XSI ZETools

[tirpider]

I'm thinking the one animation per msh is a limitation of the munge process and the way SWBF calls them in-game.

All the munge does is put everything into one file anyway, but it expects them to be prepared a specific way so it knows how to index them when it squishes them together.

Everything I have read about animations indicates SWBF relies heavily on the actual names of the animation msh files. If the filename convention (banks and specific names and such) breaks down or isn't followed, the animations just don't get used, even if the munger passes them with no error.

To get all-in-one animation files, we could start digging up the format of the munged files

[ANDEWEGET]

That sounds plausible. Definitely better for me.
I wrote all the packers for the chunks and all needed classes/objects are set up. Now I only need to fill them with data from XSI. Right now that's just position and rotation at every frame and some smaller stuff like frame range etc.

BTW you're wrapper is already nicely incorporated into the exporter with it's own small module.

[tirpider]

Awesome!
It will be even better when we get a proper algorithm to put in.

The CRC math is just to wierd for me at the moment. My head is full of quaternion maths that I just don't fully understand. (I'm studying IK chains just to take a break from quats, heh.)

You know, it's going to be ZETools that eventually gets me to actually use XSI...

[AceMastermind]

Does anyone know if it's possible to export multiple animations in one .msh? Looking at the CYCL chunk it definitely seems possible.

I think this was a supported feature of Zero back when the devs created Star Wars: The Clone Wars (2002). The unit *.msh files also contained animations.

[ANDEWEGET]

Aye thanks Ace. One anim per .msh it is then.

@tirpider: Yeah, right now Im content with being able to calculate the crc in one direction tho. I can do polynomial division(or however its called in english) all right so if wed know the base polynomial and xor stuff the rest will be pretty easy and straight forward(I believe). I found an essay on reverse engineering crcs(the author even used python)he didnt give any code examples though. I might look into this a bit more when Im finished with the animations.
That would leave only shadow volumes and cloth. All other features the PandemicExporter had are done(not as fast and maybe not as reliable but still).

[lucasfart]

Bravo! One other question - how hard would it be to create an importer for .msh into XSI?

[ANDEWEGET]

It isnt really hard. Ive thought about it before. Depending on how fast I could import multi-k tri models(:D) with python I might have to switch to C++ again and then the struggle begins anew. Even if I dont have to use C++ this would still take a few months depebding on my motivation.

[lucasfart]

Wow. If just doing that would take you a couple of months, I admire your dedication with the rest of this addon.

[minilogoguy18]

Bravo! One other question - how hard would it be to create an importer for .msh into XSI?

I think no one should be worrying about an importer until a working exporter for the newest version of Mod Tool is created, an exporter is millions of times more important than an importer. I honestly wouldn't mod this game if I had to be restrained to using Mod Tool.

I go to bed every night with the hopes that an exporter is made at some point for Mod Tool, the community would be so much better if everyone could export their own stuff especially things that are animated.

[tirpider]

Bravo! One other question - how hard would it be to create an importer for .msh into XSI?

I think no one should be worrying about an importer until a working exporter for the newest version of Mod Tool is created.

A feature complete exporter can be used to validate an importer as well.

import a MSH then export it with no changes and see what is different.

It would make developing an imorter so much easier.

[minilogoguy18]

True but an importer still shouldn't be considered since there's no valid reason to import anything, an exporter is far more useful. Why rip off the games assets when you can create something yourself that you can be far more proud of?

[kinetosimpetus]

Why import anything? To learn from it for one thing. Not everyone can make better-than-professional models their first time seeing a 3D modelling program.

[THEWULFMAN]

^That, and also because sometimes exactly what the modder needs is to import a stock model and edit it. What if you need to add interiors to one of the Mos Eisley buildings? You want consistent maps. What if you need to edit the rebel trooper model, or clone trooper, etc? You need consistent sides.

There's a dozen reasons to import a model. Sure, making your own models is fantastic, but it's not everything.

[minilogoguy18]

I only meant to get the point across that to make the community grow more than it ever has in it's entire existence would be to have the ability for anyone to be able to export custom work. I've modded multiple games and compared to most modding communities this one seems to be developing so slowly primarily because very few people can export models to the game. I'd really love to see a lot more adverse content added.

Also while modeling may not come naturally to everyone far more people would come forth to try if they at least know that they have a chance of their stuff making it into the game without MUCH longer and unneeded hassle using multiple programs and hex editing just to get a simple object in game. With an exporter all you would need is XSI, you could toss all those other programs in the recycle bin and save a lot of time.

[Marth8880]

I agree with you in that case, minilogoguy. Also, heh, ever seen Unification? There isn't going to be a single stock model or asset used in the sides, and the majority of every map's models (minus the stock ones, which will be edited as well) will be/are custom. Yeah, we have some times on our hands, so what, big whoop, wanna fight about it?

[tirpider]

True but an importer still shouldn't be considered since there's no valid reason to import anything, an exporter is far more useful. Why rip off the games assets when you can create something yourself that you can be far more proud of?

By that same logic, why bother modding any game and "ripping off" thier engine and code.

I don't see anything wrong with modifying existing assets, and don't see it as ripping anything off at all.

There is an established community standard of aquiring consent to work with other folks assets, but the permissins to create derivative works are already in the mod tools eula.

There are a ton of edits to be made to the stock assets that may not bring new things into the game, but could greatly enhance what's already there.

Besides, the learning curve for XSI is so steep, being able to imprort an already working msh could only help folks learn it.

[ANDEWEGET]

The point of modding is modifying an existing game. That means you have to use the engine the game uses and usually the game logic, too.
An importer is useful. But just to some extent.
An exporter however is very useful. To way more extent. That's why I didn't create an importer yet(and maybe never will, we have some already).

@lucasfart:
I began writing the exporter in September 2011. That means I've worked on this for 7 months now. A lot of this time was spent on reverse engineering, learning python, learning the XSI object model/API, learning a bit of C++ etc etc.

[minilogoguy18]

^Which version of XSI are you basing it on? Will it be for Mod Tool? Heard a while back that making addons for Mod Tool was different since they didn't want people using a free program to create something that profit could be made from.

[Marth8880]

Except profit can't be made from ANDE's addon. :S

[ANDEWEGET]

@mini: I'm developing with SI 2012(64-bit). But I'm testing everything with the 7.5 ModTool(not very extensively tho ). The ModTool doesn't come with a XSI_SDK, that only contains examples and header files(.h) for C++ development. I couldn't build C++ plug-ins with only the modtool but plugins aren't restricted in any way(except of course if you're using newer/older commands which aren't in 7.5 etc). All of the supported scripting languages(python, javascript, VB etc)are usable.

@Marth:
If I sold it for 2€/$ I maybe could have sold one or two copies... I'd have to create some sort of activation system, only distribute the compiled script files(which you're still able to decompile into actually meaningful source code if they're python files[which they are]) etc etc. Lots of stuff I'm way too lazy to do. And I want to stay open-source.

A small update: I didn't have as much time as expected(mostly due to having to save the Sanctuary and defeat the pure evil). I still did some stuff to ease making general changes to object structure, maintaining code etc. I put everything into classes so there aren't any class-less functions floating around anymore. New version should be released next weekend or so.