home *** CD-ROM | disk | FTP | other *** search
- Path: nntp.snfc21.pbi.net!usenet
- From: JNavas@NavasGrp.com (John Navas)
- Newsgroups: comp.dcom.modems
- Subject: Re: Why 33600?
- Date: Tue, 16 Apr 1996 19:47:09 GMT
- Organization: The Navas Group of Dublin, CA, USA
- Message-ID: <3173f628.4860238@news.pbinet.com>
- References: <4kn9v9$n6e@solaris.cc.vt.edu> <4kok44$mj9@hg.oro.net> <317237dc.836868@snews.zippo.com> <4ksjvj$tlb@hg.oro.net>
- NNTP-Posting-Host: ppp-68-35.irvn11.pbinet.com
- X-Newsreader: Forte Free Agent 1.1/16.198
-
- [Posted to comp.dcom.modems, copy sent by email]
- estarry@oro.net (Ed Starry) wrote:
-
- >... I don't need a 28.8 modem when I
- >can transfer data at 6-7,000 cps with a 14.4.
-
- You can't, so I guess you do.
-
- >... You use a 28.8 with 10 lbs. of pressure and I use a 14.4 with 100
- >lbs. of pressure. There's a matter of compression, my 14.4 modem compresses
- >data and your 28.8 doesn't!
-
- Nonsense.
-
- > It never ceases to amaze me why people think a 14.4 won't go over 1,600 >cps, they must be living in a cave.
-
- No, just the real world, in which a 14.4 modem simply cannot go faster than
- about 1600+ cps (given protocol overhead) with compressed data.
-
- >There are 14.4 modems rated at 115,200 bps
- >and as difficult as it is to believe they will run at this rate.
-
- Of course, but that's the computer-to-modem speed -- the throughput is also
- limited by the modem-to-modem speed of 14,400 bps.
-
- >Any modem
- >will run at their advertised DTE rate if the serial port is cached and the
- >operator knows how to configure the system.
-
- There is no such thing as a "cached" serial port. Regardless, it is
- definitely possible to run a modem at 115,200 bps with a standard UART -- I
- do it all the time.
-
- >You've been told that modems
- >won't compress for such a long time you believe it. Compression is alive and
- >well for those of us that know how to make it work!
-
- The rest of us know that modems do support data compression, usually MNP5 or
- V.42bis; we know that compressed data (e.g., ZIP files) cannot be compressed
- further by modems; and we know that compression of real-world text files
- rarely exceeds about 2:1, or about 3400 cps on a V.42/V.42bis 14.4 Kbps
- connection.
-
- >Anyone that sets their
- >DTE rate below specifications is guilty of shooting themselves in the foot.
- >Are you running your modem at its rated DTE speed? I am! 14.4@115,200.
-
- There is no benefit from setting the DTE rate excesssively high -- 57,600 is
- more than enough for a 14.4 Kbps modem with real-world data.
-
- > A 14.4 modem running at 115,200 is no slower than a 28.8 running at >115,200.
-
- Nonsense.
-
- >PS: If you don't know what a cached serial port is and how they work don't
- >condemn them. ...
-
- I know that there is no such thing. Regardless, I am quite familiar with
- the Hayes ESP (which you claim is "cached," something that would a big
- surprise to Hayes, which makes no such claim).
-
- >... DTE is the measure of data flow and 115,200 is 115,200 whether
- >it be from a 14.4 or a 28.8!!!
-
- Of course -- but it won't help if the speed between the modems limits
- throughput to a much lower rate, as is the case with 14.4 modems.
-
- ===
-
- So are you willing to put your money where your mouth is? Here's:
-
- the Navas-Starry Challenge
-
- You use your 14.4 modem with whatever interface you want (Hayes ESP
- included). I will use a USR Sportster 28800 Vi with a standard 16550 UART.
-
- Both of us will identify one real-world file (i.e., not a synthetic test
- file) publically available on each of three different BBS systems or FTP/Web
- servers. The size of each file shall not be less than 250 kilobytes. We
- will each then measure the total time it takes to download all six (6) files
- with our respective modems using ZMODEM or FTP/HTTP transfer as appropriate,
- measuring the time with a stopwatch in each case from the download command
- (not the start of transfer) to download complete.
-
- Either of us may challenge the results of the other, in which case the tests
- will be repeated by a mutually agreeable third party with our hardware. I
- will concede that you are the winner if you can come within 10% of my time.
-
- The loser will give the winner five (5) of the modems used by the winner.
- Both of us will escrow sufficient funds to cover the cost with a mutually
- agreeable third party before the start of the contest.
-
- Fair enough? Now as the saying goes, either put up or shut up. ;-)
-
- --
- Best regards,
- John mailto:JNavas@NavasGrp.com http://web.aimnet.com/~jnavas/
- 28800 Modem FAQ: http://web.aimnet.com/~jnavas/modem/faq.html
-