W3C home > Mailing lists > Public > www-style@w3.org > February 2006

Re: [BULK] - Re: [XHTML2] Spirit of "1.1.3. XHTML 2 and Presentation" (PR#7759)

From: Laurens Holst <lholst@students.cs.uu.nl>
Date: Mon, 20 Feb 2006 21:24:35 +0100
Message-ID: <43FA2583.3020803@students.cs.uu.nl>
To: Paul Mitchell <paul@paul-mitchell.me.uk>
Cc: "www-style@w3.org" <www-style@w3.org>

Paul Mitchell schreef:
> As far as I know, I'm describing how two things have been done in HTML 
> for a long time, and proposing extending that to XML and all 
> derivative tagged document formats, no exceptions, forever, as a good 
> idea.

Yes, but XML is not HTML, fortunately.

You’re proposing quite a change from the current practice (think e.g. 
xml:id, which also wasn’t implemented as simply id). Don’t you think 
your original wish (to be able to generically add style and script to a 
document) will have more chance of success if you *don’t* propose 
something so radical and /backwards incompatible/?

>> Namespaces exist for a reason, that is to prevent naming conflicts, 
>> and your solution would make life a lot harder for everyone who is 
>> using the ‘global namespace’ (I assume with this you mean the empty 
>> namespace?).
> Yes - the empty namespace. The one place where everyone has to agree 
> with each other for things to work properly, and the very reason 
> namespaces exist, for the many and various XML applications to avoid 
> unnecessary conflict in that domain.

No, the empty namespace is often used by applications in closed 
environments which have no intention of combining multiple technologies. 
It is used very frequently in e.g. configuration files. XML does not 
require namespaces in any way, after all. That said, the whole idea of 
an empty (no) namespace is that you don’t consider any other 
applications. That makes the empty namespace in no way global, in the 
contrary, it makes it local!

If I’d use e.g. the Internet Explorer’s XML engine for parsing a 
configuration file of an application of mine, which uses <style> to say, 
designate the style of my skin, and Microsoft would make such a change 
to have your suggestion work, my application could stop working!

And what’s next? <base>? <root>? Where does it end? If this goes on, it 
will never be possible to make a future-proof XML document in the empty 
namespace!

And all this while all prefixes starting with ‘xml…’ were reserved 
*especially* for the purpose of having a global namespace that can apply 
to all documents.

> Some things are universal, and deserve a special place in the world. 
> Style and script are universal to browsers, so I just accept that HTML 
> was right first time, at least in terms of what the respective 
> elements were called, what they represented, how people understand 
> those things and how browsers implement them already.
>> It will also cause the feature not to be implemented by implementors, 
>> in fear of breaking existing content.
> If the standard were something like "unless explicitly specified 
> otherwise, in all tagged documents, in all contexts, <style> and 
> <script> elements are *browser directives* and *not* to be considered 
> part of document content", don't you think implementors /everywhere/ 
> would cheer and get busy adapting, designing and coding?

WHY? Why is writing say, xml:style and xml:script SO HORRIBLE that 
current practice until now, good coding standards and backwards 
compatibility have to go down the drain?

And if it isn’t bad enough, your additional statement “*not* to be 
considered part of document content” will make it even more 
backwards-incompatible.

Do you also realise that if it is in the empty namespace, these 
attributes would in many cases need to always have a namespace 
declaration on them? <style xmlns=""></style> is more of a bother to 
write than <xml:style></xml:style> (or let’s consider <empty:style 
xmlns:empty=""></empty:style>). How would this be convenient?

And why would implementors ‘cheer’ for that? Do you really think that 
implementors would want to risk breaking existing content while they 
could have it work by simply matching on ‘xml:style’ instead of 
‘"":style’? And why would they like it any more than say, I do?

I’m sorry, I really can’t comprehend this line of thought, and frankly, 
don’t really care about discussing this further. It just makes too 
little sense.


~Grauw

-- 
Ushiko-san! Kimi wa doushite, Ushiko-san!!
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Laurens Holst, student, university of Utrecht, the Netherlands.
Website: www.grauw.nl. Backbase employee; www.backbase.com.
Received on Monday, 20 February 2006 22:58:13 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:43 GMT