Re: Issue 27 and IETF Prague

Ian,

On 23/03/2011, at 6:36 AM, Ian Hickson wrote:
[...]

> If we're going to be moving deadlines around, it may also be worth 
> allowing an additional change proposal to be submitted. One of the ideas 
> that has been discussed recently (and which to me sounds rather better 
> than any of the proposals on the table) is to have the spec be the main 
> registry, with an openly-editable registry being used just to queue up 
> proposals in a way that avoids name-clashes. This would more closely align 
> with how things have been done in practice -- we need to handle the case 
> of proposals being made but never implemented or used (or even sometimes 
> specified), which just clutter up the registry and theoretically prevent 
> others using the same keyword for other uses; we need to handle the 
> judgement of the chairs that a keyword should be removed from the spec, 
> which currently is essentially meaningless since the value can just as 
> easily be registered; and we need to handle the issue of implementors 
> having good guidance for all well-known values, something which the 
> current registries fail to do since they don't have anyone responsible 
> for making sure the specs are worth anything.


I agree that's a concern. 

Next week, we're going to continue discussion among IETF folks and parts of the W3C community about how the various Web-related registries work, with the aim of addressing concerns we've seen -- including those presented by the HTML5 community, as well as issues we've observed separately. 

Issues we've talked about so far include:
  - Expert Reviewers placing too high a bar on entry
  - SLA for registration
  - Lack of clarity around points of contact
  - Lack of transparency into a request's status
  - Need for clear ways to update and deprecate registered values
  - Need to link to multiple specifications for format/appliation-specific requirements
  - Ease of use for non-insiders; e.g., clear and available guides, etc.

If anyone would like to attend and will be in Prague, please ping me. So far the discussion has been informal, but expect there will be proposals for concrete changes circulated soon, so that everyone else can weigh in.

What I tried to express to Philippe was that this process is going to take *some* time, and the timing of the poll was therefore pretty unfortunate. 

Observation -- your proposal above makes sense, in that the impact of a new value upon HTML's processing model always needs to be described *somewhere." It sounds like the IANA registry *could* be the mechanism that prevents name clashes (its primary function anyway), with the HTML WG pulling candidates for elevation from that.

Regards,

--
Mark Nottingham   http://www.mnot.net/

Received on Tuesday, 22 March 2011 22:52:04 UTC