[Bug 11562] address tag definition has no relation to it's name

http://www.w3.org/Bugs/Public/show_bug.cgi?id=11562

Benjamin Hawkes-Lewis <bhawkeslewis@googlemail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mike@w3.org
          Component|HTML5: The Markup Language  |HTML5 spec (editor: Ian
                   |(editor: Michael(tm) Smith) |Hickson)
         AssignedTo|mike@w3.org                 |ian@hixie.ch

--- Comment #3 from Benjamin Hawkes-Lewis <bhawkeslewis@googlemail.com> 2010-12-16 18:17:17 UTC ---
(In reply to comment #2)
> > > 1. HTML needs a tag to specify a location, as this is a common meaningful
> > > information.
> > 
> > Why?
> 
> as I said - it seems like a very common information type, that IMO should be
> markupable.

Verb and noun and sentence are even more common units of information and we
don't have markup to distinguish those.

You need to work out what problem you're trying to solve, before we can
consider possible solutions such as changing the meaning of "address":

http://wiki.whatwg.org/wiki/FAQ#Is_there_a_process_for_adding_new_features_to_a_specification.3F

Please note that HTML already includes a native way to unambiguously mark up
locations using hyperlinks:

For example:

<a href="http://example.com">Destination description</a>

<a href="mailto:somebody@example.com">somebody@example.com</a>

<a href="tel:+358-555-1234567">Tel: +358-555-1234567</a>

<a href="fax:+358.555.1234567">Fax: +358.555.1234567</a>

<a href="geo:48.2010,16.3695,183">Vienna, Austria</a>

See also:

   * http://tools.ietf.org/html/rfc6068 for mailto

   * http://tools.ietf.org/html/rfc5341 for tel and fax

   * http://tools.ietf.org/html/rfc5870 for geo

> both MF and MD are not official, and thus are not as reliable as
> a true markup tag that is standard, and they are not a part of HTML. It feels
> strange that such a generic content would need to use "outside" markup.

Microdata and HTML+RDFa (another option for annotating HTML5 with additional
semantics) have the same "standard" status as HTML5: they are on track to
becoming W3C Recommendations:

http://www.w3.org/TR/microdata/

http://www.w3.org/TR/rdfa-in-html/

I think it's ironic that you are arguing we should *change* the definition of
the "address" element from author contact information - the definition it has
had since 1993 - to simply mean any old address on the basis that HTML is a
reliable standard!

http://www.w3.org/MarkUp/draft-ietf-iiir-html-01.txt

http://www.w3.org/MarkUp/html-spec/html-spec_5.html#SEC5.5.3

http://www.w3.org/TR/html401/struct/global.html#h-7.5.6

> > What is the ultimate difference between "contact information" and "an address
> > or location, virtual or physical", in your view?
> 
> In most discussions I've participated on the subject, such markup as:
> <li class='game'>
>    <h3>Some1 vs Some2</h3>
>   <address>Some city</address>
> </li>
> was considered off the spec.

Correctly so. That is not author contact information, it's just a location.

> > The key is not that it is "contact information" but that it is a contact
> > information for an author responsible for the document or section.
> 
> first of all, the new specs say only "The address element represents contact
> information." http://dev.w3.org/html5/markup/address.html

First, read down on that page and you'll see it clarifies:

"If an address element applies to a body element, then it represents contact
information for the document as a whole. If an address element applies to a
section of a document, then it represents contact information for that section
only."

Second the "new specs" you're quoting are intended as a mere "non-normative
reference".

http://dev.w3.org/html5/markup/Overview.html#toc

The actual normative specification is clearer:

"The address element represents the contact information for its nearest article
or body element ancestor. If that is the body element, then the contact
information applies to the document as a whole."

http://dev.w3.org/html5/spec/sections.html#the-address-element

Since you're proposing a change to the normative specification, I'll change the
Component field so that your bug gets put into the queue of the correct editor,
if that's okay. :)

> 2nd of all - why is that type of information more generic and more suited
> than a generic address

It's more suited since that has *always* been the definition of "address".

> As I see it, it makes more sense that the tag should be used as a generic
> address/contact info, rather than a very specific, less useful "contact the
> owner of the document" tag (which can be expressed by many other means, such
> as title, rel and other MF/MD which are for special cases scenarios).  IMO,
> HTML should be as generic as possible, as to allow a large set of different
> valid markups..

Precisely because HTML is supposed to be a reliable standard, we should be wary
of arbitrarily changing the semantics of its elements and attributes.

The name of an element/attribute alone is a very poor reason to change its
semantics.

> > > As a side note, I have seen the tag used as a location indicator on many
> > > accounts,
> > 
> > Do you have data that it is used to mean "an address or location, virtual
> > or physical" more often than contact information for authors?
> 
> Yes, as you can see with the example above. 

No, that's an example of it being used incorrectly. It's not data showing that
it is used incorrectly more often than correctly in the web corpus.
Incidentally, it wouldn't particularly surprise me if it were used more often
incorrectly, but we should be careful to make decisions based on actual data.

> > > and have seen search engines recognize it as such.
> > 
> > Which search engines? Can you prove that they are recognizing it as a "an
> > address or location, virtual or physical" as opposed to author contact
> > information? In particular, how do you know they are recognizing the
> > element rather than the contents of the element (i.e. just picking up on
> > text that looks like a postcode or whatever)?
> 
> Obviously I have no such proof. It is as likely that they simply extract text
> structures and analyze the raw text. But I have seen the tag used for other
> situations (as mentioned above) and indexed properly throughout the years. In
> fact, I only recently found out that I have been using it off the specs...

I think we can conclude the claim "search engines recognize it as such" is
unsubstantiated and should not be taken into account.

-- 
Configure bugmail: http://www.w3.org/Bugs/Public/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug.

Received on Thursday, 16 December 2010 18:17:19 UTC