Yep, I see what you're saying. I tried several meshes with the replace_tex bitmap on them and it appears to grab the first mesh in the node hierarchy with "Replace_tex" as the bitmap and ignores the other meshes. But it's still a good idea. Looking over the model again, I noticed that it's anim mesh. Can't really tell how important that particular bit is. I'm probably going to keep plugging away at the thing I was looking at for the other thread, but if anyone wants to look at it more, shoot me a PM and I'll see what I can rig up for ya.
(please ignore the following three lines) It is animesh because it is a flag, that is weighted to move with the wind. (I mean the individual vertices use a bit weight of the color red, to determine how much they react with the strength of the wind) I doubt that has any bearing on the texture bit,
UNLESS they (bio-engineers that wrote the code) are using the animesh tag as an extra bit to toggle special processing in general.
(please ignore the following two lines) I doubt that is the case here, and it would be easy to check by just assigning that particular replace_tex to a solid object, say a stone and see if something different happens. IE: The animesh tag and associated parent, are only there because the flag moves with the wind, and has nothing to do with the texture applied.
I would suspect that it simply replaces the assigned texture as you have already found.
Likely, tt is simply a way to reskin, without having to manually assign the new texture name to each mdl... IE, if all the mdls in a given set were created to use the various replace_tex, replace_tex01, replace_tex02 etc in that 2da, you could simply assign a NEW 2da, IE an overriding 2da, to re-skin without having to manually go through and either create new textures of the same name (which has override issues with OTHER sets in a given mod) or by having to manually rename each texture bit in each module. It was likely something that a particular engineer use as his/her style or something that was just never actually implemented as it was more trouble that it was worth with outsourcing etc... Bio had outsourced some tilesets and other objects, creatures, etc, and by having things outsourced, it becomes more problematic attempting to force artists to create using specific naming conventions etc...
You should know how it goes, artists in particular, are very prideful (rightfully so in some cases) and want things their own way, seldom do they like to follow rules set by others unless it is absolutely necessary. So forcing someone to name a texture for a wall to be "replace_tex02" or "replace_tex04" or whatever you choose, becomes more and more problematic... sensibly, most textures are named in a way that makes some sort of logical sense as to the object they are coloring with that texture. Using the replace tex would be powerful only if you were choosing to re-use all the objects and get them re-skinned with a single set of textures applied by that 2da, and if you were wanting to have it as a centralized, controlled re-skin process. Bioware didn't bother with that in the long run, they just re-skinned whatever needed it the same way that the community has done, either by replacing existing textures using the same name OR by manually replacing the bitmap fields in the objects that are being re-skinned. (The later being the best way as it avoids the bad replacing of OTHER objects, tiles, etc with a renamed texture.)
P.S. Edit notes: sections converted to italics above are only left so that folks can follow the entire conversation, but do reflect BAD statements created by me personally. To whit, I stated that animesh was used for the flexible flag. when in actuality that would have been accomplished using auroraflex sometimes referred to as danglymesh. Thanks to Zwerkules and OldTimeRadio for making note of my mistake. Basically, just ignore the italicized text above and you will be better off! Again, I am leaving it in place to allow the ability to completely follow the entire conversation(s) herein.
Modifié par Bannor Bloodfist, 14 mars 2016 - 10:20 .