[NNTP] Consensus?

[Thomas Gschwind]

Hi all!

On Tue, Aug 09, 2005 at 06:15:56PM -0700, Russ Allbery wrote:
> Steve Walker <nntp at nntpserver.com> writes:
> > Personally I would like the wording changed to require (a MUST) 32 bit
> > support and suggest (a SHOULD) 64 bit support.
> 
> That's two people supporting that option.  What does everyone else think
> about that?

I am also very much in favor of this solution.

On Wed, Aug 10, 2005 at 10:56:34AM +0000, Charles Lindsey wrote:
> My personal view is that a wrap around solution would be best long term
> (and I think we have discussed enough to see that such a scheme would be
> possible, but obviously not to the extent that we can define it now). So I
> don't want to make any decision now that would make it difficult later.

Recommending the use of 64 bit article numbers does not preclude us from
specifying a wrap around scheme.  It only postpones it by some more
time.  Enough time till we get out a newer version.

> Most systems are safe for several years. It is only Steve's unusual scheme
> which might fail after 2 years, and I don't think we should be forced into
> a premature decision just because of that.

What would be the problem with recommending a 64-bit counter?  So far we
know there are some people having problems if we specified a 32-bit
limit.

On Wed, Aug 10, 2005 at 07:20:51AM -0400, Scott Hollenbeck wrote:
> > I would have no problem if we want to limit it to 31 bits for
> > now, with a
> > SHOULD support 32 (i.e. SHOULD use unsigned int).
>
> I would advise against using 2119 directives with any programming
> language-
> or machine architecture-specific data types/limits.  It's OK to
> mention the
> range of numbers that can be expected, but it's best to stay far away
> from
> implementation-specific details unless they are clearly identified as
> examples.

That's actually a good point.  Also considering some "broken" languages
that don't have unsigned data types (eg Java), it might be nice to
specify the ranges maybe as MUST 1..2^31-1 and the SHOULD as 1..2^63-1.
I would be very much in favor of that.

Regards,
Thomas
-- 
Thomas Gschwind                      Email: thomasg at ieee.org
IBM Zurich Research Lab
Saeumerstrasse 4                     Tel: +41-1-724-8990
CH-8803 Rueschlikon, Switzerland     Fax: +41-1-724-8953