[whatwg] RDFa

Tab Atkins Jr. jackalmage at gmail.com
Sun Aug 24 14:17:32 PDT 2008

On Fri, Aug 22, 2008 at 2:53 PM, Ben Adida <ben at adida.net> wrote:

> > This doesn't allow you to say things about *another* resource, but
> > that's OK, because out-of-band metadata and data often travel their
> > separate ways.
> It's not okay for us. There are no good ways to embed metadata in media
> files that the average user can understand. So we need it in the
> enclosing HTML. With our approach, someone can take a chunk of HTML we
> give them, and paste it right in their page. We need that chunk of HTML
> to carry metadata with it.

What's the intended way to keep the metadata correct once the file leaves
the web page?  Assume this proposal is accepted and used.  At some point in
the future I see a picture on some site that I like and would to save to my
computer.  This picture is licensed under CC-SA, and this fact is indicated
using the scheme suggested in ccRel.  I do a right-click, Save As, and save
it.  Some time later, I'm cruising through my picture collection looking for
something to use in an article I'm writing, and I come across that picture.
Will I still know that it's licensed under CC-SA?  If so, how?  It seems
that, since the metadata is completely "out-of-band" that at this point the
licensing information would be completely lost.  Is this a use-case that has
been looked into?  Are there any ideas about how to keep the document's
metadata with it if this scheme is adopted?

Looking toward the ease-of-use angle, would it not be just as easy or
perhaps even easier for someone to pass their photo/video/etc. through an
encoder, either online or built into their editting application, that would
insert the licensing information directly into the resource?  If this were
done to a picture on a web site, my browser could still extract the
licensing information from it and inform me of the terms, but it would be
guaranteed to stay with the image unless I took pains to strip it out.

It's certainly true that the average user wouldn't know how to edit an image
to insert licensing information (or any other metadata), but I don't believe
they should be expected to.  Their editting application can do it for them,
or they can pass it through an online service that knows how to handle
common filetypes.  This is likely inappropriate for generic metadata, but
most metadata wouldn't be generic - it'll be some specific type that is
well-known enough to be useful.  This would still allow a generic syntax,
though (rdf in xml, frex, if text comments are available).

I'm simply somewhat confused about how this embedded metadata is supposed to
be carried around with a file.  I understand that there are many types of
metadata that you might want to embed in a webpage that *aren't* associated
with a separate resource (like contact info, geolocation, etc.) but for
specific cases like ccRel it seems like an in-band metadata storage would be
maximally useful, without appearing to add any extra cognitive load on the

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.whatwg.org/pipermail/whatwg-whatwg.org/attachments/20080824/6bc25e54/attachment-0001.htm>

More information about the whatwg mailing list