Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!mcsun!ukc!slxsys!ibmpcug!robobar!ronald From: ronald@robobar.co.uk (Ronald S H Khoo) Newsgroups: comp.dcom.modems Subject: Re: Telebit and V.32bis Message-ID: <1991Jun6.193254.3886@robobar.co.uk> Date: 6 Jun 91 19:32:54 GMT References: <1991Jun1.063101.3896@netcom.COM> <1991Jun02.231341.3315@nstar.rn.com> <98879@lll-winken.LLNL.GOV> Organization: Robobar Ltd., Perivale, Middx., ENGLAND. Lines: 34 casey@gauss.llnl.gov (Casey Leedom) writes: > / From: larry@nstar.rn.com (Larry Snyder) > | > | yes using uucp v.32bis - 1970 cps with 16 bit compressed data PEP - > \ 1300-1400 cps and v.32 900-1100 cps.. > > Yet more reason to ditch PEP as soon as possible. It no longer holds > the preeminent UUCP throughput position it once did. Well, if all your UUCP connections are like Larry's, that seems reasonable. But the info required to make global decisions like that isn't complete yet. For example, PEP has always shown unusually good performance on poor and international phone lines. A modem vendor or two have posted here that their V.32 performance on such lines is fine. There has also been anecdotal evidence from users that V.32 stinks where PEP worked fine(ish). It really be nice to have some independently confirmed DATA on exactly what kind of performance PEP, V.32 and V.32bis gets over a wide range of different phone connections. There also seems to be a "quality of implementation" issue where using different V.32 implementations would further complicate the issue. I'm sure that most of the major modem manufacturers have this info (seeing as they tend to be the ones with a few of each of everyone's modems :-) but they would not be the best people to be giving impartial advice :-) OK, so how *do* we get this info? Maybe we ought to get a comp.dcom.modems super-test going ? -- Ronald Khoo +44 81 991 1142 (O) +44 71 229 7741 (H)