home *** CD-ROM | disk | FTP | other *** search
- Path: in2.uu.net!xenitec!zswamp!zswamp!geoff
- From: geoff@zswamp.UUCP (Geoffrey Welsh)
- Newsgroups: comp.dcom.modems
- Subject: Re: Is UUCP is critical feature for Unix machine?
- Date: Sun, 21 Jan 1996 00:05:56 EST
- Message-ID: <960121.000556.8w5.rnr.w165w@zswamp.UUCP>
- References: <4ckopg$6iv@mips.pfalz.de> <4d10p3$dsm@usenety1.news.prodigy.com> <DL94wH.231@bokonon.ussinc.com>
- Organization: Izot's Swamp
- X-Newsreader: rnr v1.28
-
- stephen@bokonon.ussinc.com (Stephen M. Dunn) writes:
- >
- > In article <4d10p3$dsm@usenety1.news.prodigy.com> davidsen@tmr.com (bill
- > davidsen) writes:
- > $This is true, but the chances of an error in the serial port is
- > $pretty minimal, while the link between modems is error checked.
- >
- > It depends, to a large extent, on the hardware. If you can
- > guarantee that you'll never lose a character on links between the
- > modems and their respective serial ports, great. I'm not quite that
- > trusting in general.
- >
- > In the special case of a standard PC serial port, even with a
- > 16550, I get a little worried in expecting it to be perfectly
- > loss-free. I've seen some (albeit somewhat underpowered - I'm
- > talking 386/40) systems which, even with a 16550 and a 16550-aware
- > OS (Xenix 2.3.4), would lose data at 38.4 kbps.
-
- I've found that intelligent serial cards are very good unless horribly
- misconfigured (duh, what does "stty -ortsfl -rtsflow -ctsflow -ixon -ixoff"
- mean?), but direct UART-based hardware (including, as you point out, a 16550)
- is indeed far too vulnerable to receive buffer overflows... and, of course,
- modems connected to terminal servers are infamous for flow control problems.
-
- --
- Geoffrey Welsh, Senior Developer, InSystems Technologies Inc.
- geoff@zswamp.uucp, [xenitec.on.ca|m2xenix.psg.com]!zswamp!geoff
- USENET has become the sanctuary of the Church of the Perpetually Injured Party.
-