Re: inv-maxage Cache-Control directive syntax, was: Fwd: I-D Action: draft-nottingham-linked-cache-inv-00.txt

On 14/06/2012, at 10:12 PM, Julian Reschke wrote:

> On 2011-05-28 02:56, Mark Nottingham wrote:
>> FYI. This is an individual draft, not a WG effort. Nevertheless, feedback welcome.
>> 
>> Regards,
>> ...
> 
> In <http://tools.ietf.org/html/draft-nottingham-linked-cache-inv-02#section-5>, it would be good to define the edge parsing edge cases, such as:
> 
> - multiple instances of the directive,
> 
> - missing argument,
> 
> - use of quoted-string.

Agreed.

> I would propose to say that:
> 
> - the parameter value needs to follow the delta-seconds ABNF after applying generic C-D parsing,

C-D?

> - that a missing argument is to be treated as empty string (thus invalid), and

Yes

> - that specifying multiple instances makes the Cache-Control header invalid with respect to this directive (thus the header field should be processed as if inv-maxage wasn't specified at all).

yes

> 
> Optimally, including examples :-)

As always

Thanks,

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