T.R | Title | User | Personal Name | Date | Lines |
---|
1241.1 | no change | TOOK::CALLANDER | Jill Callander DTN 226-5316 | Fri Jul 12 1991 17:03 | 5 |
| The SRM leads the implemenatation, and at this point WITH in the SRM
isn't leading by much...there are no definitions on how WITH is to be
passed across the call interface available at this point in time.
jill
|
1241.2 | | TOOK::STRUTT | Management - the one word oxymoron | Fri Jul 12 1991 20:14 | 16 |
| and there won't be more details until we are closer to getting an
implementation of it.
The SRM reference to WITH support in in_entity is in the earlier
chapter(s) which set the stage. The later chapters detail closer to
what is implemented. As another example, Chapter 2 refers to
distributing MMs across machines - support is not in place in later
chapters or the implementation.
If you have a need for WITH to be implemented "properly", please let
us know (via mail, not this conference, please) and provide suitable
supporting information. Honestly - we know how to implement it - it's
just a simple matter of implementation (along with 1000 other things
that are just as, or more, important) :-}
Colin
|
1241.3 | Requirement could become real within a few months. | KETJE::PACCO | | Tue Jul 16 1991 13:09 | 14 |
| Although the "need" here is in an early stage, it's during the
evaluation by a potential Strategic Vendor that this question was
raised.
The vendor is evaluating how to implement an access and functional
modules to support "circuit switched" type of (sub)networks. In order
to cleanly implement certain functions in a FM, the WITH clause should
be executed as closely as possible by the network element. In this
case the AM is a better place than the FM itself. However the real
question was about the syntax and the support therefore from DECmcc.
I have no further details yet on the exact projected use of this
clause, but will try to put here a more detailed case in some time.
Dominique.
|
1241.4 | I second the need | TENERE::MCDONALD | | Thu Jul 18 1991 05:22 | 1 |
| The WITH qualifier is also needed for some PNMP requirements.
|
1241.5 | Needed for bridge management | MARVIN::COBB | Graham R. Cobb (Wide Area Comms.), REO2-G/H9, 830-3917 | Fri Jul 19 1991 15:47 | 5 |
| Setting up bridge filtering using CMIP Bridge management (as implemented in
Hastings) really requires that action directives can be issued with WITH
clauses.
Graham
|
1241.6 | not hard to do | NAC::ENGLAND | | Mon Jul 22 1991 18:54 | 3 |
| NCL already supports WITH clause in action verbs. Although its
priority is low, so is its cost.
|
1241.7 | your opinion | TOOK::CALLANDER | Jill Callander DTN 226-5316 | Wed Jul 24 1991 17:57 | 10 |
| Ben,
since you are familiar with MCC, and somewhat with the FCL, I was curious
as to how you thought I might be able to pass down the with clause to the
lower level modules.
thanks
jill
(BTW, Ben the ACLs are almost all done, hopefully by tommorrow)
|
1241.8 | amnesia | NAC::ENGLAND | | Thu Jul 25 1991 17:22 | 14 |
| Jill, thanks for trusting my memory of MCC that much, but I'm not so
sure that I remember anymore. However, I'll take a wild guess.
There was at one point a separate with clause parameter to MCC$CALL,
I don't see why that couldn't be used now to get the with Clause to the
AM. I had thought that the whole point of the With clause was that it
would be passed down to the AM, which would then either pass it on to
the remote agent or implement the WITH filtering directly. It's been
a LONG time. Seems to me that Jim Carey would know the answer to this
question better than I.
The only fix to NCL was to parse the with clause attribute as an
attribute, not as an action argument (obvious, isn't it).
Let me know if you need help testing it or anything.
|
1241.9 | a bit different | TOOK::CALLANDER | Jill Callander DTN 226-5316 | Thu Jul 25 1991 18:07 | 13 |
| As it stands now the FCL handles complex with clauses, allowing multiple
attribute "expressions" to be AND and OR together. There used to be a place
in the SRM that said the WITH would be passed down in the q_in component
but now there is talk about moving into an "entity filter" bound to the
AES spec. My main concern is how to I (the PM) express to you (the AM)
a complex expression? I would hate to have to pass down a complex ILV
encoding (like the attribute list structure) just to tell the AM what
they are to use as the WITH...more discussion in this area is definitely
needed, and the decmcc developers conference is probably a better place
for this discussion.
thanks
jill
|