Re: i15: How to tell if a client does not support persistent connections

See <http://www.w3.org/Protocols/HTTP/1.1/rfc2616bis/issues/#i15>;  
there certainly isn't a brand-new MUST level requirement here (as  
there was a blanket requirement previously), and I don't see how an  
existing client could suddenly become non-compliant as a result of  
only this change.

Even if it were to, there is wiggle room enough in the charter to  
clarify instances where interop isn't possible, without changing the  
version number.

Cheers,


On 15/12/2007, at 10:45 AM, Paul Leach wrote:

>
> If adding this as a MUST can make existing compliant HTTP/1.1  
> clients non-compliant, then I think that it's outside what the WG  
> charter allows (and logically would require a new minor version of  
> HTTP, would it not?).
>
> -----Original Message-----
> From: ietf-http-wg-request@w3.org [mailto:ietf-http-wg- 
> request@w3.org] On Behalf Of Henrik Nordstrom
> Sent: Thursday, December 13, 2007 6:16 AM
> To: Lisa Dusseault
> Cc: HTTP Working Group
>
> On ons, 2007-12-12 at 18:48 -0800, Lisa Dusseault wrote:
>> Issue i15 is closed, and the resolution suggests new text, in part:
>>
>>      "An HTTP/1.1 client that does not support persistent  
>> connections MUST
>> include the "close" connection option in every request message. "
>
>> From what I understand it's a MUST to align the rules with server  
>> side.
>
> But it's true that it doesn't need to be a MUST on the client side.  
> A SHOULD is quite sufficient.
>
> Regards
> Henrik
>


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

Received on Friday, 21 December 2007 02:01:13 UTC