Internet X.509 Public Key Infrastructure Subject Identification Method (SIM)
draft-ietf-pkix-sim-08

[Ballot comment]
I don't understand this field very well yet, and in particular I do not understand how SIITypes will be defined.  Will they be defined in a registry, or by separate drafts or other?

[Ballot discuss]
I do not believe this document adequately describes why the binding of privacy sensitive identifiers
to certificates is the appropriate way to manage the use cases they describe.  Given the description,
use of other cryptographic means to create an appropriately protected object which contains
an assertion by Alice seems to be reduce the number of people to whom Alice must reveal
this information by taking the registration authority (and possibly the certificate authority) out
of the picture. To handle this, I believe the authors must describe why something like a signed,
encrypted S/MIME  object containing a message to Bob like "My employee number is 485-777-883AG" 
does not meet the case.

This document also consistently uses the U.S. Social Security Number as an example.  While it
describes it as one of a class of privacy sensitive identifiers:

  On the other hand, such an identifier is not actually a secret.
  People choose to disclose these identifiers for certain classes of
  transactions. For example, a person may disclose his/her Social
  Security Number to open a bank account or obtain a loan. This is
  typically corroborated by presenting physical credentials (e.g., a
  driver license) that confirm the person's name or address.

the Security Considerations section's description of the short length of
and other issues with SII's make it clear that this is a very dominant use
case.  A more problematic example could hardly be found.  The use of
SSNs, and their association with specific classes of data, is the subject
of considerable legislation and legal wrangling.  The U.S. privacy act of 1974
restricted its use, and various homeland security regulations have since
expanded the cases in which it must be tightly bound to other identifiers
if it is bound at all. 

If it is to be used here as an example, I believe the authors
must at least describe why the use of SIM or PEPSI avoids implication in that
legal framework (which plaintext inclusion clearly would not).  If they cannot
do so, I strongly suggest they pick another example.

[Ballot comment]
Is there any special reason that the appendix uses the ASN.1:1988 Notation rahter than the current ASN.1:2002 notation? For differences and the process of changes between the two see http://www.itu.int/ITU-T/studygroups/com17/changing-ASN/index.html.