[NNTP] Re: Command names

Clive D.W. Feather clive at demon.net
Wed Oct 27 00:59:06 PDT 2004


Ken Murchison said:
> Looks reasonable to me, provided there aren't any other special chars 
> that might be desirable.  What is the rationale for restricting a 
> keyword to something less that all printable chars?

It makes things more readable. Do we really want unbalanced parentheses and
quotation marks in command names?

It gives us room for expansion and for special cases, by using the
characters other than those we permit for commands.

For example, if we found a need for some kind of scoping of commands
(e.g. simultaneous use of more than one authentication ID) we can do this
with a non-permitted character:
    [user=clive] GROUP local.public
...
    [user=root] GROUP local.secret
(I blithely skip over the issues of managing multiple IDs).

For example, assume we decide that my STATUS proposal isn't worth doing
but, a year down the line, discover that there's a real need to say "this
protocol is supported but not available right now". We can allow the LIST
EXTENSIONS response line to begin with (say) "!" to indicate this.

For example (and this is really blue-sky stuff; please don't criticise it)
we could allow a prefix of "&" to mean "do this asynchronously and I'll ask
for the result later". This would be a more generalised form of streaming.
Or we could allow commands to be inserted in a posting body by prefixing
them with a dot, but this relies on command names themselves not beginning
with a dot (so that they look like dot-escaped lines).

I don't know what we will want to do with it, but I'd rather not rule the
possibility out at this stage.

-- 
Clive D.W. Feather  | Work:  <clive at demon.net>   | Tel:    +44 20 8495 6138
Internet Expert     | Home:  <clive at davros.org>  | Fax:    +44 870 051 9937
Demon Internet      | WWW: http://www.davros.org | Mobile: +44 7973 377646
Thus plc            |                            |



More information about the ietf-nntp mailing list