Thread View: comp.protocols.iso
2 messages
2 total messages
Started by Ralph Mackiewicz
Fri, 28 Aug 1992 13:52
OSI Failure?
Author: Ralph Mackiewicz
Date: Fri, 28 Aug 1992 13:52
Date: Fri, 28 Aug 1992 13:52
37 lines
2661 bytes
2661 bytes
In response to Ran (atkinson@itd.nrl.navy.mil): I stand by what I said. The companies I was referring to that used MMS because they were looking for vendor independent solutions were not vendors. They were users and we have many customers who are not General Motors. I agree that GM is the reason that many vendors of MMS implement it. And, because that is the only reason, they will never be successful selling the technology. What I wrote is true. Our sales of MMS are increasing not because our customers want to promote OSI because it would be the politically correct thing to do. They are buying it because it DOES solve problems for them that other networking technologies (like TCP/IP) don't solve. The reason that other networking technologies don't solve these problems is because they are focused on networking issues that do not relate to the functions provided by MMS. let me point out that I will be the last person to suggest that if you need a network for sharing files, E-Mail, or terminal connectivity that you should only use an OSI based network. Most of the arguments I have seen in this dicussion group against OSI are based upon the premise that anyone who is in support of OSI is also in support of OSI over any other technology regardless of the specific functions that are expected of the network. WRONG! The simple truth is: MMS is currently the only non- proprietary device to computer networking protocol that has been implmented in any commercially available products. It is in use today. It works. And it uses OSI technology. This last fact doesn't make MMS useless. It just makes it harder to sell because uninformed people will assume that just because TCP/IP, Novell, Banyan, etc. do a better job at file transfer, E-Mail, etc. than OSI, then TCP/IP must also do a better job for reading and writing variables in a controller, uploading and downloading programs in a controller and reporting device faults....and this is not true. If anyone other than my company would commit to supporting MMS over TCP/IP using RFC 1006, only then would TCP/IP become a viable alternative to OSI for performing these device communications functions in a non-proprietary manner. Ralph Mackiewicz, SISCO Inc. 2943439@mcimail.com or ics.ralph@ics.org
Re: OSI Failure? (MMS and TCP/IP)
Author: Andrew McRae
Date: Sun, 30 Aug 1992 09:32
Date: Sun, 30 Aug 1992 09:32
50 lines
3123 bytes
3123 bytes
From article <32y8PB1w164w@Control.Com>, by ics.ralph@Control.Com (Ralph Mackiewicz): > In response to Ran (atkinson@itd.nrl.navy.mil): > The simple truth is: MMS is currently the only non- > proprietary device to computer networking protocol that has been implmented > in any commercially available products. It is in use today. It works. > And it uses OSI technology. This last fact doesn't make MMS useless. It > just makes it harder to sell because uninformed people will assume that > just because TCP/IP, Novell, Banyan, etc. do a better job at file transfer, > E-Mail, etc. than OSI, then TCP/IP must also do a better job for reading > and writing variables in a controller, uploading and downloading programs > in a controller and reporting device faults....and this is not true. Why isn't it true? I don't understand this logic. Let's not mix apples and oranges. MMS is an _application_ level protocol, like FTP, telnet or FTAM, whereas TCP/IP is a _transport_ and _network_ level protocol. I don't see why TCP/IP can't be used instead of OSI layers for MMS. If TCP/IP is smaller, faster and lower weight than OSI, it seems to me that it is *better* for real time device control. I agree that no other standard protocol exists at the application layer that does what MMS does, and one reason why is that MMS suffers from kitchen-sink syndrome. It may work, but the cost in resources is pretty horrendous. It takes a lot of memory and processing power to support a full OSI stack, whereas I can run my TCP/IP stack on an embedded system in 40 Kb of memory on a 7 Megahertz 68000 (and that's using what some people believe is the bloated BSD code) over 802.4 token bus and still get tens or hundreds of Kbytes throughput. > If anyone other than my company would commit to supporting MMS over TCP/IP > using RFC 1006, only then would TCP/IP become a viable alternative to OSI > for performing these device communications functions in a non-proprietary > manner. Standardisation by fiat can fail miserably; IMHO MAP/MMS is a classic committee designed solution that was frozen prematurely. I believe protocols should evolve and reach maturity *before* they become standards. > Ralph Mackiewicz, SISCO Inc. > 2943439@mcimail.com or ics.ralph@ics.org Interesting discussion. These are difficult issues. Andrew McRae inet: andrew@megadata.mega.oz.au Megadata Pty Ltd, uucp: ..!uunet!megadata.mega.oz.au!andrew North Ryde 2113 Phone: +61 2 805 0899 NSW AUSTRALIA Fax: +61 2 887 4847
Thread Navigation
This is a paginated view of messages in the thread with full content displayed inline.
Messages are displayed in chronological order, with the original post highlighted in green.
Use pagination controls to navigate through all messages in large threads.
Back to All Threads