Yes, absolutely; users will cherry-pick features from different
versions of HTML to suit their desires, and plant them into whatever
version of the format they like / their CMS supports / is the latest
fashion. They'll expect browsers to seamlessly handle this.
This only underlines the importance making the syntax and
disambiguation of @rel compatible among the different versions of
HTML; that would serve these users well. Using a different version
attribute is not an excuse to throw backwards compatibility out of the
window, and it's not like RDFa is an old specification, so ignorance
of the issues surrounding versioning and compatibility can't be claimed.
It certainly doesn't justify putting the subset of users who want to
do the right thing and have valid, unambiguous markup into a place
where they can't, because the features they need are spread out among
those incompatible versions of HTML. Not every consumer of HTML is a
browser.
Users who are attracted to RDFa today are likely to have been
influenced either directly or indirectly by Zeldman and his
brethren. They include an XHTML DOCTYPE and try to be careful about
quotes. The few that actually read specs will see that XHTML 1.0
Transitional allows the use of the text/html MIME type.
Sorry, RDFa just became mainstream, when the Creative Commons started
showing people how to use it by example. Speaking of attraction, I'm
very much reminded of the US legal concept of an attractive nuisance
here. Anyway...
From where I'm sitting, RDFa should not have gone out the door as it
is, and because it did we have some damage to contain. Likely it's not
too bad, owing to the bad state of @rel in HTML anyway, but it has
effectively created one more thing to sniff in HTML -- "what rel
convention is in use here?" -- with all of the ambiguity and issues
that entails.
So, I have a fair idea of what I'm going to write in the next Link
draft now (see recent messages to Ben for a rough idea). What I really
want to know -- and this is why the TAG is still on the CC list here
-- is what's going to be done to prevent this from happening the next
time.
Cheers,
P.S. Sam, I'm confused; you've brought up whitehouse.gov in this
context a few times, but AFAICT they don't serve RDFa on their front
page. Yes, they're serving XHTML with a text/html media type, but
that's very wide and understood practice. Please explain?
On 01/03/2009, at 9:46 PM, Sam Ruby wrote:
Mark Nottingham wrote:
On 01/03/2009, at 10:33 AM, Ben Adida wrote:
Mark Nottingham wrote:
Are people using RDFa in HTML using the profile mechanism, or
xmlns, or
both? Do they flag the use of RDFa in any way (like @version does
for
XHTML+RDFa)?
We recommend using the doctype with @version. We've determined that,
while @profile is the right approach for interpreting new values of
@rel, it doesn't cover new attributes, so it's not the right way
to flag
RDFa.
Sorry, I should have been more clear; I'm talking about in HTML4
(since CC doesn't require or recommend XHTML for CC+).
[snip]
But an HTML4 parser has absolutely no business knowing about xmlns,
period.
I think we use these terms differently than users do.
$ curl --silent --head http://www.whitehouse.gov/ | grep -i content-
type
Content-Type: text/html; charset=utf-8
$ curl --silent http://www.whitehouse.gov/ | head -3 | tail -1
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd
">
From a user perspective features are things like canvas or RDFa.
Add a little markup, include a bit of javascript and you have
awesome graphics or delectable metadata.
From a HTTP perspective, HTML is unversioned... there is no way to
indicate via a MIME type whether something is HTML4 vs HTML5, just a
way to indicate something is HTML vs XHTML; and that is done is a
way that is beyond what most users can control and is at variance
with what users understand XHTML to be and is entirely unsupported
by the dominant browser. And triggers unforgiving error recovery
strategies.
From a browser perspective, HTML is unversioned. No browser that
supports canvas will refuse to do so simply because that tag is
included in a page which includes the HTML4 DOCTYPE. Firefox (to
pick an concrete example) has no plans to have a separate HTML5
parser. Javascript libraries which will extract RDFa data will do so
even in quirks mode. Even if the attributes happen to be named
starting with the characters 'x', 'm', and 'l'.
Users who are attracted to RDFa today are likely to have been
influenced either directly or indirectly by Zeldman and his
brethren. They include an XHTML DOCTYPE and try to be careful about
quotes. The few that actually read specs will see that XHTML 1.0
Transitional allows the use of the text/html MIME type.
I am biased. I believe that we should cater to these users. They
outnumber us.
- Sam Ruby
--
Mark Nottingham http://www.mnot.net/
