XSI ZETools

[ANDEWEGET]

2.4+ in the 2.x releases will work(no 3.0+)if XSI recognizes it. I only got 2.4 working on Win7 64bit and the Mod Tool 7.5 though...
I got the 2.4 files from Blurs Installation page, there are 2.6 files, too, I couldn't get them working though. That might have been due to those downloads being for 64bit, you could try finding the corresponding 32bit versions of python 2.6.6 and pywin (pywin32-216.win-amd64-py2.6 or so). I don't have the time to test them right now.
I'd really like to use 2.6, it has some nice new features.

I've done the most/most difficult part of adding weight supports again(I believe), so it shouldn't take too much time until it's finished.

Does anyone have any wishes for functions or features BTW?

[Marth8880]

Sounds fantastic!, but the feature I'd love to see the most is polygon material flags, rather than object edit flags.

Also, I have a question that may have been answered before: If I have an object with one material flag with, for instance, glow, and then another object with a different material flag such as transparency, and merge the two objects (as well as merge their materials and UVs of course), will the newly-formed object retain those material flags?

[tirpider]

re: materials.

From what I hve noticed, each object (segm) gets one material, with whatever flags and features (transperancy, bumps, glow, ect.) set for all that objects polys and verts.

You can have many objects (segm's) per model, though, and each can have it's own material.

I don't know if ZETools keeps all objects (segm's) in the same MODL or seperates them, but the result is the same as above.

One object, one material.

[Marth8880]

I know that's how it works right now, I use the ZE Tools for almost all of my models now. I'm talking about polygon-specific material flags in future releases.

[lucasfart]

I also have the 64bit version of windows 7 ultimate, and I've just confirmed that the 32 bit of Python 2.6.6 works with your addon! Its worth noting that the 64bit version, as you stated above, doesn't seem to work. I haven't actually tried using your tools with 2.6.6 yet, but the menu itself appears just fine in the modtools.

For ease of download here are the 32 bit downloads of Python 2.6.6 and pywin32 for python 2.6.6:
pywin32
Python 2.6.6

[tirpider]

re: polygon-specific material flags
I hope I'm not talking out of turn here. I'm comming from a msh file structure perspective, as I have just enough time in XSI to know it wont work for me.

There is no place to store per-polygon settings in the msh. (Haven't seen a limit on the number of materials or objects in a MODL chunk, though.) It's a format limitation and not a tool limit.

The vertecies, triangles and poly's associated with an object are stored in a single SEGM chunk that calls a single material (via MATI).

To get different effects, it would need to be defined in a material, then, the vertecies, triangles, and polygons would need to be in their own SEGM. (doesn't have to be a seperate MODL)

It would be some fancy coding to do that programaticly. (Sorting all the proper verts, triangles, polygons, uv's, weights, and normals, into a seperate object at export.)

It would be easier to just make a new object, because that's what a program would have to do to get it into a msh.

Making this feature would be groovy, but daunting task.
(If anyone can do it, it would be ANDEWEGET.)

[Marth8880]

The Pandemic Tools do polygon material flags perfectly fine.

[AceMastermind]

What the Pandemic Tools produce:

• 1 object with 1 material and no flags = 1 MATD and 1 MODL with 1 SEGM
  1 object with 1 material and 1 flag = 2 MATD and 1 MODL with 2 SEGM
  1 object with 1 material and 2 flags = 3 MATD and 1 MODL with 3 SEGM

[tirpider]

Cool.
So it does write new SEGMs out for multiple flags.
(I didn't know the Pandemic exporter handled it that way.)

Then it would be possible to implement it in an exporter.
The trick would be coding it.

I'll just crawl back under my spreadsheets

[ANDEWEGET]

@lucasfart:
Thank you very much for testing! I'll add those download links to the FAQ.

@Marth:
Right now the flags are stored in the material(you can find them as properties of the MSHMaterials null). So they definitely get preserved. Which material an object uses is stored in a custom property of the object. I dunno which property or if both will be preserved. The exporter will ignore the second though, because it most likely will be renamed.

@tirpider, Marth, Ace:
SEGMs are a bit like clusters in XSI, they store parts of the geometry which together defines the geometry of the whole object(MODL). They appear if data is added to the object which doesn't apply to the whole object(like materials, vertex colors).
Coding that from scratch would most likely be to difficult for me but XSI is awesome and already includes a function(which is sadly only accessable from C++)which returns polygon indices per material. This still requires getting all the vertices, uvs, weights etc for the polies but it makes it a lot easier. I'd have to switch back to XSI materials and find a nice way to add the flags to those materals...
I suppose I'll try my hand at this either after I finish weights or cloth. Thanks for the nice comments BTW.

EDIT:
Improved the check selection function a bit:

Hidden/Spoiler:

Not very nice visually but it does it's job(checking your selection for problems which might break the .msh).
I added Close buttons to every dialog which deletes the property so your Scene_Root doesn't get cluttered.

[Cerfon Rournes]

EDIT:
Improved the check selection function a bit:

Hidden/Spoiler:

Not very nice visually but it does it's job(checking your selection for problems which might break the .msh).
I added Close buttons to every dialog which deletes the property so your Scene_Root doesn't get cluttered.

That will be useful. Nice update, keep it up!

[ANDEWEGET]

Thank you, I didn't get to finish weights though, but
[I'm] getting closer naaooow...
I spend quite some time today cleaning up code and trying to make it work with as few lines of code as possible. (I believe) I just need to structure the results I get from the C++ plugin then weights should work but who knows...

I'm having problems with the values I get from C++ for the weights right now, that's why it's taking so long right now. I suppose I'll go back to just doing the most basic in C++ and all the other stuff in python.
Could all the people who have python 2.4 right now try and upgrade to 2.6? I don't see the point in developing for 2.4 if 2.6 works, too. 2.6 is just way better for me(and for you, too, then; links are in the FAQ).

[Cerfon Rournes]

I'm having problems with the values I get from C++ for the weights right now, that's why it's taking so long right now. I suppose I'll go back to just doing the most basic in C++ and all the other stuff in python.

Take all the time you need.
@Python: I already have 2.6, I think.

[willinator]

Is it possible to export a walker that isn't one of the templates? (Example, exporting a walker using the spider-droid anims and skeleton?)

[ANDEWEGET]

Right now you can't, but when I finally find the time to add in weight support again it should be possible to export any skeleton.
I've stumbled upon some weird problems which need more code and time.

I have already added in code which needs python 2.6, so please make sure you update to py2.6+.
If you don't know if you've got the right version, check in your installed software in the control panel.

[willinator]

Another quick question, do we only have to check this box to apply specular, or do we have to use the pandemic flag?

Hidden/Spoiler:

[CressAlbane]

Sweet tool, ANDE. It really helped me out to have the "How-to-install" tut on the site - I was having trouble getting Python working. Thanks!

[ANDEWEGET]

@CressAlbane: Thanks!
@willinator: Pandemics exporter enables the box automatically if you choose a specmap/glossmap rendertype. My exporter doesnt afaik. So better do it. I know for sure that normal map + spec works with rendertype + specular box enabled(you can see it in pictures I posted).

Weights seem to be working now. C++ returned wrong values(unless I read something wrong)which broke the weight export. However, after exporting the unit template I got a

Code: Select all

WARNING[PC_modelmunge msh\abc.msh]:SplitSkinnedSegments: WARNING: out of space in splitting catalog!

warning. Anyone experienced that before? It made my game lag a lot(I assume the warning is the cause 'cause it ran smoothly the session before).
I still have to write the 32bit version of the C++ plugin, but that shouldn't take too long.

Hidden/Spoiler:

(Exported via the mod tool.)

Edit: 32bit version seems to be working. I've cleaned up and added some code, too. The biggest change you'll see is this:

Hidden/Spoiler:

It's a new export report(:D)which shows some additional infos(the time certain actions take individually). Before this was just printed to the console(which slowed down the export[after it finished] because that were quite a lot of lines and printing takes time). I don't suppose it'll be of much use for most of you, but this way I can find bottlenecks in the code better and you can report better because it should still pop up that report even if the export failed. The entries will stop at the point where the problem occurred.

[Cerfon Rournes]

Hidden/Spoiler:

[quote="ANDEWEGET"]
Weights seem to be working now.

It's a new export report(:D)which shows some additional infos(the time certain actions take individually).[/quote]

(Exported via the mod tool.)

Edit: 32bit version seems to be working. I've cleaned up and added some code, too. The biggest change you'll see is this:
[hide]

Hidden/Spoiler:

Great work ANDE.

[ANDEWEGET]

Changelog:
0.9.08.02:

• Added in weight support.
  Added multiple reports.
  Cleaned up code.
  Changed the folder structure a bit.
  Added in more and better comments in the code.
  Added multiple smaller functions.
  Improved performance.

Version is up at the first page. Please remove the XSIZETools folder completely from the Addons directory before installing this update. Remember to update to py 2.6+ if you haven't yet.

I didn't have the time to test much, so there could still be a lot of errors I ddin't get. Please report them as usual, if you've got any feedback apart from that, feel free to post it!