[WinZETH] 2003-10-15 Hopefully I'm getting close.

[Tormentor667]

Is there already a ACS windows-gui editor implemented? And what about a "lump" editor for all the lumps in the wad?! *g*

[Hirogen2]

Kinda.

[Kappes Buur]

• I do not know, if this just happens to me:
  
  If I zoom into an area, some linedefs simply disappear off the screen.

[randomlag]

This is problem on Windows 98 and/or if a program uses 16bit. Screen coordinates are processed as 16bit numbers in Windows98 no matter what the program uses. XP does support 32bit coordinates. So what you have to do is clip the stuff yourself since zooming causes overflow.

[PSchmitz]

Chris,

Another problem to report about WINZETH. In fact, it has forced me to abandon an editor I have used in one form then another for almost 11 years (DEU/DETH/ZETH/WINZETH).

We have moved to ZDOOM's new TX structure and ZETH/WINZETH cannot handle the new changes. Primarily it's because ZETH/WINZETH cannot support flats and textures with the same name. Crash city.

[Chris]

WinZETH supports TX textures, and it properly culls out duplicate names (the former not having been tested due to lack of wads using TX textures, however). In fact, the Ultimate Doom IWAD, and I believe the Doom2 IWAD too, has some flats and textures using the same name, and depending on if you're editing flats or walls, will properly display the one that would be shown in game.

Can you give anymore info, and/or a WAD that uses TX textures to test?

[PSchmitz]

WinZETH supports TX textures, and it properly culls out duplicate names (the former not having been tested due to lack of wads using TX textures, however). In fact, the Ultimate Doom IWAD, and I believe the Doom2 IWAD too, has some flats and textures using the same name, and depending on if you're editing flats or walls, will properly display the one that would be shown in game.

Can you give anymore info, and/or a WAD that uses TX textures to test?

I am using the WINZETH version from this thread. Maybe I'm using an older version then?

All I can tell you right now is that when we add flats/textures/patches with the same name, ZETH and WINZETH crash when accessing those resources (ZETH dislays exception errors, but WINZETH bombs back to Windows with no error messages at all). In fact, we had to rename flats/textures during Daedalus development because of this name conflict issue in the ZETH code.

Let me get back to the team and see if some arrangements can be made for you to access our resource WAD. Shouldn't be a problem though.

You'll hear back from me (or Ty) very soon.

Edit: OK, some clarification is in oder.

We haven't implmented TX yet. It's just a matter of flat/texture/patch names that are the same.

What we are doing is creating a wad with a flat and a texture (and its patch) all with the same name. It's trivial to make a wad like that--I think I can even do it here at work--and then if it crashes you, you can provide that to him.

[Chris]

Are you putting the patches between P_START and P_END? When searching for patches (either to build TEXTUREx textures or just to read a regular patch), it first looks in P_START/END, then S_START/END, and finally falling back to anywhere in the WAD. If the latter happens and the flat version is the one that's first found durring a linear search, that would cause a problem (it would attempt to read the flat as a patch). I've been meaning to filter out known namespaces when doing a global search but haven't gotten around to that. Though, this is exactly why the P_ namespace exists, since the normal data formats have very little, if any, type-checking. If putting the patches between P_START/P_END doesn't work, then I'd like to see a WAD (any WAD) that reproduces the problem.

[PSchmitz]

If the latter happens and the flat version is the one that's first found durring a linear search, that would cause a problem (it would attempt to read the flat as a patch).

That be it. First patch found is a flat. And from what I read (and if I understood you correctly), you have a fix for this problem.

Chirs, thanks for the information. I'll be sure to pass it on. If the problem persists, I'll get an example WAD to you. Thanks!

[Ty Halderman]

If putting the patches between P_START/P_END doesn't work, then I'd like to see a WAD (any WAD) that reproduces the problem.

Attached. Here are some details. The Patcher resource file is:

Code: Select all

; Testing ZETH patch/texture/flat name conflicts
.option /x:c:\games\doom2\doom2.wad
[FLATS]
ZETHTEST
[TEXTURES]
ZETHTEST 64 65
* ZETHTEST 0 0 (FLATS\ZETHTEST.BMP)

When I look at the resulting ZETHTEST.WAD in XWE, here's most of what it has to say about it:

Code: Select all

TEXTURE1	TEXTURE
PNAMES		PATCHNAMES
P_START		MARKER
ZETHTEST	PATCH		4680
P_END		MARKER
F_START		MARKER
ZETHTEST	FLOOR		4096
F_END		MARKER

(and of course the texture ZETHTEST points to patch ZETHTEST in the right-hand window)

Zeth crashes when you try to view ZETHTEST in the list, or whatever page in the browser it's on. The flat ZETHTEST does not appear in the Flats-only list.

In the wad, the patch appears first, but I think the flat's being used as the patch, and of course it'll be a different format that chokes when trying to render it.

Edit: Just noticed that 64x65 size, but I redid it and still crashed, so I'll leave this as the test file.

[Chris]

Well this is odd... it's finding the patch properly, but it thinks the width is -4593. A byte dump of the patch's header produces:

Code: Select all

0f ee 6d 6d 6e ed 97 4d

Which means the image width is 0xEE0F. That comes out to either (signed)-4593 or (unsigned)60943.

You sure this is a valid Doom RLE patch?

[Ty Halderman]

You sure this is a valid Doom RLE patch?

Not if it's trying to interpret the FLAT as a patch, probably. Could that be the issue?

[Chris]

Okay, found the problem. ZETHTEST was originally found in between P_START/P_END, but it ended up getting replaced with the one in between F_START/F_END. Didn't realize ZETH didn't keep duplicate named lumps. Though now it's only letting me pick the flat version. Need to look into that.

[PSchmitz]

Okay, found the problem. ZETHTEST was originally found in between P_START/P_END, but it ended up getting replaced with the one in between F_START/F_END. Didn't realize ZETH didn't keep duplicate named lumps. Though now it's only letting me pick the flat version. Need to look into that.

If you need someone to test your fixes, I'm the guy!

[PSchmitz]

FYGI, I think I fixed the timing problem, and the decorate items might be working (somewhat), but I need a wad that has some proper decorate objects to test it with.

Sorry for not offering this suggestion earlier:

What about Daedalus? It has about every example available for Decorate objects (at least the ones supported up to cab60 IIRC).

http://www.doomworld.com/idgames/index. ... /daedalus/