07:51 adavy: q4a: ah and I must precise you'll need to add the test to their visual.c test, and not submit NineTest. they will also want the wine_todo flag in front of the tests wine fails. That's how they distinguish regressions from improvements needed
07:55 q4a: adavy: I'm understand about `visual.c` - already added. About `wine_todo` - I thank that they will said that. It is much more difficult for me to clearly describe this multifunctional test, and even more difficult to explain why they need this test?
07:56 q4a: The problem with fog was found in Nine, but it works fine in Wine.
07:56 q4a: s/thank/think/
07:58 adavy: does it really ?
07:58 adavy: wined3d passes all the tests ?
07:59 adavy: I can write the comments needed if you need
07:59 q4a: I didn't use my test in Wine. I was talking about original fog promblem with trees.
08:00 adavy: they might not have had the original fog problem, still the fog computation was incorrect
08:00 adavy: so the trees were not exactly how they should have been, even though it looked correct
08:00 adavy: and there must be other games where there should be fog and it is incorrect but people don't realize
08:02 q4a: I think that it can be rare example that people don't use in real games.
08:04 q4a: Anyway, I'll run my test in wine and will create branch in Wine repo, but I will be grateful if you check it for nonsense / fairly complete description.
08:12 adavy: wine tests have a number of things tested that are not used in real games
08:12 adavy: the also it's also about accuracy. Even though using shader's z look ok, if you should use w, well you should use w
08:12 adavy: wine devs care about that
08:38 q4a: https://pastebin.ubuntu.com/p/WFd2TCMpJx/ - tested in Wine 8.0 on intel gpu: 33 fails in 72 cases.
08:39 q4a: AMD gpu may work better, but I'll test it later.
10:42 q4a: adavy: I'm not sure that I want to explicitly state results of the tests (in which cases and what exactly affects the fog). At least in the commit message. I hope that they will be able to analyze result and draw their own conclusions. But if there are questions, they can be discussed in the request. Anyway, I will be grateful if you check my commit message:
10:42 q4a: https://gitlab.winehq.org/q4a/wine/-/commit/79a1410e8ce186447e208381bfa42ea93668e820
10:43 q4a: * adavy: I'm not sure that I want to explicitly state results of the tests (in which cases and what exactly affects the fog). At least in the commit message. I hope that they will be able to analyze result and draw their own conclusions. But if they will ask some questions, we can discuss in merge request. Anyway, I will be grateful if you check my commit message:
10:43 q4a: https://gitlab.winehq.org/q4a/wine/-/commit/79a1410e8ce186447e208381bfa42ea93668e820
11:07 adavy: wine devs don't write any good commit message so I assume they don't read them
11:07 adavy: inspire from what is written for the other tests
11:16 q4a: Ok, let's see what wine devs will say: https://gitlab.winehq.org/wine/wine/-/merge_requests/2657
11:45 adavy: +1
14:05 adavy: q4a: after that do you plan to still contribute to nine ?
14:19 q4a: adavy: I need to write curriculum vitae and find out about internship in Collabora, Igalia or somewhere else. I'm mostly interested in Nine and Vulkan support in Panfrost. There is tasks about Nine like: https://gitlab.freedesktop.org/mesa/mesa/-/issues/8315
14:19 q4a: + Vulkan support already in progress.
14:19 q4a: Anyway I wand to contribute more to mesa and Nine is best thing for now.
14:37 DavidHeidelberg[m]: q4a: write CV, if you're interested in panfrost, Collabora could be good choice :)
14:37 DavidHeidelberg[m]: Since Alyssa leaving :D
14:41 q4a: David Heidelberg: may be you know - internship in Collabora is for students only?
15:02 DavidHeidelberg[m]: Send me the CV :) I'll do quick review and if you wsnt fwd to our HR :)