Path: utzoo!utgpu!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!ames!sgi!vjs@rhyolite.SGI.COM From: vjs@rhyolite.SGI.COM (Vernon Schryver) Newsgroups: comp.dcom.modems Subject: Re: Microport 3.0e and the Telebit Trailblazer Summary: minor corrections Message-ID: <23998@sgi.SGI.COM> Date: 27 Dec 88 19:21:38 GMT References: Sender: daemon@sgi.SGI.COM Organization: Silicon Graphics, Inc., Mountain View, CA Lines: 41 In article , eric@snark.UUCP (EricS.Raymond) writes: > ... > tb1200 =W-, "" \d\K\dATE0 OK ATS92=0S50=2S95=0DT\T CONNECT\s1200 It may help to use "DTW" here ............................^^ to wait for carrier. This is particularly handy if the line might sometimes be used by humans. That is, as far as I can tell, Trailblazers are not like Hayes 2400 & many Hayes-compatible modems which always wait for dialtone. TB's are thus more flexible, if somewhat less friendly to the unsuspecting. This suggestion is just a nit, but adding the "W" to my home computer's Dialers file restored a little domestic tranquility. Why insist on the "1200" or other speed indication from the modem? (That's the "\s1200".) If the other site changes modems, if you don't insist on checking the speed, you might still get thru. > ... > The ACUTBC entry may be better for mail and news feeds, as it enables data > compression for up to a 2:1 cut in transmission time. Measurements in this part of the world seem to contradict some of this. Compression in the modem is a Very Good Thing if the text has "low entropy," which is often the case with mail. However, I have found that shipping compressed files such as news batches is about 20% slower if you let the modem try to compress. That is, between fast machines using RTS/CTS flow control, I have measured 950-1050 B/sec thru UUCP with compression enabled, but 1200-1450 with it disabled. This is consistent what the CTS light does when transmitting compressed files. I've tried this among 80386, 68020, and MIPS-Co based machines from several vendors, including my home machine running 3.0e+DOSMerge. It would nice to know if the speed loss is because the TB is too busy trying to compress, or if the "compression" causes expansion. Of course, measurements of files smaller than 50-100KB are uninteresting, because of the buffer in the TB, as well as because of the sloppiness of UNIX timing. Vernon Schryver Silicon Graphics vjs@sgi.com