|
 |
Op 07/02/2024 om 01:32 schreef Bald Eagle:
> Thomas de Groot <tho### [at] degroot org> wrote:
>
>> Impressive. I have started to play, first with a complex mesh2 object,
>> and it works like a charm. I still need to dig deeper obviously, but I
>> love it.
>
> Thanks Thomas,
>
> I'm glad it's working well for you.
> Please try to apply a null/identity transform where the starting points are the
> same as the destination points.
>
> I rendered a loop of triangles and they all wound up exactly where they were
> supposed to, so I'm hopeful that this is all it will take.
>
OK. I shall do that today. There are a couple of questions I am hoarding
at the moment, but that can wait until I have a more complete picture of
the thing.
> I believe this is what is referred to as a "rigid body transform", as opposed to
> an affine transform, which may apply things like scaling or shear.
> So I think I just need the ordered triple, which gives us a triangle with a face
> normal, and that's all that's necessary and sufficient to completely describe
> the orientation.
>
It is smart indeed.
I dimly remembered an old post about Euler matrix rotations:
https://news.povray.org/povray.general/thread/%3Cweb.44d8fc1ca05d1ece40d56c170%40news.povray.org%3E/?ttop=438426&toff=2650
which I almost lost but was hiding in an obscure nook. At the end of the
thread there is a bit of interesting/useful code which seems related to
what you have so smartly cooked up here.
> I'm sure TOK probably coded something like 15 years ago ;)
>
Oh! I am sure he did!
--
Thomas
Post a reply to this message
|
|
