[tuhopuu-devel] Transform: feature or bug?

[Jason Saunders]

I also find the new transform a bit strange, but i agree with your
general workflow idea. The problem is that i can NOT see which axis is
"highlited" as you mention in the wiki. I get the axes popping up, but
when i hold and move MMB to hover over an axis, i see no visual
change.

If this could be fixed, i would very much like to test out this new method :)


On Fri, 28 Jan 2005 20:08:46 -0800 (PST), Martin Poirier
<theeth at yahoo.com> wrote:
> 
> --- Chris Want <cwant at ualberta.ca> wrote:
> 
> > 
> > No offense -- I know you've put a lot of thought
> > and work into this -- but I'm not crazy about this
> > new
> > behavior. I find it is slower to get what you
> > want with this than with the old way, so what
> > is the incentive to change it?
> 
> The reasoning goes a bit like this (that might sound a
> bit formal, on purpose so I can reuse it later ;) ):
> 
> The previous behavior assumed a lot from the user:
> 1- That the orientation of the global space in regard
> to the viewport was known.
> 2- In extensio, that the user knew what the 2D mouse
> motion corresponded to in global space since pressing
> MMB will constraint to the axis with the largest
> motion.
> 
> While those are safe assumptions with aligned view
> (top, front, side), it is often not the case with
> rotated viewports and prespective.
> 
> Also, for some viewport orientations, it was
> particularly hard to constrain to some axis with MMB
> alone (then again, you have the X,Y,Z keys, but these
> again assume orientation knowledge).
> 
> I think the current implementation addresses all those
> issues:
> 1 & 2- It gives visual clues to the resulting
> constraints.
> 3 - It enables easy constraining on the 3 axis
> regardless of the orientation of the viewport.
> Excluding constraints to the orthogonal axis with
> aligned view, but the transformations were it would
> actually do something (Rotate and probably Wrap)
> already have that behavior by default.
> 
> However, I didn't want to have to sacrifice speed (not
> too much anyway) to gain these.
> - This is why I used a Click - hold/move - Release
> motion. For quick motions, this is faster than click -
> move - click.
> - For aligned views, clicking MMB has the same
> behavior as before.
> 
> Another thing is that this also helps with the
> limitation of the X,Y,Z hotkeys which required a
> second of thought (if you could remember the
> orientation of the viewport) and trial-error if you
> didn't by using visual clues, it essentially drops all
> orientation awareness assumptions (which is good and
> bad, but not requiring something doesn't mean you
> shouldn't do it anyway).
> 
> I'd like to hear why you like the old one better. The
> current design isn't set in stone, so...
> 
> 
> > P.S. Transform in armature edit mode is buggy.
> > If the 'IK' button has been pressed for a
> > bone, part of the bone is left behind in
> > grab mode ... it's really odd.
> 
> Good call. Frankly, I haven't touched the bones much
> and the Edit Bones to TransData transformation wasn't
> done by me (and, IIRC is basicly a straight copy/paste
> of the old transform way without further thought) so I
> can't garanty anything. Will have to look into that
> (and Pose Mode too, obviously).
> 
> Martin
> 
> 
> __________________________________
> Do you Yahoo!?
> The all-new My Yahoo! - Get yours free!
> http://my.yahoo.com 
> 
> 
> _______________________________________________
> tuhopuu-devel mailing list
> tuhopuu-devel at projects.blender.org
> http://projects.blender.org/mailman/listinfo/tuhopuu-devel
>