[Bf-committers] IRC meeting minutes sunday 28th nov

[Martin Poirier]

--- joeedh <joeeagar at prodigy.net> wrote:

> Ton Roosendaal wrote:
> 
> > - Transform refactor
> > Martin mentions steady progress, with some bugs to
> iron out. Scaling,  
> > constraints and support for Ipo Keys are on the
> todo. He also likes 
> > to  check on unifying new transform() with the
> ones now used in 
> > UV-Image  and IpoWindow.
> 
> Hay, I did scaling.  I even got it to work almost
> exactly the same as 
> the old transform().  Hm, I thought I emailed Martin
> a faily finished 
> source a while back. . .

I redid it from scratch actually (pretty much done as
of now). Since I did many changes since you sent me
the code, I'd have to redo parts of it anyway. Also, I
redid it without looking at the old code (something I
don't think you did) so I could save myself from
reusing many horrors.

Rotate is pretty much polished now. I just redid the
angle calculation (which I had copied from the old
one) in a more accurate way (I think).

Should update the wiki latter tonigh.

>  I've been thinking of this for a while now.  I
> think that this might 
> require a more unified transform data structure,
> where all the verts (or 
> objects) are stored, instead of just the ones that
> are selected.  Hm.

Actually, it's not logically limited to the ones that
are selected (though it is right now) but rather to
every data we want to transform. For example, with the
PET tool, we'll have all data being affected on top of
those selected. Also, technically speaking, with some
slight clean up, we could even have mixed data types
in there (modify mesh vertice, curves, cvs, objects,
... at the same time).

Martin


		
__________________________________ 
Do you Yahoo!? 
Take Yahoo! Mail with you! Get it on your mobile phone. 
http://mobile.yahoo.com/maildemo