Added a Use In CMS section

[ounsworth]

Adding this section here avoids the need for a wrapper draft telling you how to use composite KEM in CMS.

Closes #12

[ounsworth]

Note-to-self: look at whether we need to specify higher levels of KDF and WRAP for the stronger algorithms -- see how DVG did it in the CMS-Kyber draft.

UPDATE: done.

[ounsworth]

Question: does the ## Certificate Conventions section collide with the recently-merged Key Usage section?

UPDATE: I think this is fine because this section is intended to be able to be standalone if LAMPS wants the alg definitions, and the "Use in CMS" to become separate documents.

[ounsworth]

We also need to add the SMIME-CAPS to the ASN.1 (find a good example doc).

[danvangeest]

Sorry! I created this review last week but forgot to actually submit it!

[ounsworth]

Thanks for the review @danvangeest

Still TODO:

• We also need to add the SMIME-CAPS to the ASN.1 (find a good example doc).
  • UPDATE: done.

[ounsworth]

Review comments from Russ:

Mike:

You are using a mix of "wrap algorithm" and "key-encryption algorithm". CMS usually uses the later. Line 683 got it just right. That is, id-aes256-Wrap identifies a particular key-encryption algorithm. I am not sure the text at the top is aligned.

Line 663: I think you need o say how that happens. In some cases a different OID is used, but in others a different parameter value is used.

Line 675: make it clear thet it is the KEM.Encaps(pk). The function is specific to the KEM identified in 'kem'.

Line 681: Do any of your composite KEM algorithms REQUIRE a ukm? I can imagine a way to use DH and ECDH that require a UKM.

Russ