W3C home > Mailing lists > Public > www-html@w3.org > December 2003

Re: a recommendation - Received on Thursday, 18 April 2002 14:01:42 EDT

From: Christian Wolfgang Hujer <Christian.Hujer@itcqis.com>
Date: Wed, 3 Dec 2003 17:35:03 +0100
To: "Fastpitch Central - Bill" <bill@fastpitchcentral.com>, <ernestcline@mindspring.com>
Cc: <www-html@w3.org>
Message-Id: <200312031735.04610.Christian.Hujer@itcqis.com>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,


Am Mittwoch, 3. Dezember 2003 16:18 schrieb Fastpitch Central - Bill:
> You said, "xml:base does not provide this level of redundancy.  "  But,
> even if it did I would still stand by my suggestion.  There are hundreds of
> millions of "simple" websites that use <base . . . > and will likely use
> <base . . . > for decades.
okay, I take a look in my glass sphere, I concentrate, I see ... a pink clowed 
surrounded by green mist ... I concentrate ... ssshhhh, be quiet there! ... I 
concentrate ... I see... I see the future ... I know the answer: No.
a) It's not so many websites that use <base/>
b) Future versions of XHTML won't have <base/>.
c) Microsoft finally publishes an XHTML 2.0 capable web browser in 2008, 4 
years later than KDE, Mozilla.org and Opera, published theirs, to keep their 
user agent's market share from dropping below the 10% mark. (Yes, they're in 
the same position and situation as Netscape was in their best times)
d) Microsoft will not use KHTML or Gecko.
http://www.gemal.dk/archives/000040.html

That's not a decade ;-)

Of course it takes time until existing sites are updated. But how many HTML or 
HTML 2.0 sites are there?
Most are tag soup based HTML 4.0.
Very few but more and more are XHTML.
All my sites are XHTML 1.1 already (with a content negotiating variant in HTML 
4.01 for outdated user agents like the current versions of Internet Explorer, 
links, lynx and w3m).
And believe me, as soon as there's a Content-Type and a User Agent for XHTML 
2.0, I will use it.


> Personally I love XML when its used for RSS type functions.  I've also
> devised my own XML tags for sports before SportsML was devised.  But,
> "xml:base . . . " has never been needed in anything I've done so far.  It's
> probably a wonderful thing - but I have no clue about it's use or value.
Here's a list of markup languages that use XML Base / @xml:base I know so far:
- - XHTML 2.0, where @xml:base replaces <base/>
http://www.w3.org/TR/xhtml2/
- - XML Inclusions (XInclude)
http://www.w3.org/TR/xinclude/
- - XForms
http://www.w3.org/TR/xforms/
- - SOAP 1.2
http://www.w3.org/TR/soap12-part0/
- - SVG 1.0 and successors
http://www.w3.org/TR/2001/REC-SVG-20010904/
http://www.w3.org/TR/SVG11/
http://www.w3.org/TR/SVG12/
- - SMIL 2.0
http://www.w3.org/TR/smil20/
- - XML Linking Language (XLink) Version 1.0
http://www.w3.org/TR/xlink/
- - XSLT 1.1
http://www.w3.org/TR/xslt11/
(discontinued, replaced by XSLT 2.0)
- - XSLT 2.0
http://www.w3.org/TR/xslt20/

Need more? ;-)
And yes, XML Base is wonderful.

> What I do recall is DNS failed and immediately the thousands of web pages
> on my own little website were unavailable.  During that short period of
> time I directed several people to my website via the dotnum.  But, it was
> impossible to walk them through more than a single page since no links
> worked.
That's because of <base/>, yes.

Not a single one of mine uses <base/>. What should I use it for? Well, 
sometimes it /might/ be useful on subsites, but only if the server 
automatically sets it (via Perl, PHP, XSLT, Servlets, JSP or whatever).
Is it neccessary for reducing the document size? No. HTTP and all user agents 
I've tested for this so far (including netscape, mozilla, ie, opera, 
konqueror, lynx, links, w3m) support the transmission of compressed data.


> If my suggestion for the dotnum addition to the <base . . . > tag was
> implemented, all links would have worked.  As I did before, I could give
> out the dotnum to get customers and prospects started.
a) I agree with those saying it's not the task of HTML to cope with flaws of 
lower level protocols and systems like DNS.
b) Extending <base/> alone doesn't solve the problem anyway, because a really 
huge number websites with several different host / domain names share the 
same ip anyway.
This is called virtual hosting. Hosts are chosen from the domain name provided 
in the Host: header of the HTTP Request.

So, see the impact it has, what you desire might also require changes in HTTP 
to work.

FTP and all other protocols share the same problem, still, the nature of the 
problem lies in DNS.

A imho far better because more generic solution would be a redundant DNS 
system that makes sure that if the main DNS system fails, a backup system is 
available.

Anyway, you'd really better suggest it to those responsible for XML Base:
It's not related to HTML anymore. The <base/> element has been replaced by a 
@xml:base attribute shared by many specifications.

This list is the wrong place to discuss this.
The problem has ceased to be a problem of HTML.

> [cut: 9/11 = 1984]


Bye
- -- 
ITCQIS GmbH
Christian Wolfgang Hujer
Geschäftsführender Gesellschafter (Shareholding CEO)
E-Mail: Christian.Hujer@itcqis.com
WWW: http://www.itcqis.com/
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2-rc1-SuSE (GNU/Linux)

iD8DBQE/zhC3zu6h7O/MKZkRApMlAJ9ZwCUeN04nC274y5VJrq+CZbBezwCeNrdd
kHLHFBNbLWgQs3suqbX+67k=
=3Tnl
-----END PGP SIGNATURE-----
Received on Wednesday, 3 December 2003 13:23:09 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 March 2012 18:15:59 GMT