Subject: Addendum to ../70-058/ne.html

 

Thanks for your great site, seems like a great help to me.

I'm preparing for my first exam: Net. Essen. and this afternoon our teacher explained us somethings about SAP's (p. 179 / 177 in 2nd ed.).

 

There appear to be several SAP's around in IT-land. Of course there is SAP as the producer of R/3, but even in networking-country there are at least 2 very different SAP's.

The one that's mentioned on your ../70-58/ne.html-page, is indeed a part of Novell Netware. This SAP is the 'Server Advertising Protocol', a way for Novell servers to make themselves known to each other. This, however has got (virtualy) nothing to do with the SAP (=Service Access Points) from the Net. Essen.-book.

Service Access Points are basicly the communication-points between the different OSI-layers.

In the OSI-standard not only the layers are defined, but also the way in wich they should communicate with the one directly on top and the one below it. (Which is actually quite logical: if you want to replace a specific implementation of an OSI-layer, for instance a specific NIC-driver in the MAC-sublayer, with a different one, the one thing that needs most to be standardised is the communication with the layer on top of it and the layer immediately below it. Whatever, ...)

These standardised points for communication between the layers are called the SAP's.

Yet, the only protocol which actually implements these SAP's (between all layers) is the OSI-protocol. But this protocol is hardly ever used (as far as I know) and is certainly not extensively explained in the MS Net.

Essen.-book. The other protocols have not realy defined these SAP's in the way OSI suggests, but they do have communication points between their layers (of course).

So, just as people are used to map other protocol(-stack)s, like TCP/IP, to the OSI-model, the OSI term SAP is used here for these comm. points.

In my opinion, this place is not even chosen badly to introduce this term.

The MAC/LLC-sublayers with their standardised NDIS/ODI-drivers, but with NIC-specific drivers is actually quite a good place to explain the principle of what SAP's are. However, MS seems to have made a very poor job of doing this explaining.

(At least I didn't understand a thing of what was going on until I got a decent explanation.)

 

I hope this helps you in even further improving your excellent web-site.

 

Btw. It might be a good idea to dedicate a page to corrections of the mistakes and errors in the MS-book. If you like, we can mail you several of them (at least from the 2nd edition, that we use in our course).

Please go on with the great job your doing.

 

Corné