- From: Marcos Caceres <w3c@marcosc.com>
- Date: Wed, 14 Dec 2011 22:16:46 +0000
- To: Shane McCarron <shane@aptest.com>
- Cc: spec-prod@w3.org
On Wednesday, 14 December 2011 at 22:06, Shane McCarron wrote: > > > On 12/14/2011 4:02 PM, Marcos Caceres wrote: > > On Wednesday, December 14, 2011 at 9:11 PM, Robin Berjon wrote: > > > > The convention is simple: > > > > [FOO] always points to the latest version, i.e. for W3C that's /TR/foo/ > > [FOO-20120315] is the dated version, i.e. for W3C /TR/2012/WD-foo-20120315/ > > [FOO-20120315] should _never_ happen, unless you are doing something non-normative: "Because of screwups in [FOO-20120315], bla bla bla" > > > > I actually disagree. If 20120315 is a REC anyway. There are LOTS of > specs that need to reference a specific version of another spec. Look > at XHTML Modularization, for example. It references XML 1.0 Fourth > Edition even though there are later editions available. There were > important technical reasons for this. Are you saying that to implement XHTML Modularization I have to support whatever bugs are in XML Fourth Edition (forever), and I need to have a separate implementation of XML Fifth Edition (which does not interact with XHTML) for other XML documents? And that I can't ever update my XML Parser to XML.Next to be used with XHTML Modularization because XHTML forces me to implement Forth Edition? > Regardless, there needs to be a > way to do this. Normatively. > If the above is true, then that is pretty bad.
Received on Wednesday, 14 December 2011 22:17:22 UTC