home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Columbia Kermit
/
kermit.zip
/
newsgroups
/
misc.19950528-19950726
/
000244_news@columbia.edu_Thu Jun 29 16:41:56 1995.msg
< prev
next >
Wrap
Internet Message Format
|
2020-01-01
|
4KB
Received: from apakabar.cc.columbia.edu by watsun.cc.columbia.edu with SMTP id AA27745
(5.65c+CU/IDA-1.4.4/HLK for <kermit.misc@watsun.cc.columbia.edu>); Thu, 29 Jun 1995 15:16:14 -0400
Received: by apakabar.cc.columbia.edu id AA20020
(5.65c+CU/IDA-1.4.4/HLK for kermit.misc@watsun); Thu, 29 Jun 1995 15:16:12 -0400
Newsgroups: comp.protocols.kermit.misc
Path: news.columbia.edu!panix!tinman.dev.prodigy.com!prodigy.com!newsjunkie.ans.net!howland.reston.ans.net!ix.netcom.com!netcom.com!jhurwit
From: jhurwit@netcom.com (Jeffrey Hurwit)
Subject: Re: Configurable APC checking in next MSK release?
Message-Id: <jhurwitDAy11x.7Jq@netcom.com>
Organization: Organization? What organization?
X-Newsreader: TIN [UNIX 1.3 950520BETA PL0]
References: <jhurwitDAuowt.ICo@netcom.com> <1995Jun29.071144.54932@cc.usu.edu>
Date: Thu, 29 Jun 1995 16:41:56 GMT
Lines: 54
Sender: jhurwit@netcom11.netcom.com
Apparently-To: kermit.misc@watsun.cc.columbia.edu
Thanks to both Frank and Joe for your comments.
In article <1995Jun29.071144.54932@cc.usu.edu>,
Joe Doupnik (jrd@cc.usu.edu) wrote:
> The reason OUTPUT is on the dangerous list is this. Something
>on the host sends APC OUTPUT \26DEL *.* ST to your desktop machine. That's
>includes a form of nasty mail trouble (^Z to suspend currrent process,
^^^^^^^^^^^^^^^^^^^^^^^^^^
>start removing files).
Hmmm, I see your point. I can see some visciousness on the part of
someone (similar to ANSI bombs, right? Someone would have to know
that I'm using Kermit), but I'm not familiar with the kinds of
system vagarities that would accidently trigger an unwanted APC
command that would in turn cause damage on the host account. I'll
take your word for it!
> I'm not sure we want to itemize commands which are dangerous
>because it is a bulky operation in the program and it adds to the doc
>complexity (must now explain how each candidate can be abused remotely
>etc).
I didn't know it would bloat the program that much. Having only
recently stepped up from an 8088 portable with no hard drive, I
really appreciate your continuing to support antiquated equipment
with minimal resources!
Ok, then let me try this suggestion: Currently, MSK will ignore a
'set apc unchecked' command in a script or macro if it was invoked
with an APC command. How about reversing this so that MSK will
recognize and act on that command? What I envision is placing a
'set apc unchecked' just before a dangerous command, and 'set apc
on' just after. This way, the window of opportunity for a system
burp setting off an unwanted command would be small, and an "APC
bomb" in an e-mail message could do nothing at all without knowing
the name of the script or macro.
In fact, better yet, how about an "unchecked' command, to be used
only in scripts or macros? Placed just before the dangerous
command (eg. 'uncheked output xxxx'), it could disable APC checking
just for the duration of that command, and reestablish it
immediately afterward.
For the docs, you could simply list out the commands which are
checked, following a brief and general discussion of the reasons
for checking them. More than that would not be necessary if the
user has the option of an 'unchecked' command.
How does this sound? Would it add too much to the size and
complexity of MSK? I think it could provide a good balance between
safety and flexibility if properly used.
Jeff