[Bf-committers] math lib reorganization

joe joeedh at gmail.com
Mon Nov 9 18:51:27 CET 2009


+1 from me, so long as we get an email warning us before this happens. :)

Joe

On Mon, Nov 9, 2009 at 8:11 AM, Campbell Barton <ideasman42 at gmail.com> wrote:
> +1 from me too
>
> On Sun, Nov 8, 2009 at 10:27 PM, Martin Poirier <theeth at yahoo.com> wrote:
>> Hi,
>>
>> Thumbs up from me.
>>
>> One little thing though, since we're renaming everything, now would be a good time to correct mistakes. For example, ortho_m3 (for Mat3Ortho) should be normal_m3, since it doesn't orthogonalize anything, it only normalizes it.
>>
>> Martin
>>
>> --- On Sat, 11/7/09, Brecht Van Lommel <brecht at blender.org> wrote:
>>
>>> From: Brecht Van Lommel <brecht at blender.org>
>>> Subject: [Bf-committers] math lib reorganization
>>> To: bf-committers at blender.org
>>> Received: Saturday, November 7, 2009, 7:53 PM
>>> Hi,
>>>
>>> I've made a patch to reorganize BLI_arithb.h, the Blender C
>>> math
>>> library which has various consistency issues.
>>>
>>> What I've done is splitting it up into multiple
>>> BLI_math_*.h files,
>>> renaming functions according the proposal by Ton/Zr a few
>>> years ago,
>>> and making function parameter order consistent. All the
>>> code is still
>>> in arithb.c, need to split up into .c files according to
>>> the .h files
>>> still. Mainly the tricky thing was changing all the source
>>> code that
>>> uses these functions to rename them and change the
>>> parameter order,
>>> which was done using a python script.
>>>
>>> The problem is merging these changes into branches, as it
>>> would give
>>> many merge conflicts, and new code that is likely also
>>> using the math
>>> functions. I think it's doable though, the following steps
>>> should make
>>> this possible without unnecessary conflicts:
>>> * merge up to the math lib refactor commit
>>> * run the python script to do the renaming
>>> * merge the following revisions
>>>
>>> There's more changes I'd like to do, like making some
>>> functions
>>> inline, parameters const, adding functions, and also
>>> changing a few
>>> infrequently used functions more drastically .. but these
>>> should not
>>> affect other source code as much and give merging issues,
>>> so could be
>>> done afterwards still.
>>>
>>> I don't want to interfere with 2.5 beta 0, so committing
>>> this can be
>>> postponed. I think this is fairly low risk though, most
>>> likely bug
>>> this would cause is if I've made a mistake in function
>>> parameter
>>> reordering, double checked the orders but it's still
>>> possible of
>>> course. Using my simple rendering test suite also survives.
>>> I guess
>>> this depends on the beta 0 schedule, but personally I think
>>> it's quite
>>> safe to commit if we don't plan a beta 0 in the next week.
>>>
>>> Here's the full patch with the new header files and all
>>> source changes
>>> included to test, and the script I used:
>>> http://users.pandora.be/blendix/arithb_reorganization.txt
>>> http://users.pandora.be/blendix/arithb_refactor.py
>>>
>>> Brecht.
>>> _______________________________________________
>>> Bf-committers mailing list
>>> Bf-committers at blender.org
>>> http://lists.blender.org/mailman/listinfo/bf-committers
>>>
>>
>>
>>      __________________________________________________________________
>> Looking for the perfect gift? Give the gift of Flickr!
>>
>> http://www.flickr.com/gift/
>> _______________________________________________
>> Bf-committers mailing list
>> Bf-committers at blender.org
>> http://lists.blender.org/mailman/listinfo/bf-committers
>>
>
>
>
> --
> - Campbell
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> http://lists.blender.org/mailman/listinfo/bf-committers
>


More information about the Bf-committers mailing list