Originally posted by nomadewolf
View Post
Another big issue is with the way wine-staging works. Too many patches seem to be hanging around in wine-staging for too long for important things. eg. Last I checked (a few months back) DRM platforms like Origin and Uplay required a bunch of patches from wine-staging to work correctly - but it wasn't clear which specifically. When testing a bug reported against the main (non-staging) Wine, testing with all the staging patches isn't all that helpful - you want the minimum required patches to do the job and you want to report precisely what was tested. Finding which patches are required to get Origin or whatever working can be both very difficult, and very time consuming as you likely need to compile many times (adding patch sets along the way) before you get it working. So if you want to test a game in Origin, you potentially have a lot of work involved before you can even start. This seems to have turned a lot of testers away (myself included), which means no bug reports to close.
And Wine-staging makes regression testing really hard! All staging patches seem to be rebased against upstream git tags, so what happens when you know 2.15-staging was working and 2.16-staging wasn't? You can't just run a git bisect and start testing from a commit in the middle, because there is no middle. Chances are, neither patches from 2.15-staging or 2.16-staging will apply cleanly, which results in a lot of burden on the tester which didn't previously exist. It used to be that staging could be ignored, but as said above, so many critical patches to get DRM gaming platforms working seem to hang around in staging forever so ignoring wine-staging is no longer an option - at least for the time being.
Comment