W3C home > Mailing lists > Public > www-tag@w3.org > June 2013

Re: References and Modularity

From: Marcos Caceres <w3c@marcosc.com>
Date: Tue, 4 Jun 2013 16:40:21 +0100
To: Karl Dubost <karl@la-grange.net>
Cc: Anne van Kesteren <annevk@annevk.nl>, Larry Masinter <masinter@adobe.com>, "L. David Baron" <dbaron@dbaron.org>, Robin Berjon <robin@w3.org>, "www-tag@w3.org" <www-tag@w3.org>
Message-ID: <A633D10D56D140B8B3CF1830666ECC37@marcosc.com>

On Tuesday, June 4, 2013 at 3:58 PM, Karl Dubost wrote:

> hehe. You are contradicting yourself ;)

nah! I just confused you with my infallible logic :)  
> The latest version:
> Le 4 juin 2013 à 19:21, Marcos Caceres a écrit :
> > whatwg.org/c (http://whatwg.org/c)
> A specific version:
> > https://dvcs.w3.org/hg/html-proposals/raw-file/9443de7ff65f/responsive-images/responsive-images.html
> The latest [frozen] version:
> http://www.w3.org/TR/html5/
> A specific version:
> http://www.w3.org/TR/2012/CR-html5-20121217/
> > So, as HTML5 shows, a process that relies on dated versions is a broken process.
> No :) As shown just above. The process doesn't rely on dated versions. These are just identifiers. The current W3C Process relies on versions with no latest or "now" identifier.
Well, that's supposed to be the "Latest Editor's Draft:" which is "http://www.w3.org/html/wg/drafts/html/CR/" according to HTML5 … but it's a broken link… yeah, it's broken as in 4-ooh-4 :)  

So, I rest my case.  
> The issue is the meaning of the latest version, not dated versions.
> The issue is NOT the choice of identifiers (dated/hash/whatever unique id).
> Suggestions: Add a link to the NOW (real latest) version of a technology.
> Some of the schemes, choices made in the past which were logical at the time and make less sense now.
> * http://www.w3.org/TR/html/ -> link to XHTML1.0 instead of http://www.w3.org/html/wg/drafts/html/master/
> * http://www.w3.org/TR/svg/ -> link to SVG 1.1 instead of https://svgwg.org/svg2-draft/
> * etc.
> These could be fixed. (Note that things become anyway hairy when the technology is being split into modules in the future. Example CSS.)
Sure. And yeah, we should probably take back http://www.w3.org/TR/html/ … though that would probably cause issues now (unless we deprecate XHTML once and for all).  
> > This problem seems to stem from the W3C process being model on a pre-digital-publishing standardization model (i.e., from a kind of "nuts-and-bolts" standardization process, where standardized products remain static and don't get changed after production). That doesn't apply to us.  
> Here there's another confusion in between a document which is always versioned, an identifier, and a UX. What we are really discussing is not related to the standard itself, but to the user experience (UX) of discovering the latest (now) version of a technology.
> Each time I go to http://www.yr.no/place/Canada/Quebec/Montreal/time_for_time_detaljert.html
> I expect to get the weather of now, and it's fine, it is the semantics of the identifier.  
> (note also the clever: "Updated at 3:02. Next update around 16:00." aka setting expectations)
> UX. UX. UX. ;)
> Here another one, the latest version (now):
> http://weather.gc.ca/city/pages/qc-147_metric_e.html
> And the dated access:
> But you have also access to, for example, "Hourly Data Report for April 07, 1994"
> http://www.climate.weatheroffice.gc.ca/climateData/hourlydata_e.html?StationID=5415&timeframe=1&Month=4&Year=1994&cmdB1=Go&Day=7

Sure, that's the same as "9443de7ff65f". It's not a contradiction to have a version history.  
> If in the specs, you had a link to the latest version you solved most of your issues with regards to implementers. :)

Ha, if only. As long as there continues to be multiple versions of the same document, bad things will continue to happen (like the 404 for HTML5). If there was one document, then we wouldn't need to have this discussion.   
Received on Tuesday, 4 June 2013 15:40:54 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:33:20 UTC