W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 2009

Re: Comments on the HTTP Sec-From Header (draft-abarth-origin)

From: Ian Hickson <ian@hixie.ch>
Date: Thu, 9 Jul 2009 07:06:34 +0000 (UTC)
To: Adam Barth <w3c@adambarth.com>
Cc: Mark Nottingham <mnot@mnot.net>, collinj@cs.stanford.edu, HTTP Working Group <ietf-http-wg@w3.org>
Message-ID: <Pine.LNX.4.62.0907090658520.1976@hixie.dreamhostps.com>
On Wed, 8 Jul 2009, Adam Barth wrote:
> 
> I'm a bit unclear on the difference between URIs and IRIs.  Is there are 
> a good reference you could point me to?

You probably want to just refer to DanC's Web Addresses draft (since I 
believe we need to be able to handle non-conforming URIs here too, e.g. 
the page "http://www.example.com/%" still has an origin, but that's 
neither a URI nor an IRI), but I hear there's work to merge that with 
IRIs, so it's unclear exactly what the state is of that. I'm not sure what 
I'm going to do with HTML5 regarding this -- I guess if this isn't 
resolved by October (last call) I'll just temporarily put the definitions 
back into HTML5, just so we have a clear place that defines the parsing.


> > * The /variable/ syntax is distracting.
> 
> Is there a standard way of distinguishing variables from non-variables?  
> In non-ASCII specs, I've seen italics used for this purpose.

(If you find an answer to this please let me know; I've got the same 
problem in the WebSocket I-D.)


> > * Specifying RFC2119 requirements as "Implementations MUST use the 
> > following algorithm..." isn't good practice, as discussed previously. 
> > E.g., section 2 would be better specified as:
> 
> The preference for one or the other seems to vary by community.  The 
> algorithmic parts of the spec originated from HTML 5, which is largely 
> written in that style.  I can change to the declarative style if Ian 
> doesn't think that will cause a problem for HTML 5.

If you use the style "the algorithm is: 1. do this. 2. do that." I can 
invoke it with a "must" from HTML5 to get the right normative conformance 
criteria for testing purposes, so if this helps you resolve issues and 
move the draft forward, go for it.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'
Received on Thursday, 9 July 2009 07:09:24 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:51:07 GMT