Re: [YANG] new pyang errors

Andy Bierman píše v Ne 20. 01. 2008 v 10:01 -0800:
> > If something is mandatory, it must exist.  If there's a default,
> > it need not exist, since the default gives the value.  Making
> > the agent/manager/user/implementor ask your Q1 and Q2 is unnecessary.
> > The question is simply "does it exist?".  If it's mandatory and it
> > doesn't exist, it's an error.
> > 
> 
> We disagree on which POV is confusing.
> IMO, it is important to be able to easily distinguish between
> some knob that the agent is not using at all (whatever that
> means is DM-specific) vs. a knob that is being used, but which
> has the DM-defined default value.

More to this point: an agent saying "I have parameter X set to the
default value but I won't tell you what it is." doesn't seem acceptable
to me. Even if standard defaults were defined, I am sure there will be
broken (optimised:-) implementations with other default values. It would
be extremely hard to debug network problems if the manager cannot verify
the actual values used by all agents.

I'd thus allow both agents and managers to have their own defaults but
always be explicit when exchanging the data. Relying on common universal
defaults would make the system fragile.

Lada

> 
> 
> > Thanks,
> >  Phil
> > 
> > 
> 
> Andy
> 
> 
> _______________________________________________
> YANG mailing list
> YANG@ietf.org
> https://www1.ietf.org/mailman/listinfo/yang
-- 
Ladislav Lhotka, CESNET
PGP Key ID: E74E8C0C

_______________________________________________
YANG mailing list
YANG@ietf.org
https://www1.ietf.org/mailman/listinfo/yang