[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: message-id attribute issues
On Tue, Dec 20, 2005 at 11:48:00AM -0800, Andy Bierman wrote:
> Let's make conscious decisions about the protocol.
> This upper limit for the manager is the lower limit for the agent.
> If the manager is allowed to send it then the agent MUST accept it.
> I don't know what this idea of a smaller implementation requirement
> is supposed to mean. Please explain.
I guess it boils down to the question how hard-wired a limit is.
The SMI specs state that an OID is 128 sub-identifiers (something
people considered a useful large number several years ago). We now
know that this limit hurts with table indexes becoming larger and if
you want to do more fancy protocol extensions (which we never got
agreement on - but at least the OID limit was one of the factors that
did hurt). And we have other somehow arbitrary limits in the SNMP
world which then led to CLRs.
The size of the message-id attribute might be less of an issue - but
who knows for sure? The question I think is whether we forsee that
sizes might change in the future and we envision that new ones can be
announced via say a capability or whether we hard-wire arbitrary size
constraints in the XSD based on what we believe is a good choice
today, which means the XSD has to change to lift such limits.
/js
PS: I did a quick scan of the IMAP specs to see whether they limit
the size of the IMAP tag (which I think is very close in nature
to netconf's message-id). I did not find a size limit - does
anyone know whether such a thing exists? Otherwise, if IMAP
can do without a specific size limit, perhaps netconf 1.0 can
do as well.
--
Juergen Schoenwaelder International University Bremen
<http://www.eecs.iu-bremen.de/> P.O. Box 750 561, 28725 Bremen, Germany
--
to unsubscribe send a message to netconf-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/netconf/>