Free Allegiance Forums

LANS wrote:QUOTE (LANS @ Apr 29 2010, 02:02 PM) Oh, ad thanks for that link. I'll have to reaad over that study.

I dunno, it may be crap, just the first thing from a google search.

Part of the problems in moving forward with the Artwork is that the existing tools are way out of date.

Take a look at the original artwork and you'll see it uses "Retained Mode" which hasn't been updated since DX3

What's "Retained Mode"?

exactly.

This project is now temporarily on hold until I clear up some IRL stuff. I'll probably have some time to continue work on this by mid-august.

As this topic evolved, I lost track of what your project is about ...

Do yo just want to fix GT bases or do a more general tool to edit/compute CVH stuff ?

Whatever, since you're busy as I am, i'm releasing the unfinished code of AMT 2.0.

You can find it herein the 'AMT2' directory of this SVN. (just checkout the AMT2 dir).

It's written in C# and you'll need SlimDXto build it.

As said, this is unfinished stuff, it barely works but load MDL, CVH and .X files and display them as well as CVH/MDL specific stuff (doors, lights,...).

There is no editing nor saving.

There is the code to compute CVH hulls. It's based on John E. Lloyd's QuickHull3D written in Java. I've ported it to C# as a separate library so you might just want to use this (or use John's if you work in Java).

There is also a library for '3D manipulators', some XNA code that I ported to SlimDX. I planned to use them for editing the CVH/MDL specific stuff (doors, lights, etc). I don't recall how far I went.

All this stuff is 2+ years old already and I haven't looked at it since. I'm not even sure you can compile it with latest SlimDX ...

So use it more as a sample/starter.

My goal with this project is to produce a verification tool that will tell you what is wrong (if anything) with any given CVH file. It will (when complete) check for lines where 3 planes intersect or where there is a hole because the lines surrounding the hole are not all in a flat plane.

This will tell us what is wrong, if anything, with any of the CVH models.

ok so you only need a 'cvh reader' .There is a C# lib which can read CVH files in the AMT2 source if you want.

From there, it's just 'pure maths' to verify/compute what you want.

OR you can use AMT2 'as is' as a tool to visualize CVH files (it renders them in 3D) and eventually detect glitches in them by just looking at them ... sometimes it's simpler to just use "the human eye"

I have a pretty good idea of how to do what I want mathematically, including how I'll need to read and store the data from the CVH files. I'll have a look at AMT2 when I have time. What's unfinished about it? I could probably finish some of it instead, and maybe even integrate this project into the AMT2 code.

edit: And I don't work in java. I stick with C/C++, and I've never programmed anything for windows outside of school assignments, which I was required to use C# for (and these were very simple).

AMT2 is HIGHLY unfinished.. it's more a prototype actually... I don't recommend trying to finish it.

I've a modular approach when coding with .Net so I started gathering or building 'modules' (cvh reader, mdl reader, quickhull lib, manipulators) and then I quickly drafted AMT2 using these modules. But just to be able to test the modules.

So the central part is a real mess not worthy looking at, it's just a quick hack to use the modules:

I used the (deprecated) Managed Direct X DXUT framework which I ported to SlimDX (oh god...). This clearly not the good way to do. Using another 3D rendering engine or the framework included in SlimDX is more recommended.

The part that does all alleg stuff and need reworking is in frames.cs. It's the quick hack/attempt and it needs comments and rework. It mainly tries to redo what the legacy tools do (mdlc/xmunge) when processing .X files to generate mdl/cvh files.

Also the thing is still named 'MDLView' because it was based on my old MDLView project.