home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!usc!randvax!edhall
- From: edhall@rand.org (Ed Hall)
- Newsgroups: comp.dcom.modems
- Subject: Re: Boom! Our lawyers are tougher than your lawyers. You're Dead.
- Message-ID: <3627@randvax.rand.org>
- Date: 22 Jul 92 20:29:12 GMT
- References: <BrMHC9.Ct6@wsrcc.com> <1992Jul20.034928.14560@cubetech.com> <BrsyEw.Dst@wsrcc.com>
- Sender: news@randvax.rand.org
- Organization: RAND Corporation.
- Lines: 19
- Nntp-Posting-Host: ives.rand.org
-
- In article <BrsyEw.Dst@wsrcc.com> wolfgang@wsrcc.com (Wolfgang S. Rupprecht) writes:
- >Anyone relying on "<pause>+++<pause>" not occurring during a data transfer is
- >not writing good defensive code. Good programmers don't *hope* that a
- >certain set of characters and delays never happens. They turn off the
- >inband +++ escape kludge and use a good *out-of-band* escape.
-
- --If such is available.
-
- In datacomm we deal with probabilities and not certainties. Adding the
- guard times makes in-band escapes a few orders of magnitude more
- reliable. Sure, a reliable--and reliably /available/--out-of-band
- escape is to be prefered. But given the state of systems today, with
- almost-right serial ports and pretty-good serial drivers, in-band
- escapes might be the best option for a "good programmer" to select.
-
- Software which doesn't work on a particular system has /zero/ reliability.
-
- -Ed Hall
- edhall@rand.org
-