Re: [css4-image] element() comments

On Tue, Feb 4, 2014 at 6:34 AM, Tab Atkins Jr. <jackalmage@gmail.com> wrote:

> On Sun, Feb 2, 2014 at 2:06 PM, Robert O'Callahan <robert@ocallahan.org>
> wrote:
> > http://dev.w3.org/csswg/css-images/#element-notation
> >>
> >> The function represents an image with its intrinsic size equal to the
> >> decorated bounding box of the referenced element
> >
> >
> > Giving element() an intrinsic size is actually super bad. It creates
> almost
> > arbitrarily bad circular layout dependencies; e.g. any <li> element's
> size
> > can now depend on the size of any other element in the document!
> Detecting
> > and fixing the circularity isn't easy either, because you can combine
> this
> > with existing dependencies to create cycles in all kinds of ways. Since
> this
> > is mostly useless anyway, I propose specifying that element()s have no
> > intrinsic dimensions at all.
>
> Unless I'm misunderstanding, the spec is describing Mozilla's current
> behavior.  This is illustrated by the first example in
> <https://hacks.mozilla.org/2010/08/mozelement/> (the one with white
> text on an orange background), and a few others in that page.
>

Good point. Our implementation doesn't support list-style-image:element()
yet, and so element() intrinsic size currently can't affect layout in
Gecko. But we're adding that now and hitting this problem. So either we add
a parameter to the object-sizing algorithm to distinguish between the cases
where object sizing can affect layout and those where it can't, or we
change our behavior. I prefer the latter since object sizing is already too
complex for my taste.

Rob
-- 
Jtehsauts  tshaei dS,o n" Wohfy  Mdaon  yhoaus  eanuttehrotraiitny  eovni
le atrhtohu gthot sf oirng iyvoeu rs ihnesa.r"t sS?o  Whhei csha iids  teoa
stiheer :p atroa lsyazye,d  'mYaonu,r  "sGients  uapr,e  tfaokreg iyvoeunr,
'm aotr  atnod  sgaoy ,h o'mGee.t"  uTph eann dt hwea lmka'n?  gBoutt  uIp
waanndt  wyeonut  thoo mken.o w