Re: RADEXT charter, Take 8; dictionaries

[wolfgang.beck01@t-online.de (Wolfgang Beck)]

David Nelson wrote:
> Hmmm...  I guess if the evidence presented to date does not convince
> you, then it entirely possible that you will never be convinced.  Maybe
> it has to do with how one defines "break".  I define it as (a) deviation
> from the established protocol model, and (b) requiring changes to the
> protocol parsing engine of RADIUS implementations (as opposed the
> changes in the semantics enforcement engine, which is obviously required
> each time a new data element is introduced).

The problem with (b) is, that implementation details - like the concept
of a dictionary - are not standardized. You are saying that changes to the
attribute parser are painful and this might be true for some
implementations. In other implementations, parser extensions are the
easy part and changing the semantic enforcement engine is hard.

(a) is more a matter of style. String attributes may only be structured if
the components cannot be used individually. We may need some more
guidelines here.

The definition of "break" can only be derived from the RADIUS RfCs. No
WG could ever judge if an extension would work with every implementation
out there.


Wolfgang


--
to unsubscribe send a message to radiusext-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/radiusext/>