home *** CD-ROM | disk | FTP | other *** search
- Path: news.gate.net!not-for-mail
- From: dhaire@gate.net (doug haire)
- Newsgroups: comp.dcom.modems
- Subject: Re: Is USR going to support 42bis+ on future courier upgrades?
- Date: 25 Mar 1996 08:52:59 -0500
- Organization: CyberGate, Inc.
- Message-ID: <4j68fr$14go@navajo.gate.net>
- References: <4j2fv1$8kf@nnrp1.news.primenet.com> <4j2iun$a3t@newsbf02.news.aol.com> <4j3r1f$1tc4@seminole.gate.net> <4j43mu$m9t@freenet-news.carleton.ca>
- NNTP-Posting-Host: navajo.gate.net
- X-Newsreader: TIN [UNIX 1.3 950824BETA PL0]
-
- Anthony Hill (an171@FreeNet.Carleton.CA) wrote:
- :
- : doug haire (dhaire@gate.net) writes:
- : > RobertN141 (robertn141@aol.com) wrote:
- : > : I find this information very interesting. My company manufactures the
- : > : AMQUEST HyperModem V.34. This product will support up to 230,000 BPS
- : > : throughput and has a enhanced microcontroller architecture to better
- : > : compress
- : > : and decompress data. The fundamental problem with many modems is that
- : > : they are all rated 28.8 V.34 115,200 max. However, they lack the
- : > : horsepower
- : > : to handle above 88,000 bps throughput. Since in the past many people were
- : >
- : > More hype.
- : >
- : > Listen, first of all, I'd say V.34 modems (and even V.32bis modems) are
- : > fully capable of handling a throughput speed above 88k. In fact, I know
- :
- : Try running bi-dirrection transfer with both DTEs set to 115.2kbps
- : and apair of modems using the Rockwell controller. You'll find that they
- : do tend to have problems keeping up.
-
- Unfortunately (or maybe fortunately), I don't have two Rockwell based
- V.34 modems to test this. I have noted that my Couriers do slow down on
- bi0directional transfers of highly compressible files but I have no way
- to compare to a paired Rockwell set.
-
- : > this is true. Second, I'd like to point out that comm overruns are a
- : > fault that lies with the operating system software of the CPU. I recently
- : > tested this on a comparison with a large file and two different platforms
- : > (MS-DOS and Linux) on the same CPU. The sender was a 486dx4/100 CPU running
- : > MS-DOS and using a USR Courier V.34 v.everything. The receiver was a
- : > 486sx33 with another Courier (same model) running linux and MS-DOS
- : > (separate partitions, dual boot). Modems were connected via an 8 ft
- : > telephone cable and synched at 33,600 bps with V.42 LAPM and V.42bis
- : > compression.
- : >
- : > The linux platform received the file with no errors, no comm overruns, no
- : > garbled subpackets, no problems.
- : >
- : > The MS-DOS platform received constant errors such as comm overruns and
- : > garbled subpackets.
- :
- : It isn't so much a case of what operating system you're using, but
- : what sort of software is talking to you're com port. Dos doesn't come
- : with any useful com drivers, so everyone writes their own, to varying
- : degrees of success. A really well writing DOS com driver can run at up to
- : 115.2kbps with a 16450 UART without getting any errors. Of course, well
- : writing DOS com software can be kinda rare. Linux, on the other hand,
- : does use a com driver (like just about every other OS out there). Again
- : though, overruns may occure with one com driver and not the other.
-
- Generally speaking, the DOS drivers suck big time (I just love technical
- terms). I used the linux platform as a comparison. I ran other stuff in
- the background while transferring these files on the linux system
- (installing software on another login) and got no glitches whatsoever. I
- did no multi-tasking on the DOS platform at all.
-
- I think it's pretty clear...
-
- : > When the common computer software platform is capable of handling 115200
- : > properly perhaps we can then consider the 230k UART speed.
- :
- : Software IS capable of 115.2 speeds now. In fact, with the right
- : hardware it's capable of MUCH higher DTE speeds then that (eg Hayes ESP
- : cards will run at up to 900kbps or so without overruns). The problem is
- : that 16550 UARTs can't really go faster then 115.2kbps. Both 16650s and
- : 16750s can, and both of those chips have on board flow control, which will
- : completely eliminate com overruns.
-
- Well, we agree to some extent here (I say it isn't just the hardware but
- the software as well) and we also agree that it isn't necessary to open a
- port higher than 115200 in the "real world" (I read your other post
- addressing this).
-
-