Re: PPP

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



I kinda working on my own tcp/ip lib  and this is how I interprete it.

Your dumb terminal scripter makes connection

that activates PPP (with LCP confsync)

if that get an IP and return good then you can splat (encapulate)
IP/TCP/UDP packets 
out the line

er. and I must warn you I havnt got a working version so dont listen to me,
I am a techno moron.

Why do they call it TCP/IP  ?   that sound reversed. it should be
IP/TCP-UDP   as that makes sense in 
my head.


At 02:25 AM 3/1/02 -0500, Bill Cunningham wrote:
>I have received several responses and most people say it's in the data
>layer, and a couple of people think it's in the network layer. I don't
>really pay much attention to the OSI model, I think it complicates the
>complicated. I try to focus more on TCP/IP. Does PPP establish a link, then
>terminate, or continue throughout session in UDP and TCP? I posted this
>question on the PPP mailing list with less familiaritive response than ietf
>general list.
>----- Original Message -----
>From: "Brian Lloyd" <brian@lloyd.com>
>To: "Bill Cunningham" <billcu@CITYNET.NET>
>Cc: <ietf@ietf.org>
>Sent: Thursday, February 28, 2002 11:52 AM
>Subject: Re: PPP
>
>
>> At 03:55 AM 2/28/2002, you wrote:
>> >In what layer is PPP in the TCP/IP suite?
>>
>> I have read some of the other responses and it reinforces my belief that
>> most people don't understand PPP's relationship to IP and either the
>> 5-layer (internet) or 7-layer (ISO) models.
>>
>> PPP is really both the link and lower network layers. (The ISORMites
>> discovered that layer three was really several layers in itself but found
>> it difficult to say that the 7-layer model was really a 9-layer model so
>> they created sublayers, i.e. layers 3A, 3B, and 3C.  Something about
>> Padlipsky comes to mind here.) The best way to think of PPP is a
>degenerate
>> network of two nodes, not a link between two devices.  If you think of it
>> in this way, things like multilink and L2TP begin to make some sense.  The
>> problem occurs when people forget this.
>>
>> The way that I think of it is that the LCP negotiation represents
>> configuration of the link layer while the NCP negotiation configuration at
>> the network layer.
>>
>> And I continue to kick myself for allowing negotiation of multilink as
>part
>> of LCP instead of doing it after authentication.  I fear that this helped
>> screw up L2TP too.  I admit I caved to people who were worried about how
>> long it took PPP to complete negotiation, something that just isn't very
>> important.
>>
>>
>> Brian Lloyd
>> brian@lloyd.com
>> +1.530.676.1113 - voice
>> +1.360.838.9669 - fax
>>
>
>


[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Fedora Users]