GNOME Bugzilla – Bug 779976
Patches from Endless
Last modified: 2017-03-13 21:01:53 UTC
Created attachment 347824 [details] [review]
ide: don't try to steal stream data if it failed to close
We have some patches at the Endless for of the GNOME Builder, so we would like to upstream them, instead of maintaining them ourselves.
All of them actually seem to be written by Cosimo, so I'm CCing him here.
Created attachment 347825 [details] [review]
gca: don't return an error when not available
Created attachment 347827 [details] [review]
flatpak: don't call json_node_* methods on NULL nodes
Created attachment 347828 [details] [review]
ide-subprocess: instrument another function
Created attachment 347829 [details] [review]
ide-subprocess: count outstanding operations beforehand
Thanks a lot for submitting those and for trying to keep things upstream, we appreciate it!
What were those patches generated with? Are they kept in git at Endless? “seem to be written by Cosimo” is not really reassuring, you should have a way to track precisely that sort of things.
In general, Endless has the gnome-builder fork here - https://github.com/endlessm/gnome-builder, the PR with these patches is here - https://github.com/endlessm/gnome-builder/pull/6, you should find those patches in the very recent history, on a first page even - https://github.com/endlessm/gnome-builder/commits/master
I recently did a rebase of the fork on the current master upstream branch, so all the not-yet-upstreamed patched surfaced. The result is here - https://github.com/kinvolk/gnome-builder/commits/krnowak/endless-master-rebase, it's not yet pushed to the fork, but I was asked to upstream some of them. I didn't upstream the first one as it certainly wouldn't be accepted and the last one which is IMO wrong.
Given your patches come from git, you should use `git format-patch` to generate them. That will ensure they contain the necessary metadata (author, commit they apply on).
But I did generate them with git format-patch…
Nevermind, I followed the wrong link and got the diff without the header. Thanks again for your submission!
BTW. The patches were generated from this branch, which is recent upstream master + those 5 patches:
Thanks, Krzesimir! I also confirm that I'm the original author of these patches :-)
Excellent, everything looks great!
Attachment 347824 [details] pushed as 0a9fb79 - ide: don't try to steal stream data if it failed to close
Attachment 347825 [details] pushed as fd96d48 - gca: don't return an error when not available
Attachment 347827 [details] pushed as f0c4ba7 - flatpak: don't call json_node_* methods on NULL nodes
Attachment 347828 [details] pushed as 6d5d991 - ide-subprocess: instrument another function
Attachment 347829 [details] pushed as 5e25df7 - ide-subprocess: count outstanding operations beforehand