Re: Issue 13.3.1 -- I say "go for it"

[Andy Bierman <abierman@cisco.com>]

At 09:04 PM 2/6/2004, Phil Shafer wrote:
>Eliot Lear writes:
>>The notion of a "create" operation is not new, dating back to at least 
>>O_CREAT/creat(2), and possibly the days of Benjamin Franklin.  SNMP has 
>>it for good reason, and we should have it as well.  Do people really 
>>need to be sold on it?
>
>I'd be against requiring an explicit create.  Implicit
>creation is fine for >90% of the cases and the other
><10% can do a query to see if the target exists.

This would be a rather inconsistent approach.
We should try to get the protocol operations to be
as robust as possible from the beginning.  Some application
developers have requested this, not only because it's
more efficient, but it provides some bug detection
and database corruption prevention capabilities.

As it is, an agent has to be capable of identifying
the portion of the data model that corresponds to
a particular instance identifier, in order to
support the 'merge', 'replace' and 'delete' operators.  
Since this code is mandatory, a little extra code to kick back
an 'ALREADY_EXISTS' or 'OPERATION_FAILED' error if
this parameter is 'create' is not a big deal.


>Thanks,
> Phil

Andy


--
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/>