home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
kermit.columbia.edu
/
kermit.columbia.edu.tar
/
kermit.columbia.edu
/
newsgroups
/
misc.20020314-20021006
/
000187_fdc@columbia.edu_Sun Jun 30 13:53:44 EDT 2002.msg
< prev
next >
Wrap
Text File
|
2002-10-06
|
2KB
|
47 lines
Article: 13493 of comp.protocols.kermit.misc
Path: newsmaster.cc.columbia.edu!news.columbia.edu!news-not-for-mail
From: fdc@columbia.edu (Frank da Cruz)
Newsgroups: comp.protocols.kermit.misc
Subject: Re: REPEAT COUNTS
Date: 30 Jun 2002 13:53:32 -0400
Organization: Columbia University
Lines: 30
Message-ID: <afngis$84o$1@watsol.cc.columbia.edu>
References: <3D1DCAEE.18B634CC@peoplepc.com> <afkrr0$alp$1@watsol.cc.columbia.edu> <3D1F3AFE.FC0D1092@peoplepc.com>
NNTP-Posting-Host: watsol.cc.columbia.edu
X-Trace: newsmaster.cc.columbia.edu 1025459613 22878 128.59.39.139 (30 Jun 2002 17:53:33 GMT)
X-Complaints-To: postmaster@columbia.edu
NNTP-Posting-Date: 30 Jun 2002 17:53:33 GMT
Xref: newsmaster.cc.columbia.edu comp.protocols.kermit.misc:13493
In article <3D1F3AFE.FC0D1092@peoplepc.com>,
Jack Patteeuw <jjpatteeuw@peoplepc.com> wrote:
: Frank da Cruz wrote:
: > It's a byte-oriented variation of run-length encoding. Simple but
: > sometimes remarkably effective. Like everything else in Kermit, the user
: > can control it -- turn it on or off, choose the repeat-count prefix, etc,
: > although there is rarely a need to bother. It's on by default.
:
: You guys keep making it easier and easier !!
:
: Speaking of "High-Speed Kermit FIle Transfer", we have been doing some
: benchmarks and the defaults for C-Kermit 7.x and K95 2.x seem to have all
: of the "tricks" discussed in Chapter 12 turned on. However, when
: downloading out test case (lptest 127 8192) from a VAX/VMS V7.2, STA
: /VERBOSE says it used 0 windows ? Why ? And is there anything we are
: missing to get maximum performance ?
:
When using modern Kermit versions (like the ones you have), the protocol
automatically switches to streaming mode instead of sliding windows if
either of the Kermit programs knows that it has a reliable connection:
http://www.columbia.edu/kermit/ckermit70.html#x4.20
Even when sliding windows are used, the receiver is not conscious of them
unless there have been transmission errors, and the sender uses only as
many window slots as it actually needs. The sender's window grows
only when there is a long round-trip delay, or when there are
transmission errors.
- Frank