W3C home > Mailing lists > Public > www-html@w3.org > January 2000

Re: ISO/IEC refs to HTML4

From: Dan Connolly <connolly@w3.org>
Date: Tue, 18 Jan 2000 21:26:25 -0600
Message-ID: <38852EE1.75A27300@w3.org>
To: Murray Altheim <altheim@eng.sun.com>
CC: Jan Roland Eriksson <jrexon@newsguy.com>, www-html@w3.org
Murray Altheim wrote:
> Jan Roland Eriksson wrote:
> >
> > I have had reason to browse through the ISO/IEC Standard for HTML4...
> [...]
> > My question is; is it correct of the W3 server to redirect a request for
> > the older HTML4 spec to the latest HTML4.01 in this case (or in any case
> > for that matter)?
> You ought not get me started on this one. Just as the W3C posts their
> philosophical rant "Cool URIs don't change"

i.e. http://www.w3.org/Provider/Style/URI

> they pull this number, which
> runs completely counter to it.

You're going to run us over the coals for a mistake that was cleared
up after a few days?

How does this run counter to "Cool URIs don't change"?

Since 18-Dec-1997, to this day, the URI
http://www.w3.org/TR/REC-html40 has meant
	"whatever W3C publishes as the latest HTML 4.0 specification"

If you use that address to cite HTML 4.0, you mean for your
readers to get updates when W3C publishes a new HTML 4.0 spec.

just like Yahoo gives the current austin weather at

If you don't want updates, cite the "this version" address,
i.e. http://www.w3.org/TR/REC-html40-971218 for the 18 Dec 1997

As of 18-Dec-1997, the content of http://www.w3.org/TR/REC-html40
was the same as the content of
http://www.w3.org/TR/REC-html40-971218 . But on 24-Apr-1998, we
revised the HTML 4.0 spec; the new (archival) publication was
http://www.w3.org/TR/1998/REC-html40-19980424 and
http://www.w3.org/TR/REC-html40 was updated to match.

Again, on 24 December 1999, HTML 4.0 was updated; the new archival
document was: http://www.w3.org/TR/1999/REC-html401-19991224/
and http://www.w3.org/TR/REC-html40 was updated to match.

Unfortunately, it was done in such a way that some
http://www.w3.org/TR/REC-html40/* addresses got broken in a
pre-holiday rush to publish a bunch of stuff. This has been
fixed now.

CRAP! No it hasn't!
is still broken!

Sorry... I'll get that fixed ASAP.

> Very confusing for everyone, and in my
> opinion just plain wrongheaded. If TimBL is correct and names and addresses
> are the same thing, then a URI should be the equivalent of a name. Obviously,
> they are making the case against their own argument.

It's not obvious to me at all. Please explain.

Dan Connolly
Received on Tuesday, 18 January 2000 22:27:01 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:05:52 UTC