We have tons of issues open on github. They are sorted with labels which describe what the issue is about (UI, federation, browser specific, etc).
BUT we have no way to sort issues according to their importance. Some issue are open because we think it could be nice to work on that point (for example, improve the atom feed) but they are absolutely not an emergency. At the opposite, youtube embedding is broken in Firefox, this affects thousands of users everyday, and it’s open for months without any activity.
I think we should clean the labels, removing the ones we do not use anymore, and adding some new defining three levels: “Urgent” (or ASAP), “Needed” and “could be nice to have” (I need native speakers to fine better words here, but I think you got the idea.
Note: This discussion was imported from Loomio. Click here to view the original discussion.
Which labels do you consider not used anymore?
We can add classical priority labels, “Urgent”, “High”, “Medium”, “Low” but I don’t think we’d see an effect on development.
PS: The issue you mentioned is basically an upstream bug, in both Youtube and Firefox. All we would do to fix it would just be a hack.
I created a list a while back of all issues open but untouched for more than 6 months. It’s not prioritised at all, but should help if anyone feels up to the task of doing some housekeeping, bumping issues which are still important and need attention and closing or adding to the discussion on other ones which may not be needed any more - or not at this time.
Rather than list them here, which would take up a lot of space, here’s a link to a post I made with them: https://pod.orkz.net/posts/581381
(All the ones concerning the old single-post view have now been closed.)
Well, I guess we could probably remove some existing labels. “Quick fix” almost always comes with “Newcomer”. “Security issue” never stays longer because it’s immediately fixed. “needs more info” is not really useful either, we should simply ask the reporter to be more precise.
I guess we should think the other way: labels are filters. So, on which criteria do we want to filter issue. Is “postgreSQL” a criteria really needed, for example?
Apart from that, I’m all in favour of classical priority labels. I’d love to help with that, did Github make advancement in permission separation?
I disagree, I find all of them useful.
Quickfix usually still does involve quite deep knowledge of codebase, they’re rarely newcomer friendly. OTOH a newcomer task is seldom quickly done.
Postgres is a very valuable filter for the same reason backbone is a valuable filter, it makes it easy for experts in those areas to find things to contribute.
Security is a useful filter in case you’re unsure if the issue you’re thinking about might be fixed already and you just rediscovered it in an older version of the code.
“needs more info” can provide a elegant way for non developers to contribute, finding a clear way to reproduce an issue often points out to a developer where the bug is quite easily. Reproducing is probably the hardest part in fixing a bug. So we could stamp that out a bit more and write some guidelines and tips on how one could help us out here, similar to the newcomer initiative.
@jonnehass I think adding priority labels could improve development process. Maybe not everybody would use them, but some people would. Of course, it would be nice to list “Urgent” before “Normal” but we have to use what we have.