[uf-discuss] HTML5, Microformats and RDFa

[Manu Sporny]

There have been several threads discussing Microformats, RDFa and HTML5
that are occurring on the WHATWG mailing list. The discussion relates to
whether or not HTML5 should depend on the Microformats community to
solve HTML5's semantic markup issues, or if both Microformats and RDFa
should be considered for semantic web markup issues.

The start of the discussion is here:

http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-August/015860.html

and continues here:

http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-August/015875.html

I have authored a blog reply, stating that HTML5 should not depend on
the Microformats community to develop all semantic web vocabularies, the
reasoning can be viewed here:

http://blog.digitalbazaar.com/2008/08/23/html5-rdfa-and-microformats/

and my first response to the WHATWG mailing list

http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-August/015949.html

Things start getting dicey here:

http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-August/015892.html

and here:

http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-August/015950.html

and my second response to the WHATWG mailing list, outlining some of the
shortcomings of Microformats and stating what differentiates RDFa in
it's approach:

http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-August/015957.html

Posting to this list because there are many on here that would be
interested in the WHATWG's current position on web semantics: "not
important enough to consider as part of the HTML language". Note that
the XHTML1.1 and XHTML2 workgroups have already accepted the position
that: "web semantics are important and a standard method of semantics
expression is necessary for the future development of the web".

-- manu
_______________________________________________
microformats-discuss mailing list
microforma...@microformats.org
http://microformats.org/mailman/listinfo/microformats-discuss

[Scott Reynen]

On [Aug 25], at [ Aug 25] 8:47 , Manu Sporny wrote:

> There have been several threads discussing Microformats, RDFa and
> HTML5
> that are occurring on the WHATWG mailing list. The discussion
> relates to
> whether or not HTML5 should depend on the Microformats community to
> solve HTML5's semantic markup issues, or if both Microformats and RDFa
> should be considered for semantic web markup issues.
>
> The start of the discussion is here:
>
> http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-August/015860.html
>
> and continues here:
>
> http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-August/015875.html
>
> I have authored a blog reply, stating that HTML5 should not depend on
> the Microformats community to develop all semantic web vocabularies,
> the
> reasoning can be viewed here:
>
> http://blog.digitalbazaar.com/2008/08/23/html5-rdfa-and-microformats/

Manu,

I agree it's unfortunate microformats, created to fill gaps in HTML,
are now suggested as a reason to not fill those gaps. That said, it
seems to me you're misreading your opposition here. Microformats are
based entirely on HTML (which Ian fully understands, having
participated early on in the microformats community), so the
underlying argument being made against RDFa is that *HTML* is already
sufficient, that there is no need for it to solve the wider problem
RDFa would solve. As Ian said (with no mention of microformats):

> It would be helpful if you could send a separate message that is
> specifically asking for the changes you desire, and explaining what
> problem it is they address, and what research shows that that is an
> important enough problem that we should address it.

Whatever shortcomings microformats or the process have should be
irrelevant to making such a case for RDFa. Microformats explicitly do
not seek to solve the wider problem as RDFa does, so rather than
trying to convince people that RDFa solves the problem better than
microformats, I suggest you convince them that the wider problem would
actually be useful to solve. (That microformats don't solve it should
then be self-evident, as microformats do not even attempt to solve
it.) I think comparing RDFa to microformats actually hurts your
argument by suggesting they solve the same problem and reinforcing the
notion that the wider problem RDFa seeks to solve is unimportant.
Rather, I would interpret the mentions of microformats as an
indication that people are missing the wider problem RDFa would solve,
and focus on making that clearer, by talking about what RDFa does that
microformats don't even attempt to do.

Peace,
Scott

[Ben Ward]

On 25 Aug 2008, at 19:47, Manu Sporny wrote:

> There have been several threads discussing Microformats, RDFa and
> HTML5
> that are occurring on the WHATWG mailing list. The discussion
> relates to
> whether or not HTML5 should depend on the Microformats community to
> solve HTML5's semantic markup issues, or if both Microformats and RDFa
> should be considered for semantic web markup issues.

I've been out of touch with HTML5 development for a bit, but the way
you describe this paragraph is somewhat alarming.

We, the microformats community, absolutely *should not* be relied on
the fill every gap in HTML. That they would not specify minority
concerns in the HTML language is perfectly understandable, but the
Microformats Community is itself not designed to do that either. This
community, with this development process, is completely inappropriate
for filling every single extended use for HTML that people might have.

HOWEVER, there may just be misinterpretation here. Perhaps rather than
intending to depend on our specific community, the intention is that
the gaps be filled with ‘microformat-like patterns’. Patterns, class-
patterns, ‘posh’… whatever you want to call it. Microformats.org does
not own the class attribute and anyone working on techniques that are
incompatible with our process can do so.

It seems to me the case is not about ‘microformats.org’, but instead
about the capabilities of the class attribute itself. Is it just that
the word ‘microformats’ is being used as a generic catch-all for
semantic class name patterns?

It seems quite reasonable that the HTML working group be considering
the use case of ‘extended semantic description in HTML’ and
considering its existing capabilities (which are proving very capable
in the specific case of microformats), rather than a use case of
‘support RDFa in HTML’, which is just one solution.

I think Scott is correct in that you may need to reframe your
argument. Any push to have RDFa made a part of HTML5 should be focused
on the capabilities of RDFa compared to the class attribute, not the
(often intentional) limitations of one particular user of the class
attribute (us).

Ben

[Manu Sporny]

Scott Reynen wrote:
> I think comparing RDFa to microformats actually hurts your argument by
> suggesting they solve the same problem and reinforcing the notion that
> the wider problem RDFa seeks to solve is unimportant. Rather, I would
> interpret the mentions of microformats as an indication that people are
> missing the wider problem RDFa would solve, and focus on making that
> clearer, by talking about what RDFa does that microformats don't even
> attempt to do.

Scott, Ben - thanks for the feedback, both of you make some very good
points and I've adjusted my argumentation a bit to follow advice
expressed by both of you. Things are being clarified in some ways on the
HTML5 list and muddied in others.

The one thing that is clear is that most of those on the list are not as
up-to-speed with web semantics as either this community or the RDFa
community would expect. Certainly, I was a bit blind-sided by some of
the false assertions those on the list were making about semantics in
general.

The very long thread continues,

RDFa Problem Statement and Features
http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-August/015957.html

Intro to RDFa
http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-August/015974.html

RDFa markup consistency
http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-August/015992.html

CSS-based approach to semantic data on the Web (Microformats and RDFa)
http://lists.whatwg.org/pipermail/whatwg-whatwg.org/2008-August/015996.html

Of particular note is the last thread - the CSS-based approach to
semantic data markup. It's a proposal that, while interesting, ignores
the hard work that this community and the RDFa community has done over
the past several years. I could be mis-reading the various threads, so
some feedback from this list would be appreciated.

-- manu

--
Manu Sporny
President/CEO - Digital Bazaar, Inc.
blog: Bitmunk 3.0 Website Launches
http://blog.digitalbazaar.com/2008/07/03/bitmunk-3-website-launches

[André Luís]

Manu,
the css based approach is somethin that has come up in discussions
about semantics with fellow workers. I believe it does not trash all
of the hard work the communities have don so far. All it does, from
what i gathered, is move the semantics from html and places it in a
separate file/place. The vocabulary used could be one specified by
ufs. For instance: #tags a { rel: "tag"; }
it all comes down to: do we want to separate semantics from our markup?

thanks for the heads up on this matter.
Cheers,
André Luís

--
Sent from Gmail for mobile | mobile.google.com

[Manu Sporny]

André Luís wrote:
> Manu,
> the css based approach is somethin that has come up in discussions
> about semantics with fellow workers. I believe it does not trash all
> of the hard work the communities have don so far.

I never said the discussions "trashed" all of our hard work. I said that
some of the discussions "ignore" (some) of the hard work performed by
this community as well as the RDFa community.

> All it does, from
> what i gathered, is move the semantics from html and places it in a
> separate file/place.

Right - which both this community and the RDFa community are opposed to:

1. We do not want semantics to be placed in separate files.
2. We do not want vocabularies to be re-defined from site to site.
3. We want semantic markup to be easy to author for regular people - CSS
is /not/ easy to author.

That's what I was attempting to point out with my statement. Apologies
if I was not clear :)

-- manu

[André Luís]

I had misread you there, Manu. I apologize.

Thanks for clearing it up.

Those 3x bullet points are a great summary. Well done.

--
André