ietf-nntp Wildmats

[Russ Allbery]

Charles Lindsey <chl at clw.cs.man.ac.uk> writes:
> Andrew Gierth <andrew at erlenstar.demon.co.uk> writes:
>>>>>>> "Clive" == Clive D W Feather <clive at demon.net> writes:

>> Clive> [W6: the grammar does not treat , in sets as special. This
>> Clive> means that the wildmat "a[b,c]d" is a single pattern that
>> Clive> matches the three strings "abd", "a,d", and "acd". Are we
>> Clive> happy with this ? It matches existing practice

>> it does no such thing. Existing practice, where a list of wildmats is
>> expected, treats ',' as a delimiter without any escape mechanism (and
>> none is needed).

> There is NO EXISTING PRACTICE on this.

This is not correct.  There is too existing practice, in INN among other
places, and the current implementations do exactly what Andrew says that
they do.

> For a start, the whole wildmat-set business is new since RFC 977, so
> there is no existing definition in the matter (and it only arises in the
> NEWNEWS command currently).

Andrew did specify "existing practice."

> Then, in the second place, wildmats have (apart from XPAT) only ever
> been used to match newsgroup names (and this will remain true). Since
> ',' cannot occur within a newsgroup name (for obvious reasons) it is
> impossible for this case ever to have arisen in the past; nor will it
> ever arise under our present proposal.

Just because ',' cannot occur in a newsgroup name doesn't prevent a user
from sending a pattern with a wildmat set containing a comma.  If they did
so, it would be treated as Andrew describes.

> And in the case of XPAT, if it has ever arisen it SHOULD NOT have been
> treated as a separator of alternative patterns because that has not been
> a feature of wildmats until now.

Since XPAT is not a standardized command, I see little basis for this
statement.  These concepts don't spring into existence only when we write
them down in an RFC and not before.

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