Re: Clarifications for mteObjectsEntry and mteObjectsIndex DESCRIPTIONs: Issue-3

Hi -

> From: "Benoit Claise" <[email protected]>
> To: "Disman (E-mail)" <[email protected]>
> Sent: Tuesday, July 19, 2005 3:09 AM
> Subject: [Disman] Clarifications for mteObjectsEntry and mteObjectsIndex DESCRIPTIONs: Issue-3
>
> Dear all,
>
> Regardless of the order (trigger, trigger test, event) discussion just
> posted as an Issue-2, I would like to come back to the DESCRIPTIONs of
> mteObjectsEntry and mteObjectsIndex.
>
> mteObjectsEntry OBJECT-TYPE
>     SYNTAX      MteObjectsEntry
>     MAX-ACCESS  not-accessible
>     STATUS      current
>     DESCRIPTION
>         "A group of objects.  Applications create and delete entries
>         using mteObjectsEntryStatus.
>
>         When adding objects to a notification they are added in the
>         lexical order of their index in this table.  Those associated
>         with a trigger come first, then trigger test, then event."
>     INDEX       { mteOwner, mteObjectsName, mteObjectsIndex }
>     ::= { mteObjectsTable 1 }
>
> I think that the last 2 sentences are not very clear.
> I guess that the correct behavior is that the second one takes
> precedence over the first one, and that the goal is to say:
> trigger object group first, trigger test object group next, and event
> notification object last. Within the object group (mteTrigger*Objects or
> mteEventNotificationObjects), then we take the numerical order of the
> mteObjectsIndex, the third index in mteObjectsEntry.
>
> If this is the correct behaviour, I'm confused by "When adding objects
> to a notification they are added in the lexical order of their _index
> _in this table. Those associated with a trigger come first, then trigger
> test, then event.".
> Which index do we speak about? mteOwner, mteObjectName, or mteObjectsIndex?

Clearly it's mteObjectsIndex.  For a group, all the entries will have the same value
for their mteOwner and for their mteObjectName.

> mteObjectsIndex OBJECT-TYPE
>     SYNTAX      Unsigned32 (1..4294967295)
>     MAX-ACCESS  not-accessible
>     STATUS      current
>     DESCRIPTION
>         "An arbitrary integer for the purpose of identifying
>         individual objects within a mteObjectsName group.
>
>         Objects within a group are placed in the notification in the
>         numerical order of this index.
>
>         Groups are placed in the notification in the order of the
>         selections for overall trigger, trigger test, and event.
>         Within trigger test they are in the same order as the
>         numerical values of the bits defined for mteTriggerTest.

Note that this last paragraph provides some of the missing information
about ordering!

>         Bad object identifiers or a mismatch between truncating the
>         identifier and the value of mteDeltaDiscontinuityIDWildcard
>         result in operation as one would expect when providing the
>         wrong identifier to a Get operation.  The Get will fail or get
>         the wrong object.  If the object is not available it is omitted
>         from the notification."
>     ::= { mteObjectsEntry 2 }
>
> Again, I think the second and third sentences are confusing. If I read
> first "Objects within a group are placed in the notification in the
> numerical order of this index.", I'm asking myself what a group is! I

A group is a set of entries with the same values for the (mteOwner,
mteObjectName) tuple.  If that's not clear, we should make it explicit.

> guess/understand that a group in the case of mteObjectsIndex refers to
> trigger, trigger test, or event but this don't think this is specified
> anywhere in the draft. What is even more confusing is that
> mteObjectsEntry refers to "a group of objects" but the meaning is
> completely different.

No, it's identical: the set of entries with the same (mteOwner, mteObjectsName)
tuple in their indexes constitutes a group.  If it's not clear enough, we
should make it more explicit.

> Am I the only one to consider this very confusing?
> Should we think of some improve text?

If a document's editor finds it confusing, that alone is good reason to
improve the text's clarity, in my opinion.

> Regards, Benoit

Randy