[regext] AD review of draft-ietf-regext-data-escrow-04

Barry Leiba <barryleiba@computer.org> Fri, 14 February 2020 04:38 UTC

MIME-Version: 1.0
From: Barry Leiba <barryleiba@computer.org>
Date: Fri, 14 Feb 2020 04:38:19 +0000
Message-ID: <CALaySJJijZeiSiu3ChspN8rDVb=DCFhM9E76xwyBVQaEMedtpA@mail.gmail.com>
To: "draft-ietf-regext-data-escrow.all@ietf.org" <draft-ietf-regext-data-escrow.all@ietf.org>
Cc: regext <regext@ietf.org>
Content-Type: multipart/alternative; boundary="000000000000020034059e81c75e"
Archived-At: <https://mailarchive.ietf.org/arch/msg/regext/UNo6YxapgjyerAYv0223zEuzjFk>
Subject: [regext] AD review of draft-ietf-regext-data-escrow-04
Precedence: list

Good work on this document, and thanks.  I have only some small points to
comment on, which can be fixed later, and I’ll request last call as soon as
I’ve sent this message.  My apologies for being swamped and not getting to
this as soon as I should have.

Barry

— Section 3 —

   While the main motivation for developing this specification is rooted
   on the domain name industry, the specification has been designed to
   be as general as possible.

This strikes me as awkward English.  May I suggest a rewording?:

NEW
   While the domain name industry has been the main target for
   this specification, it has been designed to be as general as possible.
END

— Section 5.1 —

      Each registry is responsible for maintaining its
      own escrow deposits identifier space to ensure uniqueness.

Nit: Make it  << deposits’ >>, with the trailing apostrophe to make it
possessive.

   o  An OPTIONAL "prevId" attribute that can be used to identify the
      previous Incremental, Differential or Full Deposit.  This
      attribute MUST be used in Differential Deposits ("DIFF" type).

It seems that labelling this as OPTIONAL isn’t really right, because its
use depends upon the deposit type, right?  Perhaps this?:

NEW
   o  A "prevId" attribute that can be used to identify the previous
      Incremental, Differential or Full Deposit.  This attribute is REQUIRED
      in Differential Deposits ("DIFF" type), is OPTIONAL in Incremental
      Deposits (“INCR” type), and is not used in Full Deposits (“FULL
      type).
END

— Section 5.4 —

   This section of the deposit SHOULD NOT be present in Full Deposits.
   When rebuilding a registry it SHOULD be ignored if present in a Full
   Deposit.

Questions about both “SHOULD”s:
For the “SHOULD NOT”, why might an implementation need to put that section
in?  That is, why is this not “MUST NOT”?
For the “SHOULD”, why is it not “MUST”?  What else but ignoring it could a
rebuilder possibly do?

— Section 10 —

   Depending on local policies, some elements or most likely, the whole
   deposit will be considered confidential.  As such the registry
   transmitting the data to the escrow agent SHOULD take all the
   necessary precautions like encrypting the data itself and/or the
   transport channel to avoid inadvertent disclosure of private data.

Nits: This needs a comma after “or” in “or, most likely,” and a comma after
“As such”.  Also, “like encrypting” should be “such as encrypting”.

   It is also of the utmost importance the authentication of the parties
   passing data escrow deposit files.

NEW
   Authentication of the parties passing data escrow deposit files
   is also of the utmost importance.
END

Apart from that rewording, it seems odd that something “of the utmost
importance” is labelled as SHOULD, rather than MUST, yes?

   RECOMMENDED to ensure not only the file was transmitted correctly
   from the registry, but also the contents are also "meaningful".

Nits: this needs “that” before both “the file” and “the contents”.

[regext] AD review of draft-ietf-regext-data-escr… Barry Leiba