NetNews Usenet Archive 1992 #16/NN_1992_16.iso/spool/comp/dcom/modems/11014

Text File | 1992-07-23 | 2.7 KB | 51 lines

Newsgroups: comp.dcom.modems
Path: sparky!uunet!sun-barr!cs.utexas.edu!torn!cunews!hobbit.gandalf.ca!dcarr
From: [e-mail] (Dave Carr)
Subject: Re: Modems, compression, and SLIP or PPP (was Re: Using Telebit PEP with SLIP)
Message-ID: <[e-mail]>
Organization: Gandalf Data Ltd.
References: <[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]>
Date: Thu, 23 Jul 1992 12:57:50 GMT
Lines: 40

In <[e-mail]> [e-mail] (Ed Hall) writes:

>In article <[e-mail]> [e-mail] (Jim Shankland) writes:
>>In article <[e-mail]> [e-mail] (Bob Sutterfield) writes:
>>[Lots of interesting stuff, the upshot of which is:]
>>
>>Thanks; that's very useful information.  Does anybody have any thoughts
>>about SLIP with v.42bis and V-J header compression?  Does v.42bis
>>eliminate the need for V-J header compression?  Or vice versa?
>>Would it be a mistake to use both?  And what about PPP vs. SLIP?

>No.  V-J header compression only affects TCP/IP's protocol-specific
>information, which it compresses much more than any general compression
>algorithm like V.42bis.  The latter will do a good job of compressing the
>rest, if it is compressible.  So the two are complementary.

>Some modems seem to have increased latency when V.42bis is used, so there
>might be some specific applications where turning it off would increase
>performance with such modems.  But generally V.42bis is a win with SLIP or
>PPP.  When in doubt, try it both ways /with your particular application(s)/.
>Simple benchmarks are useless in this case.

>PPP has a higher "overhead" than SLIP, since it adds its own header and
>checksum.  The checksumming and more robust framing is of only marginal
>use when the connection is near-error-free, such as with V.42 or MNP4, and
>does add a small amount of latency.  But there are other aspects of PPP
>which makes it useful: it allows for better security and greater
>flexibility for dialup use, and it allows for the selective escaping of
>control characters (such as XON/XOFF)--quite useful when going through a
>switch or in situations where hardware flow-control isn't possible.
>And it supports non-IP protocols.

You should turn off MNP4 and/or V.42 when running PPP.  Also, PPP can
optionally turn off the address and control fields.  Compression over
PPP was a hot topic at the IETF-PPP meeting last week in Boston.

--
Dave Carr               | [e-mail]       | It's what you learn,
Principal Designer      | TEL [phone]     | after you know it all,
Gandalf Data Limited    | FAX [phone]     | that counts.