MCU decomposition delayed contribution

[OKUBO Sakae]

Paul,

If empty lists make no sense, they should have been constrained to _not_ be
empty, e.g., X SEQUENCE SIZE(1..) OF Y! 8-| I'm not necessarilly picking the
editors, but, IMO, it is real sloppy to go to the trouble of defining the
ASN.1 syntax for something and then later on saying, "Oh, an empty list, why
would you want to encode that? Let's go back and make it semantically
illegal." Or, worse, the behavior is not defined and we have interop
problems. IMO, every SEQUENCE OF and SET OF should be constrained for
maximum and for minimum when it makes sense, e.g., X SET SIZE(1..256) OF Y.

Also, let's not have anymore optional empty lists (e.g., X SEQUENCE OF Y
OPTIONAL) unless an absent list is semantically different from an empty one,
which is extremely rare.

There, I feel better. :-)

Paul Long
ipDialog, Inc.

-----Original Message-----
From: Mailing list for parties associated with ITU-T Study Group 16
[mailto:ITU-SG16 at mailbag.cps.INTEL.COM]On Behalf Of Paul E. Jones
Sent: Saturday, May 12, 2001 12:01 PM
To: ITU-SG16 at mailbag.cps.INTEL.COM
Subject: Re: PER ASN.1 encoding of Sequence Of Elements - Boundary
Conditions


Manoj,

It is legal, but one should not do this unless the protocol prescribes it.
For example, in H.323 version 4, an empty list of alternate gatekeepers is
used to signal to the endpoint that it should delete any list it may have
previously received.  Usage like this, which has a well-prescribed meaning,
is useful.  However, empty lists in places that make no sense will probably
do nothing but create confusion.  In any case, entities must be prepared to
decode them.

Paul

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
For help on this mail list, send "HELP ITU-SG16" in a message to
listserv at mailbag.intel.com