Subject: SDL-News: Comments about SDL-Algorithimcs
From: Sanjiv Agarwal (agarwal1#earthlink.net)
Date: Wed Mar 18 1998 - 18:58:36 GMT
Hello philippe/all sdl users,
This mail is in reply to philippe's request for the proposed upcoming SDL extensions.
It's good that the commitee is regularly working for future upgrades but i think that the
proposed extensions will not provide any needfull change to SDL functionality.
It is better if we do not mix the GR with text. The proposed extension will have its own pros and cons.
PRONS: It is sure that by implementing the proposed extensions we can get rid of Labels/Connect, which
will increase the readability of the GR file.
CONS: This might make the life more harder for the vendors working in the directions of automatic code generators.
This is sth to be discussed.
It would be nice if SDL std's can work something in the direction of "ORGANIZATION OF CONTEXTS".
This means following things:
1) STATIC PROCESS CTX's.
2) DYNAMIC PROCESS CTX's.
Some work can be done in the direction of defining the data structure of the context's. For e.g.,
Consider a call processing unit with multiple static entities (Time slots's, User Port's, PCM's etc). These static entities are created during the startup of the exchange and their life exists till the shutdown of exchange. For every call initiated inside the exchange these resources are reserved and later released. That means there is always some static data available for each resource defining the state of this resource. And whenever a call is initiated, the system will create the dynamic process contexts for each required resource which will work on the previously created STATIC context. At the end of call these dynamic ctx's will be released but the STATIC always exist and telling about the latest status of resource.
Now in this scenario if some SDL std's are available concerning the STATIC/DYNAMIC ctx then it will help in solving lot of issues of telecom industry.
For example., for every process ctx we can define :
1) some SDL area (set of bytes telling things like., some information used for message routing, info about PID's, info about SDL state, info about the msg received etc),
2) some area specific to the process (set of bytes used for defining process specific var's).
This kind of structure will help in solving lot of issuses like the one raised Raj Singh <singh#cig.mot.com> concerning the PID of the process sending a msg etc.
Please discard this mail if it does not make any sence to you.
This archive was generated by hypermail 2a23
: Sun Jun 16 2013 - 10:41:40 GMT
And if it does then feel free to contact me for any further details.
Senior Application Engineer
353 Kiely Blvd, ## E 304
San Jose, CA 95129
Fax : (408)260-2325
-----End text from "Sanjiv Agarwal"
This archive was generated by hypermail 2a23 : Sun Jun 16 2013 - 10:41:40 GMT