[Bf-python] Last minute addition

[Campbell Barton]

Im not so against this but it opens up a can of worms...

IMHO
Noise could also be a Mathutils submodule

And Pose, Group, and Modifier should be Object submodules.

Not sure weather spesific constructs for line/triangle's are that 
important, why not just use a tuple of vectors for a line and a 3 for a  
triangle?
is there much gain in having a spesific datatype? - Less type checking I 
suppose..





Gilbert, Joseph T. wrote:
> Post-issue commentary....
>
> I have to agree with Campbell that Draw is for GUI and that mathutils
> was written for the purpose of general math utilities for importers. I
> can see that this method was written more for working with geometry
> import however.
>
> It was my intention originally to add some math primitives such as
> triangle, cube, etc. to allow for some geometric math as well. Because
> none of this code was added to mathutils originally we ended up adding
> theeth's line intersection code to Mathutils. 
>
> Since we now have a geometry module with a general purpose utility it in
> I would feel more comfortable moving it as a submodule of mathutils as
> Mathutils.Geometry.  We could then add some objects to represent Line,
> Triangle, etc. and some math for boundary checks, and intersections,
> etc. while the Geometry module (like mathutils) contains general
> geometric utilities for importers/exporters.
>
> Joseph
>
> -----Original Message-----
> From: bf-python-bounces at projects.blender.org
> [mailto:bf-python-bounces at projects.blender.org] On Behalf Of Campbell
> Barton
> Sent: Sunday, July 02, 2006 2:00 AM
> To: Blender Foundation Python list
> Subject: Re: [Bf-python] Last minute addition
>
> All valid points, but your kind of missing some issues that casued me to
>
> add it.
> 1) Mesh.fill() requires an object and mesh data, cycles editmode on and 
> off and is also tagged experemental.
> This means that for every fgon in an imported model it needs to go and 
> create mesh data, a new object, add the verts then edges, fill in the 
> data, get the faces indicies and return them. then clean up the mesh, 
> object, restore  editmode data.
>
> 2) I have tested 500+ lightwave files out of 1700 that I have to import.
>
> many of these use a lot of fgons  so this function can be called a lot 
> for each import. I have tested with .fill() and PolyFill() - polyfill is
>
> of course a lot faster, and dosent make the cursor flicker (seterate 
> issue with cycline editmode a lot).
>
> Having good (fast+high quality) importers for blender is important, this
>
> fuction effects Lightwave and Obj.
>
> When testing fill() I kept on finding bugs and fill crashed blender - (3
>
> seperate bugs in fill()) - its good that blenders scanfill is now more 
> stable, but directly  accessing the function is much neater.
> In exasperation! - (trying to get these importers not to crash with 
> blenders fill()), I asked Ton (who fixedall the bugs) if I could  just 
> access scanfill directly and add a hook in python. he seemd to think 
> this would be okay, and gave me some help adding it in.
>
> The PolyFill Function its self isnt that complex, I have tested on 
> importing meny models with no crashes.
>
> If you guys want Ill remove if but Im just putting forward a rationale 
> for adding it in the first place, and that I added it when fixing bugs 
> in Lightwave importer for release.
>
> Could we move it to BGL? Seems more closly linked to BGL then Draw.
>
> - Cam
>
>
> Martin Poirier wrote:
>   
>> --- Martin Poirier <theeth at yahoo.com> wrote:
>>
>>   
>>     
>>> --- Ken Hughes <khughes at pacific.edu> wrote:
>>>
>>>     
>>>       
>>>> If we need to "kick this idea around" then I have
>>>>       
>>>>         
>>> to
>>>     
>>>       
>>>> ask, why is this 
>>>> being added during a CVS freeze?  It doesn't
>>>>       
>>>>         
>>> really
>>>     
>>>       
>>>> fix a bug, and I 
>>>> don't think it's filling a hole in the API. I
>>>> believe the commit said it 
>>>> was to avoid creating a mesh just to fill it, but
>>>> that just means this 
>>>> is redundant.  This is not coming from me as the
>>>> "Mesh API father", it 
>>>> just seems rushed.
>>>>       
>>>>         
>>> I'll second that.
>>>
>>> Anything concerning the API has to be thought out,
>>> because releases are "contracts" with *are* users
>>>     
>>>       
>> that
>>   
>>     
>>> about the fonctionality and structure of the API.
>>>     
>>>       
>> Obviously, that sentence should be read as: ...because
>> releases are "contracts" with users, that *are* about
>> the fonctionality and structure of the API.
>>
>> How that "are" moved two words to the left, nobody
>> shall every know... :P
>>
>> Martin
>>
>> __________________________________________________
>> Do You Yahoo!?
>> Tired of spam?  Yahoo! Mail has the best spam protection around 
>> http://mail.yahoo.com 
>> _______________________________________________
>> Bf-python mailing list
>> Bf-python at projects.blender.org
>> http://projects.blender.org/mailman/listinfo/bf-python
>>
>>   
>>     
>
>
>   


-- 
Campbell J Barton

133 Hope Street
Geelong West, Victoria 3218 Australia

URL:    http://www.metavr.com
e-mail: cbarton at metavr.com
phone: AU (03) 5229 0241