home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.dcom.modems:11252 can.uucp:179
- Newsgroups: comp.dcom.modems,can.uucp
- Path: sparky!uunet!wupost!csus.edu!netcomsv!ntg!dplatt
- From: dplatt@ntg.com (Dave Platt)
- Subject: Re: UUCP 'g' vs. MNP & V.42
- Message-ID: <1992Jul27.172137.25556@ntg.com>
- Organization: New Technologies Group, Inc. Palo Alto CA
- References: <1992Jul20.052318.29102@zooid.guild.org> <1992Jul23.180658.20724@eci386.uucp> <1992Jul24.052107.1990@zooid.guild.org>
- Date: Mon, 27 Jul 1992 17:21:37 GMT
- Lines: 23
-
- In article <1992Jul24.052107.1990@zooid.guild.org> ross@zooid.guild.org (Ross Ridge) writes:
-
- >I've seen no benefit from using either v.42 or v.42bis with UUCP.
-
- My experiments over the last year lead me to the following conclusion:
-
- [1] If your uucp 'g' implementation runs with the default window size
- (three 64-byte packets), both MNP and V.42 error control hurt
- throughput quite a bit and should not be used. The error-control
- latency leads to protocol stalling.
-
- [2] If your 'g' implementation runs with a larger window size (e.g. seven
- packets) or supports larger packets, V.42 can often lead to some
- modest performance improvement (5-10%).
-
- [3] If [2] is true, and if you're sending mostly email messages, then
- using V.42bis compression on top of V.42 error control can lead to
- even more performance improvement (20%-40%). If you're sending
- compressed newsbatches, V.42bis doesn't buy you anything.
- --
- Dave Platt VOICE: (415) 813-8917
- Domain: dplatt@ntg.com UUCP: ...netcomsv!ntg!dplatt
- USNAIL: New Technologies Group Inc. 2468 Embarcardero Way, Palo Alto CA 94303
-