Merge of PR 1674: a potential problem?
Posted: 13 May 2018, 14:13
I am on arch linux, which is a rolling release. Since OSG got a new version, had to compile openmw all over again -- I am not using the stable package in the arch repos, but the openmw-git package that pulls the last version from the openmw repo; the last compilation I did was 2 or 3 weeks ago. On testing some mods, I found that some textures were not being replaced. At first I though there was some problem with my load order, but after commenting out some data paths and mods, I concluded that the problem could not be found there. Then I went to the list of PR's merged and noticed https://github.com/OpenMW/openmw/pull/1674 and a little light bulb went on in my head. I have the habit of converting all tga textures in mods to dds because openmw cannot read some tga formats -- I think this has been fixed by now, but the habit remains, and more importantly, I never did this for the *original* morrowind Textures folder. So went ahead and did the conversion, and voila, everything was working fine. Or not. Because then I noticed that the icon replacers were also not kicking in. Now, I do *not* do a tga conversion because there was never any need, but I suspected the problem was the same, and so the solution should be the same, and sure enough, it was.
This of course raises a problem: when the next 0.44 stable version is out, texture and icon replacers will suddenly stop working. Now, if I diagnosed the problem correctly, that the source of the problem is in the merging of https://github.com/OpenMW/openmw/pull/1674, then we have what I would consider a regression. The other option of course, is that I misdiagnosed the problem, but then could someone explain to me what I got wrong?
This of course raises a problem: when the next 0.44 stable version is out, texture and icon replacers will suddenly stop working. Now, if I diagnosed the problem correctly, that the source of the problem is in the merging of https://github.com/OpenMW/openmw/pull/1674, then we have what I would consider a regression. The other option of course, is that I misdiagnosed the problem, but then could someone explain to me what I got wrong?