Re: [Ipsec] Discrepency RFC4301 and RFC4305

At 8:04 PM -0800 1/8/06, Vishwas Manral wrote:
>Content-class: urn:content-classes:message
>Content-Type: multipart/alternative;
>	boundary="----_=_NextPart_001_01C614D1.CA0C5D20"
>
>Hi,
>
>I had brought out the issue more then a year back that:
>
>RFC4301 states
>             - confidentiality-only (MAY be supported)
>             - integrity only (MUST be supported)
>             - confidentiality and integrity (MUST be supported)
>
>However RFC4305 states that NULL authentication support is a MUST.
>
>I had brought out the issue with the draft which became RFC4305. 
>Stephen Kent had supported the change and stated
>"since we changed the requirements for encryption-only support in 
>this round of document revisions, I think a SHOULD here is correct."
><http://130.230.52.14/list-archive/ipsec/msg05576.html>http://130.230.52.14/list-archive/ipsec/msg05576.html
>
>however Donald Eastlake had stated
>@@@ I think draft-ietf-ipsec-esp-v3-09 should be changed.
><http://130.230.52.14/list-archive/ipsec/msg05578.html>http://130.230.52.14/list-archive/ipsec/msg05578.html
>
>The issue never got resolved and we now have this discrepancy in the 
>RFC's. Should I send an errata for RFC4305 regarding the same?
>
>Thanks,
>Vishwas

Whoops.  Sorry that this one fell through the cracks in the 
intervening year after you noted the discrepancy.

I still think a SHOULD is appropriate for ESP, given the changes in 
the architecture document. Since this is a significant change (from a 
MUST to a SHOULD), it cannot be an errata, as Paul noted. I'll ask 
Russ how he would like to handle this.

Steve