home *** CD-ROM | disk | FTP | other *** search
- Path: mail2news.demon.co.uk!genesis.demon.co.uk
- From: Lawrence Kirby <fred@genesis.demon.co.uk>
- Newsgroups: comp.dcom.modems
- Subject: Re: Astounding high CPS rate!
- Date: Tue, 16 Jan 96 21:32:29 GMT
- Organization: none
- Message-ID: <821827949snz@genesis.demon.co.uk>
- References: <4cvsii$h0@horus.infinet.com> <4db4ra$56@hg.oro.net>
- Reply-To: fred@genesis.demon.co.uk
- X-NNTP-Posting-Host: genesis.demon.co.uk
- X-Newsreader: Demon Internet Simple News v1.27
- X-Mail2News-Path: genesis.demon.co.uk
-
- In article <4db4ra$56@hg.oro.net> estarry@oro.net "Ed Starry" writes:
-
- > For properly configured modems compression of <*.zip> files is no big
- > deal.
- >Actually, zip file transfers of only 2,100 cps is slow for a 14.4!
-
- The raw bit rate of 14400bps corresponds to 14400/8 which is 1800cps
- (assuming the usual 8 bit characters). With EC framing overhead the
- actual maximum is about 1730cps (can't remember if that is for 128 or
- 256 byte packets). Anything above that means that either
-
- 1. compression is enabled and you are passing compressible data, or
-
- 2. The speed measurements or calculations are wrong.
-
- > Most, if not all, BBS's & ISP's 'Do Not' have compression enabled.
-
- If you're seeing that you need to fix your modem settings. I have yet to
- come across a BBS which has compression disabled (I've called a few in the
- USA as well as many in the UK).
-
- >It takes
- >*TWO* properly configured modems and with a BBS on one end you will never
- >see true high speed compression. Not knowing exactly what your set-up is
- >there's a high probability you've inadvertently stumbled onto what most
- >people believe to be impossible, 'Zip files ARE compressible!'
-
- Try rezipping a ZIP file sometime. You might get a 1-2% reduction if you
- are lucky and that most likely comes from compressing the ZIP internal
- catalogue which is not stored compressed. ZIP uses a far stronger
- compression algorithm than V.42bis and the latter really has no chance.
- What 'people believe' to be impossible is supported by well established
- compression theory. Read the comp.compression FAQ to find out more.
-
- >If you have
- >16550 UART's and are able to set the TxFIFOBuffer to 16 give it a try, now
- >send to yourself on the other PC. Depending on your modems capabilities and
- >their configuration you could see rates of 5,000 cps, or higher, for <*.zip>
- >files!
-
- Only if the ZIP file was created using the 'store' mode or a 2nd order
- degenerate file was originally compressed, or, I suppose if the files
- being archived were so small that the ZIP file consisted mainly of a
- catalogue.
-
- > Incidentally, internal modems are 'usually' better at compression than
- >externals. Not because of the 16550 but because internal modems buffer the
- >UART via 'Additional RAM'. An external modem has 'Zero RAM' unless using a
- >Hayes ESP card.
-
- This is complete rubbish. As long as the Modem's transmit buffer is kept
- reasonably full (it is often about 2-3K on both internal and external
- modems) V.42bis can work at full efficiency. The 'RAM' on an ESP is not in
- the modem and has no effect on compression in the modem.
-
- >LAPM modems are better than MNP, higher compression ratios.
-
- V.42bis usually compresses better than MNP5 (exception: uuencoded ZIP files).
- LAPM is not a compression protocol but it is true that V.42bis is most
- usually implemented on top of LAPM.
-
- --
- -----------------------------------------
- Lawrence Kirby | fred@genesis.demon.co.uk
- Wilts, England | 70734.126@compuserve.com
- -----------------------------------------
-