Augmented BNF for Syntax Specifications: ABNF
draft-crocker-abnf-rfc2234bis-00

[Note]: 'This document obsoletes RFC 2234.  It was produced to address comments received during the IESG evaluation of moving 2234 to Draft Standard status.  Implementation report found here: http://www.ietf.org/IESG/Implementations/RFC2234-implementation-report.txt' added by Amy Vezza

[Note]: 'This document obsoletes RFC 2234.  It was produced to address comments received during the IESG evaluation of moving 2234 to Draft Standard status.  Implementation report found here: http://www.ietf.org/IESG/Implementations/RFC2234-implementation-report.txt' added by Scott Hollenbeck

[Note]: 'This document obsoletes RFC 2234.  It was produced to address comments received during the IESG evaluation of moving 2234 to Draft Standard status.  Implementation report found here: http://www.ietf.org/IESG/Implementations/RFC2234-implementation-report.txt' added by Amy Vezza

[Ballot comment]
If Pekka is talking about parsers, my parser implements all of the features of this I-D.  I'm pretty sure that Harald's parser does a large subset.

Very minor concerns:

Should the "Changes in the latest version of this Internet Draft" section be removed or change to "Changes since RFC 2234"?

Section 3.1 has "NOTE: NOTE: ...".

[Note]: 'This document obsoletes RFC 2234.  It was produced to address comments received during the IESG evaluation of moving 2234 to Draft Standard status.  Implementation report found here:

http://www.ietf.org/IESG/Implementations/RFC2234-implementation-report.txt' added by Scott Hollenbeck

[Ballot comment]
> 1.  INTRODUCTION
>
>    Internet technical specifications often need to define a format
>    syntax and are free to employ whatever notation their authors deem
>    useful.

surely s/format/formal/? (also in Abstract)

>    Changes in the latest version of this Internet Draft:
>
>      In Section 3.7 the phrase: "That is, exactly à occurrences of
>      ." was correct to: "That is, exactly à occurrences of
>      ."
>
>      Some continuation comment lines needed to be corrected to begin
>      with comment character (";").

should be flagged for removal by RFC-Ed

Where is list of changes from RFC 2234?

Isn't there an Ur-reference for ABNF prior to RFC 733? I remember seeing
ABNF used in the late 1960's in compiler land.

[Ballot discuss]
Does this meet the interoperability requirement? Or do we
declare that irrelevant?

From Suzanne Wolf:

I'd have no trouble saying this is ready to advance as Proposed
Standard. The absence of an IANA Considerations section bugs me a
little, but it would be just like the Security Considerations section
anyway, so why enslave the electrons to say so?

However....I'm assuming it's flagged as Draft because RFC 2234 is
Proposed, but I share the concern of a couple of IESG members about
the lack of implementation report details. (It sounds like there is
one....)

It seems especially important to be able to assert there are multiple
parsers that handle ABNF as specified here. First it demonstrates that
the spec allows for self-consistent and interoperable implementations,
as required for Draft. But it seems even more important that we be
sure ABNF as documented here is powerful enough to support the use
intended, and serve as infrastructure for defining new protocol.

It's entirely possible the part that seems to be missing isn't really
and I just don't know where it is. But the concerns raised about the
requirements for Draft seem reasonable to me.

[Ballot comment]
If Pekka is talking about parsers, my parser implements all of the features of this I-D.  I'm pretty sure that Harald's parser does a large subset.

[Ballot discuss]
Very minor concerns:

Should the "Changes in the latest version of this Internet Draft" section be removed or change to "Changes since RFC 2234"?

Section 3.1 has "NOTE: NOTE: ...".

[Ballot comment]
Comments from the OPS directorate by Pekka Savola (Mar 30 17:47:13 PST 2005):

Good that someone wants to push these kind of docs forward in the
standards track.  Maybe this should go for full standard later on?

substantial:
- the implementation report lists documents which use all the
  features of ABNF (good), and points out two specific implementations
  of ABNF.  However, said two (or some other) implementations have not
  been analyzed whether they really implement the whole ABNF language.
  If they do, mention that in the implementation report.  If not, find
  new implementations which do or remove the feature(s).

- the document should have a section, e.g., at the appendix, which
  explicitly describes the changes (or lack thereof) since RFC2234.

editorial:
- in references, s/Descriptive/Informative/
- add a dummy IANA considerations section
- there are two non-ascii chars in the draft.  The 'Changes in the
  latest version of this Internet Draft' in Introduction should be
  clearly marked to be removed by RFC-editor prior to publication.

[Note]: 'This document obsoletes RFC 2234.  It was produced to address comments received during the IESG evaluation of moving 2234 to Draft Standard status.' added by Scott Hollenbeck

[Note]: 'This document obsoletes RFC 2234.  It was produced to address comments received during the IESG evaluation of moving 2234 to Draft Standard status.' added by Scott Hollenbeck