Re: xywh on images with no clearly defined size

On Tue, May 17, 2011 at 6:32 AM, Jack Jansen <Jack.Jansen@cwi.nl> wrote:
>
> On  16-May-2011, at 00:19 , Silvia Pfeiffer wrote:
>
>> Hi Jack,
>>
>> I am concerned about this position. I shared it when we started the
>> group and I never thought spatial fragments would become relevant in
>> the near future. However, now it has happened and the CSS WG wants to
>> pick it up. Therefore, we cannot leave this issue undefined. We can
>> state that we need help by the CSS WG and the SVG WG to resolve this
>> and maybe we need a F2F meeting or phone conference with them or
>> something to discuss this through. But I don't believe putting our
>> heads in the sand is going to help - it will just result in the CSS WG
>> having to define it themselves.
>
>
> Silvia,
> are you sure about this? We're talking here about an extreme corner-case: semantics of applying an xywh-media fragment to an image-container that contains multiple images of different sizes. (Unless I miss something, and this isn't the corner case I think it is)
>
> If you indeed have an indication that this is important enough for CSS that they would define their own spec for this case, then I fully agree that we should take the lead in this. Otherwise, there are numerous other corner cases that we've also left for the future, and that I would personally consider much more important, such as temporal media with non-continuous timelines or interaction with HTTP internationalisation  and content negotiation (to name just two thing that spring to mind).
>
> All in all, I'm not against putting effort in this, but I would like to have an indication that it's the best way to spend our time.

Hi Jack,

The pictures that consist of multiple images are just one case.
There is also the case of SVG and the links inside it.
They may be corner cases, but details is what we are looking at now.

I think we should add a sentence for both of these use cases. I don't
think they are terribly difficult to resolve, so while another group
has such a need, we should look into it. The CSS working group is
clearly looking at this issue right now and so it makes sense to
address it right now and resolve it with them. The SVG case is likely
the more complicated one and will need a discussion in both groups
IMO. It has been shown already that browsers react inconsistently to
fragment addressing on SVG, so it's a serious issue. Is somebody here
a member of the SVG WG and can sort it out?

As for temporal media with non-continuous timelines: do you have any
examples that we should work on? Has somebody raised a problem to you?

As for HTTP internationalisation and content negotiation: what is the
problem? Is it that HTTP may indicate a Content-Language of x, while
the request is looking for a track=y where y is a language that may be
different from what the server can provide? I think that is something
that we will have to leave to server configuration to sort out. (Also,
in reality, HTTP content negotiation is much less used than in common
belief - typically, people create completely separate Web sites when
they need to offer a site in multiple languages.)

Cheers,
Silvia.

Received on Monday, 16 May 2011 21:58:39 UTC