[microformats-discuss] Visible metadata

Robert Bachmann rbach at rbach.priv.at
Sun Jul 24 18:03:54 PDT 2005

Ryan King wrote:
> As many have noticed, the <abbr> for date/time gets reused alot...so,
> I think it could stand to be refactored out into its own elemental
> microformat. What do you think?

I think every date/time on the web should be expressed using
<abbr title="YYYY-MM-DDTHH:MM">...</abbr>.

There is a great benefit because authors can use their favorite
date notation and users which are used to other conventions
could still understand them.

AFAIK in the US a common date/time format is MM/DD/YYYY hh:mm
(with 12 hours and am/pm)
In Austria a common format is DD.MM.YYYY hh:mm (with 24 hours)

So if I read "02/07/2005 11:30 pm" I have to engage my brain:
But with <abbr title="2004-02-07T22:30">02/07/2005 11:30 pm</abbr>, I
simply move the mouse cursor over it and my browser will display the
date in an unambiguous notation.

(Note: I intentionally used the long notation with dashes because it's
easier to read. I strongly recommend this to everybody, because title
attributes are not only for machines but for humans too)

But perhaps we could go even one step further.
Imagine a plugin, an Opera UserJavaScript or a Seamonkey script which
extends the mouse over
tool tip from
 Title: 2004-02-07T22:30
 Title: 2004-02-07T22:30
 Regional notation #1: 7.2.2004 22:30
 Regional notation #2: 7. Februar 2004 22:30

Ryan, you once pointed me to http://www.cl.cam.ac.uk/~mgk25/iso-time.html
and I learned that ISO8601 even allows you to specify your time zone, so
a plugin could even translate the time provided to the user's regional

>> The author(s) of a page could be denoted with
>>  <p>Authors:
>>  <span class="author">John Doe</span> and
>>  <span class="author">Winston Smith</span>
>>  </p>
>> or
>>  <p>Authors:
>>  <a href="http://example.org/~jd" class="author">John Doe</a> and
>>  <a href="http://example.org/~ws" class="author">Winston Smith</a>
>>  </p>
> Take a look at <address> and hcard again and make sure it can't
> supply your needs.
> http://www.w3.org/TR/REC-html40/struct/global.html#h-7.5.6
> http://microformats.org/wiki/hcard

Thanks Ryan, you brought me on the right way.
I already thought about <address> and I thought about hCard but
I didn't think about address _and_ hCard(s) ;-).

    <span class="vcard">
     <span class="fn">John Doe</span> and
    <span class="vcard">
     <span class="fn">Winston Smith</span>

    <span class="vcard">
     <a class="url fn"
        rel="alternate">John Doe</a></span>
    <span class="vcard">
     <a class="url fn"
        rel="alternate">Winston Smith</a>

Relationship rel="alternate" was added because http://example.org/~jd
and http://example.org/~ws should both contain hCards which provide
more contact details (email, phone, etc.).

One general note on hCard:

Parser could make the following assertion: If the container element
of a hCard is an inline element (e.g: <span>) and contains an <a>
element with class "url" and rel="alternate" _perhaps_ the page pointed
to by href contains the "full" version of the Person's hCard.

>> Using
>>  <p class="summary">...</p>
>> one could mark the summary, so that robots can extract it instead of
>> the description meta tag.
> I'm not sure about this.
> Do people use this meta tag?

Some people use it.
Goggle displays the contents of this meta tag in search results.

I've searched for "homepage" on http://www.google.com/search?q=homepage
The second result is "BBC - bbc.co.uk homepage - Home of the BBC on the
The text displayed below ("bbc.co.uk offers a varied range of sites
including news, sport, community, education, children's,
and lifestyle sites, with TV programme support, ...") comes from
the description meta tag.

> And do they publish the same content in html?
(I suppose you mean inside of <body> ...)
At least Jakob Nielsen does it with his articles on useit.com and
I guess others do it too. I'll write more on this topic later, when I
know more about the existing practice.

Robert Bachmann <rbach at rbach.priv.at> (OpenPGP KeyID: 0x4A5CCF10)

More information about the microformats-discuss mailing list