Fabio Manganiello on Nostr: a good example is Unite. A very popular extension to customize the windows behaviour, ...
a good example is Unite. A very popular extension to customize the windows behaviour, mentioned all over the Web, which is basically broken - both because it hasn't been updated in three months, and because it relies on X.Org-specific tooling.
Pano clipboard manager is another one. Sound Input & Output Device Chooser too. And Multi Monitors Add-On. In general, if you just go to [extensions.gnome.org](http://extensions.gnome.org ) and try to install some random extensions from the first two pages of results, there's literally a >50% chance that the extension won't work on the latest version of Gnome Shell.
I understand that the "power" of Gnome extensions come from customizing the source code of Gnome itself. But I wonder if a better extensions model is possible. The current model basically requires developers to be very proactive and release a new version of their etensions as soon as a new version of the Gnome Shell is out, even if technically nothing has changed either on the Gnome side nor the extension side.
If the Gnome Shell has a Web-inspired design, I wonder if it wasn't easier for extensions to just override (without patching) the current CSS style, or for Gnome itself to emit events that extensions can listen to in order to modify the WM just like you would modify a DOM.
Probably it wouldn't be as powerful as patching the Gnome code itself, but it'd definitely be less brittle, it would require less maintenance efforts from the developers side, and it would probably results in less extensions being broken if a developer was on vacation when a new Gnome version came out.
Pano clipboard manager is another one. Sound Input & Output Device Chooser too. And Multi Monitors Add-On. In general, if you just go to [extensions.gnome.org](http://extensions.gnome.org ) and try to install some random extensions from the first two pages of results, there's literally a >50% chance that the extension won't work on the latest version of Gnome Shell.
I understand that the "power" of Gnome extensions come from customizing the source code of Gnome itself. But I wonder if a better extensions model is possible. The current model basically requires developers to be very proactive and release a new version of their etensions as soon as a new version of the Gnome Shell is out, even if technically nothing has changed either on the Gnome side nor the extension side.
If the Gnome Shell has a Web-inspired design, I wonder if it wasn't easier for extensions to just override (without patching) the current CSS style, or for Gnome itself to emit events that extensions can listen to in order to modify the WM just like you would modify a DOM.
Probably it wouldn't be as powerful as patching the Gnome code itself, but it'd definitely be less brittle, it would require less maintenance efforts from the developers side, and it would probably results in less extensions being broken if a developer was on vacation when a new Gnome version came out.