-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

A RELOAD node needs to receive the certificate of the signer of a kind in the
same document than this kind.  The usual way of doing this is to add the
signer certificate in the SecurityBlock that is encoded in the kind-signature
element.  But if there is multiple kinds, all signed by the same signer, the
certificate is essentially duplicated.  This is in addition to the certificate
that will be embedded in the SecurityBlock in the signature element, if they
share the same signer.  Having duplicate certificates is not an issue in
itself, but increasing the size of the configuration will probably trigger
fragmentation when it is sent in a ConfigUpdate request over UDP, so I would
like to optimize this case.

My current idea is to never add a certificate in a kind-signature, but to add
all the certificates required to verify all the signatures in the enclosing
signature element.  This requires the following:

1. When verifying the signatures of a configuration file, the certificates
must be searched in the signature SecurityBlock, in addition of the
SecurityBlock in the kind-signature.

2. When sending a ConfigUpdate request with one or more kind-block element
(type==kind), the SecurityBlock inside the kind-signature element must be
modified by adding the certificate, if it is not already present.

AFAIK, this is compatible with the RELOAD specification, but I can be wrong.
If it is the case, which part of the specification explains this?

Thanks.

- -- 
Marc Petit-Huguenin
Email: [email protected]
Blog: http://blog.marc.petit-huguenin.org
Profile: http://www.linkedin.com/in/petithug
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
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=Yfd9
-----END PGP SIGNATURE-----
_______________________________________________
P2PSIP mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/p2psip

Reply via email to