ietf-nntp new draft of base document available

[Russ Allbery]

Howard Swinehart <howard at redrose.net> writes:
> From: "Russ Allbery" <rra at stanford.edu>

>> ! is not at all ambiguous in existing implementations; the code that
>> implements it produces a well-defined result given any input.  What
>> about it do you think is ambiguous?

> Section 5.2 of the current draft gives these examples of a negated
> wildmat pattern:
>      e)!bc*d -- matches any string that does not start with
>         "bc" and end with "d" (without quotes)
>      f)!\\x  -- matches any string that does not start with
>         "\x" (without quotes)

> This to me implies that a command should return everything that doesn't
> match the pattern after the !.

Any wildmat expression whose first pattern begins with ! is treated as if
an implicit "*," were appended to the beginning of the pattern.

(I'd use prepended here, but the OED claims this is not an appropriate use
of that term, and gives a definition of append that allows this usage.
Interesting.)

> Should a ! in a comma separated list have a different meaning?  That is,
> should it exclude rather than include results like in the examples?  If
> ! acts differently depending on the context, then it would be helpful if
> this was documented.

Does the addition of the above sentence fully clarify it?

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>