Return to BSD News archive
Newsgroups: comp.os.386bsd.misc Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!bunyip.cc.uq.oz.au!munnari.oz.au!spool.mu.edu!howland.reston.ans.net!usc!elroy.jpl.nasa.gov!decwrl!netcomsv!netcomsv!calcite!vjs From: vjs@calcite.rhyolite.com (Vernon Schryver) Subject: Re: Auto-dialing, Auto-redialing under SLIP [FreeBSD 1.1] Message-ID: <CuJ6vp.8rv@calcite.rhyolite.com> Organization: Rhyolite Software Date: Sun, 14 Aug 1994 15:26:13 GMT References: <RICK.94Jul29131023@vox.trystero.com> <32j131$1i0@news.bu.edu> Lines: 41 In article <32j131$1i0@news.bu.edu> heiser@bu.edu (Bill Heiser) writes: >Richard E. Nickle (rick@vox.trystero.com) wrote: > >: SLIP works just great. I don't even want to try pppd right now because it >: works so nice. I'd only play with PPP if someone would convince me that >: there is a tremendous performance/reliability gain to be had. Any thoughts? >I don't know if it is true or not, but I've read that PPP is "the >protocol of the future". Apparently, although I don't know the details, >there is some sort of "standard" that PPP adheres to (I wish I could >remember where I saw the article that described this :-) The same >article also stated that PPP is more robust than SLIP. > >P.s. I'm typing this across a SLIP connection from my FreeBSD "test box". >The performance of this SLIP connection feels significantly better in >terms of interactive response than that of my LINUX SLIP connection. SLIP uses 1 byte of framing, while async PPP uses at least 4. The resulting performance difference is not noticable, and in practice not measurable. As long as both PPP and SLIP are using header compression (cslip) and PPP is not using host-compression (CCP), they'll have practically identical performance on comparable systems, modems, and phone lines. On the other hand, v.32bis modems vary by a factor of ten (10) in how fast they carry SLIP or PPP data. About the only time I'd expect one system to be faster than another on the same modems is if one is more likely to drop bytes, and so cause TCP/IP timeouts and retransmissions. See RFC-1331 for PPP. It's no longer current, but it stands alone better than its replacements and searching for "Obsoletes.*1331" will find them. SLIP is as standardized as PPP, despite the "this is not a standard" words in the SLIP RFC and the Declaration From A Standards Committee That PPP is a Standards Track Recommend Standard. Only standard committee go-ers, lawyers, and fools care more about such Declarations than what is needed and what works. SLIP is less flexible than PPP, since SLIP can handle only IP, while PPP also can do Appletalk, IPX, and on and on. PPP installations can be easier because they negotiate some parameters that many people have trouble understanding (e.g. IP addresses). Vernon Schryver vjs@rhyolite.com