born during an era of much more significant shell-upheaval. As a result, there are numerous features which have been added over the years which are challenging to support and which drastically increase the complexity of the project. Many don't even work on the latest Shell versions.
This is my plan to fix-up my favorite extension.
Work Done:
1. No Custom Theme Support
In the past, DPT has attempted to "blend in" to any custom theme with varying degrees of success (the "remove excess styling" option exists because shell themes are notoriously difficult and varied).
In future versions it will no longer attempt to blend in to custom themes and users will have to manually adjust the style preferences if they wish the panel to "blend"
2. Wallpaper-based Styling
Instead of relying on themes to determine panel coloring by default, DPT will attempt to use the user's wallpaper to "guess" its appearance.
More users change their wallpapers at this point than change their themes, so I think this change is overdo. See #48.
3. No per-app settings.
This added a lot of complexity to the extension and became difficult to maintain. Instead, if DPT is unsure what to do with an app it will default to a solid panel.
4. Reasonable customization.
I support and love the ability to customize things, but the number of settings and preferences this extension offers is simply too much to support adequately.
Knowing this, I'm working on reducing the number of preferences offered and opting for sane defaults where supporting detailed preferences is challenging or adds unreasonable complexity.
This PR will serve as the staging ground for this revamp.
Expect things to be:
faster
,less buggy
,more stream-lined
, and generally much-improved.