2 ms·
Is there an example you have in mind where they were objectively wrong and did not admit it? In the examples I've seen their decisions generally seem reasonabl
by asoneth 4y ago
Is there an example you have in mind where they were objectively wrong and did not admit it?
In the examples I've seen their decisions generally seem reasonable and well-considered, even in cases where I disagreed with the outcome. The most I can say is that it might benefit some contributors to consider tone and messaging a little.
With respect to closing tickets, unless someone is paying for the work I generally close feature requests that aren't compatible with a project's vision. Focusing is about saying "no" and building every feature every user requests eventually results in a bloated, unmaintainable project.
Finally, with respect to downstream vendors and projects that incorporate pieces of the GNOME shell into their own software the GNOME team has made it pretty clear that (for better or worse) their focus is their own desktop environment. Asking contributors if they're willing to do additional work to support outside projects is reasonable, but demanding that they do so is not a good look.
- lokedhs 4y agoI can think of one specific case. At some point, it all of a sudden wasn't possibly to disable cursor blinking in gnome-terminal anymore. It's fixed now, but it took years and absolutely no admission that perhaps they didn't handle it very well. The entire story is outlined here. https://geoff.greer.fm/2016/08/26/gnome-terminal-cursor-blinking-saga/ https://geoff.greer.fm/2016/08/26/gnome-terminal-cursor-blin...