W3C home > Mailing lists > Public > ietf-http-wg@w3.org > April to June 2010

Re: How many coding registries should we have? (Issue 143), was: Issue 189, Re: Content/Transfer-Codings organization/IANA considerations (issue 143, 188, 189)

From: Mark Nottingham <mnot@mnot.net>
Date: Fri, 9 Apr 2010 13:10:59 +1000
Cc: HTTP Working Group <ietf-http-wg@w3.org>
Message-Id: <34617E1E-A8DB-4D5B-9FDB-FFF13804068F@mnot.net>
To: Julian Reschke <julian.reschke@gmx.de>
Personally, I like how IANA has it set up now; a little bit of duplication never hurt anyone, and it avoids the "oh, I didn't notice that column" defence.

Really, though, as long as it's clear where each can be used, I don't have a problem.



On 09/04/2010, at 12:06 AM, Julian Reschke wrote:

> On 07.08.2009 18:59, Julian Reschke wrote:
>> ...
>> So they really should be defined separately from Transfer-Coding and
>> Content-Coding, and be collected in the same registry (surprise: IANA
>> already has both in "http-parameters").
>> 
>> Feedback appreciated...
>> ...
> 
> As far as I can tell, this is the one remaining issue related to ticket 143 (<http://trac.tools.ietf.org/wg/httpbis/trac/ticket/143>).
> 
> The current status is:
> 
> - IANA has two registries (transfer coding and content coding), which *both* are maintained at <http://www.iana.org/assignments/http-parameters>
> 
> - Part 1 defines Transfer Codings, the Transfer Coding registry, and updates the transfer coding registrations
> 
> - Part 3 defines Content Codings (referring to Part 1 for the compression codings), the Content Coding registry, and updates the content coding registrations
> 
> - Both registries have been clarified/modified to require spec & expert review
> 
> The open question is:
> 
> - Do these two registries share the same namespace? That is, is it allowed for a coding named X to be both a transfer and a content coding, and have incompatible definitions?
> 
> I think this would be a very bad idea, so we really should unify the registries to a single coding registry, and then have new entries state as which type (transfer/content) they can be used. My assumption would be that any transfer coding could be a content coding, but that the opposite is not true.
> 
> Feedback appreciated,
> 
> Julain
> 
> 
> 


--
Mark Nottingham     http://www.mnot.net/
Received on Friday, 9 April 2010 03:11:31 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:51:18 GMT