Re: SLIP and Cu-SeeMe cont'd

Paul Flores (
Thu, 4 Aug 1994 10:47:14 -0400

> On Tue, 2 Aug 1994, Geoff Rehn wrote:
> > At 5:26 PM 8/2/94 -0400, Jim B. Krone wrote:
> > I am still undecided as to whether or not to use:
> > 1. Compressed SLIP (ie. CSLIP) (ie. header compression)
> > 2. Compressed data ie. MNP-5 or V42bis data compression.
> >
> > But on CSLIP for header compression, I don't know.
> > I guess not. Does CU-SeeMe video data have headers, Cornell?
> >
> > But I think no CSLIP and no MNP5 or V42bis might give the best results.
> This is how CSLIP header compression works: Each IP packet coming to
> your machine or going out of your machien has a header on it. How CSLIP
> works is basically this:
> header represented in caps data in lower case:
> XXXX XXXX XXXX XXXX aaaa aaaa aaaa aaaa aaaa aaaa aaaa aaaa
> (this is your original packet)
> (the next packet coming down the pipe looks like this:)
> XXXX XXXX XXXX XXXB aaaa aaaa aaaa aaaa aaaa aaaa aaaa bbbb
> So, what CSLIP does, is instead of sending the entire header on each
> packet it just sends what the difference from one packte to the next is.
> So, in my example it would, instead of sending the entire header on the
> second packet, send something to the effect that everything was the same
> except bit 16 which is now B.
> CSLIP will do nothing but speed up your connection, because it saves
> data being transmitted. The worst thing that can happen is that you
> transmit a full header, which is what would happen anyway if you had
> CSLIP turned off.
> Educating the masses, (sha, as if!)
> -Rusty
> (aka Michael)

CSLIP is ok but PPP is better, both use VJ Compression on the TCP
headers, but PPP has more 'slots' to work with than CSLIP. Also seems
to be alot more bullet proof from noise, etc in my experience.

Paul Flores
Internet Services Manager, Phoenix Data Systems Inc
Dial Up/Dedicated Internet Access for Houston, Tx!!!!

Remember the Golden Rule :
He who has the gold, PAYS THE TAXES!!!!
(Unless you run for a prominent political post)