RE: Changing enum labels considered harmful

["Presuhn, Randy" <Randy_Presuhn@bmc.com>]

Hi -

> From: C. M. Heard [mailto:heard@pobox.com]
> Sent: Friday, January 17, 2003 11:08
> To: mreview@ops.ietf.org
> Subject: RE: Changing enum labels considered harmful
...
> But a problem _can_ arise if the NMS imports two modules, one with the
> definition of an object or TC that uses updated versions of the enum
> labels, and another module that uses the TC or object but did not get
> updated.  The problem happens if the second module uses the 
> label alone
> without a number -- as would happen in a DEFVAL clause.  This is not
> just a matter of a MIB compiler being picky, it's a matter of it not
> being able to figure out what the DEFVAL clause means.  A human being
> could not so so either without having the old version of the 
> definition
> available.  That's the point I was trying to get across in 
> the examples.
...

The problem (DEFVAL references to enum labels from other modules)
is worse for instrumentation code-generating compilers, since
they really do care what the DEFVAL resolves to.  For a manager,
knowing what the DEFVAL value is provides at most an optimization
opportunity.

Randy