[uf-discuss] Signalling use of microformats?

Thu Feb 9 10:46:55 PST 2006

On 2/9/06 7:24 AM, "David Osolkowski" <qidydl at gmail.com> wrote:

> On 2/9/06, Angus McIntyre <angus at pobox.com> wrote:
>> 
>> Is there a way to indicate - for the benefit of robots, for example -
>> that a given web page contains markup that conforms to one or more
>> microformats? I'm imagining something along the lines of:
>> 
>>         <meta name="microformats" content="tag hAtom xFolk" />
> 
> Ironically, this is exactly the sort of thing microformats are against
> using--it's not human-readable.

On its surface this is a true statement, and while the *proper* way to
indicate presence of microformats is also invisible:

 <head profile="microformatProfileURL1 microformatProfileURL2 ...">

...we are exploring alternatives which would allow these declarations to be
visible with rel="profile" hyperlinks.

> This means it's likely to be out of
> date, even when it's present.

Not necessarily.  The "profile" attribute is essentially the microformat
equivalent of the DOCTYPE at the top of the document.  That's invisible, and
yet we still include it.  Why?

To create valid web pages, that's why.

In the same way, I predict that we will in the not too distant future have
an XMDP-aware validator, that reads the <head profile> parses those XMDPs
and uses them to validate the microformats present in the document.

Thus just like an out of date DTD would be detected by the W3C validator, an
out of date profile list would be detected by an XMDP-aware validator.

> Leave it to the code to determine if a
> page uses microformats, instead of putting the burden on the author.

Certainly a lot of code will do this, for quite some time, just as HTML
parsers (browsers) ignored DTDs for quite some time.  But eventually, just
as all modern browsers now pay close attention to the DTD, I believe we will
see the same with the profile list, though it may take several months or
years.

Thanks,

Tantek