hxn8477%njitx.decnet@NJITC.NJIT.EDU ("NJITX::HXN8477") (10/24/88)
PERSONAL PROTOCOLS: WHERE DO THEY FIT IN THE BIG PICTURE
People have been asking me this question. Personal protocols (such as
kermit, xmodem, ymodem, <banana>modem, tlink, sealink, <banana>link,
cis a, cis b, cis <banana>, etc..etc.(where banana = any thing)), where
do they fit in the ISO reference model?
I say, well. If you are calling a node over the
phone network, as opposed to a packet switching network, then they
represent the link layer protocol. That is because they insure safe
data transfer over the logical link between the two calling points.
An example here is if you transfer data to or from a BBS.
But if you are calling a node over a packet switching network, then they
represent the transport layer. That is because they insure end-to-end
error and flow control. An example here would be transfering data with
Compuserve over its proprietary net or over Telenet or Tymenet. While
the link and network protocols are provided by the PSN, the protocol
you use provides the transport layer protocol.
.;;;. ..
/\ //^|^\\ .'..;
/ \ /\ | _/~~\_.'.'
/ \/ \ /\ | ,(_ ) "Blanked until further notice"
/ \ \/ \ | ' \~~~|'
/ \ \ \ | | |
+---------------------------+------------------------------------------------+
|Hamed Nassar |Internet : hxn8477%njitx.decnet@njitc.njit.edu |
|EE Department |UUCP : bellcore!argus!mars!nancy |
|NJ Institute of Technology |CompuServe: 74000,130 |
+---------------------------+------------------------------------------------+
------DEGROOT@hwalhw50.BITNET ("Kees de Groot ", DEGROOT@HWALHW50) (10/24/88)
Hamed Nassar wonders: > === personal protocols: where do they fit in the big picture === > People have been asking me this question. Personal protocols (such as > kermit, xmodem, ymodem, <banana>modem, tlink, sealink, <banana>link, >cis a, cis b, cis <banana>, etc..etc.(where banana = any thing)), where >do they fit in the ISO reference model? I would say they represent the application layer. There are two applications at both sides communicating with each other, more or less duplicating functions of the lower layers. If the lower layers provide more functionality you can choose a more appropriate protocol. Eg. if you use one of those ISDN-killing-modems which transport data error-free at astonishing speeds then you can use a simpler protocol. But if you have no error-handling what so ever then you should use a protocol that handles this itself. Modems as the Telebit trailblazer do 'protocol-spoofing'. They collect packets ACKing each packet directly to the host and send them in one long datapacket to the other modem which handles them one by one to the receiving host. I have always thought you could see this as a (confusing) implementation of the layers 1,2 and 7 (physical, data- and application layer). Comments? Tel. +31-8370- .KeesdeGroot (DEGROOT@HWALHW50.BITNET) o\/o THERE AINT NO (8)3557/ Wageningen Agricultural University [] SUCH THING AS 4030 Computer-centre, the Netherlands .==. A FREE LUNCH! X25: PSI%(+204)18370060638::DEGROOT DISCLAIMER: My opinions are my own alone and do not represent any official position of my employer. - if you go too far to the east, you find yourself in the west .. -
sjs@jcricket.ctt.bellcore.com (Stan Switzer) (10/25/88)
Kees de Groot ( DEGROOT@HWALHW50) writes: > Hamed Nassar wonders: > > > People have been asking me this question. Personal protocols (such as > > kermit, xmodem, ymodem, <banana>modem, tlink, sealink, <banana>link, > >cis a, cis b, cis <banana>, etc..etc.(where banana = any thing)), where > >do they fit in the ISO reference model? > > I would say they represent the application layer. ISORM is just a model. It is useful only to the extent that it helps to clarify and organize understanding. There is no sense force-fitting things to this model when it only serves to confuse. Knowing what "level" it is won't make it any more "standard" and it won't make it interoperate any better either. The protocols you mention have functionality assigned to several of the ISO levels, and are layered differently. So one answer is that they don't fit at all. When it comes down to using them with or over "standard" protocols, you will have to make certain map-or-wrap choices, though, and I suppose that you could say (if you must say it) that the protocols are onel level higher than where you started to "wrap" your protocol. Nevertheless, I do not find this approach very useful. And anyway, it's the OSI reference model. Speaking for myself, Stan Switzer sjs@ctt.bellcore.com