OpenMW 0.46.0

Anything related to PR, release planning and any other non-technical idea how to move the project forward should be discussed here.
User avatar
Capostrophic
Posts: 779
Joined: 22 Feb 2016, 20:32

Re: OpenMW 0.46.0

Post by Capostrophic » 12 Apr 2020, 23:14

Status update:
1) AnyOldName3 knows the reason for the alpha channel reading issue - basically the specific TGA file Is malformed - but it's not decided what to do with it; technically OSG's interpretation of the improperly set up alpha data as set up by AON3 is just as valid as Morrowind's.
2) Shadow transparency fix works but needs more testing
3) The NiCollisionSwitch optimizer issue fix is fixed in master, needs cherry-picking to 0.46
4) There's a door unlocking regression (the fix is up)
5) There's another radial fog regression with water, currently unclear how to solve it and whether to solve it for the release at all instead of just making underwater fog non-radial for now. There's something that can slightly improve it by making it not abusable, but it won't look good anyway.
6) There's a small issue with the console. The reason is known but there might be caveats with the theoretical solution.

nrh
Posts: 4
Joined: 29 Mar 2019, 03:02

Re: OpenMW 0.46.0

Post by nrh » 13 Apr 2020, 05:58

Alphas always give me trouble every game I go. Super pesky!
Nice job on shadow transparency. I'm stoked to see it.
Skip the underwater radial fog for 0.46. It's 0.47 or 0.48 material.
Any meaty details on the console problem? Is it syntax?
Thanks a lot. The predictable release curve was severely altered by 0.46.

User avatar
AnyOldName3
Posts: 2096
Joined: 26 Nov 2015, 03:25

Re: OpenMW 0.46.0

Post by AnyOldName3 » 13 Apr 2020, 11:48

There's nothing exciting about the shadow transparency fix. Sometimes we were using the alpha channel of textures and/or vertex colours to make things not cast shadows when they were actually opaque. Now, opaque things always cast shadows properly.
AnyOldName3, Master of Shadows

User avatar
Capostrophic
Posts: 779
Joined: 22 Feb 2016, 20:32

Re: OpenMW 0.46.0

Post by Capostrophic » 14 Apr 2020, 15:12

Status update:
1) The fix for the TGA issue has been submitted to OpenSceneGraph. Once it's there (or even if it's not or if it takes a while), I assume we can update OSGoS TGA readerwriter and add the necessary option string to OpenMW (0.46 branch and master) with little effort (I believe it's not even necessary to put it behind a preprocessor check?).
2) Nothing changed.
3) NiCollisionSwitch fix cherry-picked to 0.46.
4) Door unlocking regression fix cherry-picked to 0.46.
5) Nothing changed but it's not a big deal.
6) The issue with the console is that since it belongs to generic GUI window layer but can be present while an exclusive GUI mode is enabled it has lower priority than book view. I've submitted something that makes book view have lower priority than other GUI windows but maybe it could break something. Otherwise it's undesirable for the console to belong to a dedicated layer while the book view must have a dedicated layer.

User avatar
Capostrophic
Posts: 779
Joined: 22 Feb 2016, 20:32

Re: OpenMW 0.46.0

Post by Capostrophic » 20 Apr 2020, 21:26

Status update:
1) Still the same.
2) Shadow alpha testing fix needs some more time (will need alpha threshold handling revamped).
3) Done.
4) Done.
5) Underwater fog won't be radial in 0.46.0, but it's not broken anymore.
6) That issue has been fixed, but there seems to be another minor console issue (it gets overlapped by the inventory after a MenuTest debug instruction call)

There has been a bunch of other regression fixes in the mean time.

Release video is "finished" but Atahualpa wanted to give johnnyhostile some feedback on it.

User avatar
psi29a
Posts: 5031
Joined: 29 Sep 2011, 10:13
Location: Belgium
Gitlab profile: https://gitlab.com/psi29a/
Contact:

Re: OpenMW 0.46.0

Post by psi29a » 20 Apr 2020, 23:32

Getting there. :)

User avatar
Greendogo
Posts: 1460
Joined: 26 Aug 2011, 02:04

Re: OpenMW 0.46.0

Post by Greendogo » 21 Apr 2020, 08:33

excite.

Edit:
That fix list is gargantuan.

User avatar
sjek
Posts: 441
Joined: 22 Nov 2014, 10:51

Re: OpenMW 0.46.0

Post by sjek » 21 Apr 2020, 08:45

so. to keep things short :)

There's has been an effort to test issues in 0.46 changelog via google docs.
https://docs.google.com/document/d/1x-x ... sp=sharing

Basically almost third (78) of them is more or less tested and
have marked with a little green background the ones with test cases written.
We started this before capo released his changelog so it's based on changelog on github.
Document is on comment and suggestion rights.

This is an ongoing project and it takes easily days to write all test cases so putting this here.
but anyhow, this is not gonna be ready for release xD

Feeback would be welcome.
This effort's own server is
https://discord.gg/buh6BKc "some casual openmw testing server"
"life is crazy"
"craziness has beauty which only crazies understand" some movie clip in the head.
https://wiki.openmw.org/index.php?title=Testing

User avatar
raevol
Posts: 3093
Joined: 07 Aug 2011, 01:12
Location: Caldera

Re: OpenMW 0.46.0

Post by raevol » 21 Apr 2020, 17:08

sjek wrote:
21 Apr 2020, 08:45
There's has been an effort to test issues in 0.46 changelog via google docs.
https://docs.google.com/document/d/1x-x ... sp=sharing

Basically almost third (78) of them is more or less tested and
have marked with a little green background the ones with test cases written.
We started this before capo released his changelog so it's based on changelog on github.
Document is on comment and suggestion rights.
This is awesome. This may need to be spun out into a different thread, but is there a web tool we could use to manage this? Like a Gitlab issue tracker but for test cases? The only one I've ever used was when I was in the industry and was proprietary, and was like 10 years ago.

User avatar
sjek
Posts: 441
Joined: 22 Nov 2014, 10:51

Re: OpenMW 0.46.0

Post by sjek » 22 Apr 2020, 14:54

i think there would be lot of overhead making this integral part of the development cycle.
even when in same workplace could, as files need to be made and test cases written continuously.
it would be good for documenting purposes but not much help if nobody actively making said document.
that's why i think this is better as unofficial effort.

on the web tools.
dropbox and google both has file hosting services, dropbox embedded file sharing on it's document
and google docs having much better editing possibilities.

then again there is privacy, security and slick reporting interface reasons to use issue tracking tools like gitlab, jira. bugzilla etc.
but if someone offers hosting, document and issue tracking services it's probaply paid one.
if not meant to run on own server, i guess or has a people limit.
thought wikipedia got extensive list of issue tracking systems and project management software.

but i think the document hosting is needed to give overall look more easily.
in issue tracking services it goes to many paths to specified information so it's harder to read and maintain.
with test cases, what solved, what's still an issue, what needs research, etc.
there is a point when issue tracking becomes simply better for complexity reasons.
and the number of simultaniously solvable issues is important.
2 cents.

but yeah. off topic.
"life is crazy"
"craziness has beauty which only crazies understand" some movie clip in the head.
https://wiki.openmw.org/index.php?title=Testing

Post Reply