Re: Proposal for i23: no-store invalidation

tor 2009-06-11 klockan 09:00 +1000 skrev Mark Nottingham:

> ... I think this issue is related in no small way:
>    http://trac.tools.ietf.org/wg/httpbis/trac/ticket/110

Indeed, if one want to make a general blanket rule independent of method
etc..

But not for a rule restricted to successful responses to
GET/HEAD(/POST).

> Yes; as mentioned before, the concept of "cache correctness" was  
> dropped from p6,

Sorry, been a little too much in the shadows lately...

> because it didn't add any binding requirements, and  
> it was fuzzy at best anyway. If you think something was lost, please  
> say so...

Not sure anything was lost. As you say it was very fuzzy.. better make
something that makes sense.

> Which requirements are you referring to when you say "caching of  
> negotiated resources sections"?

2.6 last paragraph, about deleting obviously old cache copies. Already
have a note to discuss if it's at all needed...

Regards
Henrik