ietf-smtp
[Top] [All Lists]

RE: RFC 5321 VRFY and quoting syntax

2011-05-11 16:54:47

-----Original Message-----
From: A. Rothman [mailto:amichai(_at_)amichais(_dot_)net]
Sent: Wednesday, May 11, 2011 1:26 PM
To: Murray S. Kucherawy
Cc: IETF SMTP
Subject: Re: RFC 5321 VRFY and quoting syntax

This is not exactly an issue of conflict, but rather of leaving room for
interpretation. The ABNF is indeed well-defined on it's own as you
state, however when one reads that the argument 'cannot be further
restricted' (is that before or after quoting?) and sees many examples
out in the wild, most notably the brackets but also others (as in
Hector's previous response which uses backslash escaping outside of a
quoted-string), it raises doubt of whether 'cannot be further
restricted' means 'you can do anything you want' or indeed still must
abide by the rules of quoting. Having many examples violating the rules
of quoting gives more weight to the first interpretation. Or, in more
practical terms - since there are many common violations, the RFC
wording might have some room for improvement in regards to being more
explicit and preventing these violations.

It has worked for me, in such situations, to err on the side of implementing 
the ABNF and not the examples.  In my experience there's more attention paid to 
getting that right than there is to getting the examples right.  However, this 
might not be one of those occasions.  :-)

If you believe the ABNF, that's correct. 

See above.  :-)

If you believe the text in that
paragraph... not so sure any more. Even in the example within the
paragraph of "Joe\,Smith", it is unclear if the quotes are part of the
actual argument, or just a way to embed the example within a paragraph
of explanatory text. Here too, I think the fact that there are various
violations out there which can legitimately point to this paragraph and
say 'it says so in the RFC'... it would be better imho to update the
spec's wording to be unambiguous, e.g. "Note that the backslash, "\",
when used within a quoted string, is a quote character..."

I would recommend opening an errata item then.  They get reviewed and 
eventually assigned to be included in the next update, and as it happens 
there's a working group chartered to take care of those sorts of things.

ok, I'll do that when this thread ends, if no one comes up with a
counter-explanation.

Let me know if you need help navigating the process.

-MSK