Özkan Sezer

> no lightmap with TEX_MISSING flag Well, I wasn't expecting any solutions to this thing at all to begin with. I'm still not sure that your patch whether your patch...

@ericwa, @andrei-drexler, @Novum ?

P.S.: Have you bisected this?

It's possible that it was introduced between 0.93.1 and 0.94.x when re-release support patches went in.

Vispatch'ed id1 looks OK to me? (i686 linux, qs from today's git, both with `r_wateralpha 0.4`) ID1 1.06 + vispatch:  ID1 - rerelease: 

Build-tested _only_, no issues on my old setup. Attached patch to regenerate configure: [configure-regenerate.txt](https://github.com/libsdl-org/SDL/files/6702936/configure-regenerate.txt)

Here is the working and clean build fix against your latest commit: ```diff diff --git a/src/physfs_archiver_tar.c b/src/physfs_archiver_tar.c index c0d5f44..41a3920 100644 --- a/src/physfs_archiver_tar.c +++ b/src/physfs_archiver_tar.c @@ -14,10 +14,14 @@ static bool...

The mode procedures aren't used: I'm guessing that commenting them out will make windows builds succeed.

The following patch makes a clean-up of the whole thing: ```diff diff --git a/src/physfs_tar.h b/src/physfs_tar.h index e15bc68..aca7e40 100644 --- a/src/physfs_tar.h +++ b/src/physfs_tar.h @@ -1,15 +1,7 @@ #ifndef _INCLUDE_PHYSFS_TAR_H_ #define _INCLUDE_PHYSFS_TAR_H_...

> @sezero: Thanks for the changes and hints I've integrated them into the pull request. Thanks: all my change requests are addressed. It's in @icculus hands now.