W3C home > Mailing lists > Public > public-vocabs@w3.org > May 2012

Re: setting multiple properties: name and location in a href?

From: Stéphane Corlosquet <scorlosquet@gmail.com>
Date: Thu, 31 May 2012 11:01:54 -0400
Message-ID: <CAGR+nnGFioAQHMWLptO-L=B+gErS=T5TSWPQ3SmdomC9hozDXw@mail.gmail.com>
To: Michael Below <below@judiz.de>
Cc: public-vocabs@w3.org
Hi Michael,

On Thu, May 31, 2012 at 10:21 AM, Michael Below <below@judiz.de> wrote:

> I just noticed another difference between Google and Bing and I am
> wondering how to do it right:
> I have a list of old blog entries on an "archive" page of my blog. I
> described that page in microdata as a Blog, and I am trying to integrate
> the microdata for the individual entries like this:
> <li itemscope itemtype="http://schema.org/BlogPosting"><a
> itemprop="contentLocation name headline"
> href="/aktuelles/atomhaftung/">Haftung für Atomunfälle: Wer soll das
> bezahlen?</a>
> <p class="blogdate">Geschrieben am <time itemprop="dateCreated"
> datetime="2011-03-30 14:41" pubdate>30.03.2011</time></p></li>
> According to Google, I have set all properties to a text value and to a
> href value both (which seems fine). According to Bing, I have set it all
> to the href values.

Bing is right and the microdata spec is clear on this: an itemprop
attribute in an 'a' element will target the href, and the href only. so all
three properties that you are using have the href "/aktuelles/atomhaftung/"
as value. To markup the text of a link, you need to add an extra span
element inside that link to explicitly markup the text, like this:

<li itemscope itemtype="http://schema.org/BlogPosting">
  <a itemprop="contentLocation name
headline" href="/aktuelles/atomhaftung/">
    <span itemprop="headline">Haftung für Atomunfälle: Wer soll
das bezahlen?</span>
  <p class="blogdate">Geschrieben am <time itemprop="dateCreated"
datetime="2011-03-30 14:41" pubdate>30.03.2011</time>

properties in @itemprop should only apply to the href, the other should go
in the @itemprop of the inner span.

The extracted rich snippet data that Google shows is misleading here, it
extracts more data than it should.

Received on Thursday, 31 May 2012 15:02:49 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:29:23 UTC