Re: protocol: ASN.1 for Attribute

[Steven Legg <steven.legg@adacel.com.au>]

Hallvard,

Hallvard B Furuseth wrote:
> I'd like to join my belated vote against the hairy
>
>         Attribute ::= PartialAttribute(WITH COMPONENTS {
>              ...,
>              vals (SIZE(1..MAX))})
>
> definition to John McMeeking's message in the 'REVISED protocol
> review notes [LONG]' thread.  All other ASN.1 in [Protocol] is quite
> simple, and easy to understand.  This exception is not necessary.
> As John said, this is not even longer:
>
>           Attribute ::= SEQUENCE {
>                type       AttributeDescription,
>                vals       SET SIZE(1..MAX) OF value AttributeValue }
>
> I don't see any arguments in _favor_ of the current notation in that
> thread, except that it's cool.

The current notation allows ASN.1 compilers to recognize that Attribute
is related by derivation to PartialAttribute. Compilers can then
produce a programming language object class hierarchy with PartialAttribute
as the base class and Attribute as a subclass. This allows instances of
Attribute to be used interchangably with instances of PartialAttribute
in application code (insofar as the minimum number of values is of no
importance).

Regards,
Steven


> While it's cool to be cool, I don't
> think that's a very strong argument.