ietf-dkim
[Top] [All Lists]

[ietf-dkim] What is your view on these three SSP topics?

2008-01-21 19:55:22
As SSP is currently defined, a restricted or unrestricted key must be on-behalf-of the From header to be compliant with SSP "strict".

1) To be SSP "strict" or "all" compliant, the identity associated with a signature:

a) must match an email-address within the From header.

b) must match an email-address associated an entity introducing the message contained within the following headers:

   From, Sender, Resent-From, Resent-Sender.

c) may not match with any header and may not be defined.

d) may not match with any header and may not be defined except for SSP "strict" which is limited to not being defined or matching the first email-address within the From. (as currently defined)


The number of DNS transactions required to support SSP is affected by the number of From email-addresses being protected. Currently SSP is defined to reference SSP only for the first From email-address.

2) SSP references should be done for:

a) Only the first From email-address. (as currently defined)

b) All From email-addresses.

c) More than one From email-address is not be SSP compliant and is to be considered to produce an invalid signature.

d) More than two From addresses is not be SSP compliant and is to be considered to produce an invalid signature. When two From addresses are used, both domains should be checked for SSP compliance.


RFC 4871 Section B.1.1 4th paragraph makes this statement:

[An outsourcing company can supply g= restricted keys] "using a unique selector that authorizes a specific From header field Local-part. For example, a client with the domain "example.com" could have the selector record specify "g=winter-promotions" so that this signature is only valid for mail with a From address of "winter-promotions(_at_)example(_dot_)com "."

This comment is within the appendix and there is nothing within RFC 4871 nominative section that limits use of a g= restricted keys to the just the From header, although this seems implied by section B.1.1.

Currently SSP is defined where g= restricted keys are compliant with "all".

3) Keys restricted by a g= parameter are treated as valid and are:

a) SSP "all" or "strict" compliant only when on-behalf-of a From header email-address.

b) SSP "all" or "strict" compliant regardless of the identity signed on-behalf-of.

c) SSP "strict" compliant only when on-behalf-of the first From header email-address. (as currently defined)



1-c

2-d

3-a

-Doug



_______________________________________________
NOTE WELL: This list operates according to http://mipassoc.org/dkim/ietf-list-rules.html