[Bf-committers] MFace and edcode

[Tobias Oelgarte]

Campbell Barton schrieb:
> On Sat, Oct 9, 2010 at 12:39 PM, Tobias Oelgarte
> <tobias.oelgarte at googlemail.com> wrote:
>   
>> Hello,
>>
>> I make big steps forward, but i get some unexpected results. Now i guess
>> i made an error to ignore "edcode" from "MFace". Somehow i can assume
>> multiple meanings for this variable. They define in which order the quad
>> is build, but are the codes,
>>
>> #define ME_V1V2            1
>> #define ME_V2V3            2
>> #define ME_V3V1            4
>> #define ME_V3V4            4
>> #define ME_V4V1            8
>>
>> supposed to work?
>>
>> Greetings Tobias Oelgarte
>>     
>
> Looks like the only reason this is kept is so old files can be
> correctly loaded - this was set if the side of a face would draw an
> edge (before blender had edge data).
> The flags are used by the rendering engine but in the VlakRen struct,
> unrelated to modifiers.
>
> So for writing a modifier you can just ignore the edcode flag.
>   
Then I'm wondering which order the faces have. Going after the right 
hand rule i would assume that it is always [(v1,v2,v3),(v1,v3,v4)] for 
the two triangles of a quad. (First for triangle only). If this 
assumption is right then it should be save to test with MFace.v4 == 0 if 
its a triangle or quad.

It tried it out and i got expected results. But maybe it is just 
coincidence?