Re: I-D ACTION:draft-ietf-ldapbis-models-12.txt

[Hallvard B Furuseth <h.b.furuseth@usit.uio.no>]

RL 'Bob' Morgan writes:
> This document has been recommended to the IESG for consideration as a
> Proposed Standard.  Note that comments will be solicited during
> IETF-wide last call also.

This comment was outstanding, though it's not particularly important.
Not sure if anyone is interested; the only response was a request for
clarification from Ron.

  From: Hallvard B Furuseth <h.b.furuseth@usit.uio.no>
  Date: Sat, 6 Nov 2004 17:24:42 +0100
  To: ietf-ldapbis@OpenLDAP.org
  Subject: ABNF of numericoid

  I've discovered that I've lost this old thread again.

  I had the impression that we landed on using the exact syntax for
  numericoid, but maybe I confused the issue too much:-(
  Anyway, [bcp64] and [models] say:

      numericoid = number 1*( DOT number )

  I wonder if we could change that to Kurt's suggestion (3) in the old
  thread, so the syntax expresses exactly which values are valid:

      numericoid = %x30-31 DOT [ %x31-33 ] DIGIT *( DOT number )
                 / %x32 1*( DOT number )
                 ; Starts with 0-1.0-39 or 2.number


  BTW, I think I was not the only one who had the misconception that the
  BER encoding's first component (40 * value1 + value2) can be only 1
  octet, and therefore that value2 in the syntax was not unlimited even
  with value1 = 2.  Where did others get that idea from?  I got it from
  RSA's "A Layman's Guide to a Subset of ASN.1, BER, and DER".  On the
  Asn1 mailinglist, I was told guide has a number of errors.

-- 
Hallvard