Re: [round-display] [motion-path] Orientation of elements due to merging polar positioning and motion path

On Thu, May 19, 2016 at 4:21 AM fantasai <fantasai.lists@inkedblade.net>
wrote:

> On 05/17/2016 08:11 PM, Shane Stephens wrote:
> >
> >> Wait, really? That seems very confusing to me. I would have
> >> thought that the motion-* and rotate-* were all independent
> >> and 'translate' operated on top of that result.
> >
> > You need to specify an order that everything gets assembled into a
> transformation matrix.
> >
> > Currently we have:
> >
> > apply motion (= motion translation, motion rotation)
> > apply translate
> > apply rotate
> > apply scale
> > apply transform
> >
> > The translate, rotate, scale ordering is important so that
> > those three properties appear to operate independently of each
> > other. We could move the motion rotation to be at the same
> > place as rotate, but it seems weird to split the motion
> > components like that.
> >
> >> You're saying that 'rotate' affects the coordinate system of
> 'translate'?
> >
> >No, motion rotation does.
>
> This seems very counter-intuitive to me. I would not expect
> the coordinate system of translate/rotate/scale/transform
> to be affected by motion at all. The goal is for these all
> to behave as if they were independent right? But with this
> ordering they're not.
>

I'm not sure that is a goal. Would you expect motion rotation to affect
transforms? Or transforms to affect motion? You have to pick one. I think
translate/scale/etc. should act the same as the transform property, and we
can't split that up to insert motion components.

Actually, based on your reactions here I'm leaning towards us adding a full
motion transform function. There isn't a perfect ordering of transform and
motion components where everything always works as expected but at least
that way more advanced users can select an order.


>
> What if you did
>
> apply translate
> apply motion
> apply rotate
> apply scale
> apply transform
>
> ?
>

This would keep translate components in the global coordinate system but
then
translate: 100px;
and
transform: translate(100px);
would sometimes act differently to each other, which is weird.


>
> p.s. Please either use plaintext email on www-style, or
> successfully harass the Gmail team to fix b/19483003 ?
>

Ah, sorry :( I think that whatever messed up happened because I manually
quoted an email to merge two replies. I won't do that again. Unfortunately
inbox has no plain text option yet.

Cheers,
    -Shane


>
> ~fantasai
>