home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.org.eff.talk
- Path: sparky!uunet!gatech!emory!nastar!phardie
- From: phardie@nastar.uucp (Pete Hardie)
- Subject: Re: Beneficial Virus?
- Message-ID: <1993Jan12.150940.10789@nastar.uucp>
- Organization: Digital Transmission Systems, Duluth, GA.
- References: <19530@mindlink.bc.ca>
- Date: Tue, 12 Jan 1993 15:09:40 GMT
- Lines: 63
-
- In article <19530@mindlink.bc.ca> Clayten_Hamacher@mindlink.bc.ca (Clayten Hamacher) writes:
- >How many people do you know who have a unix workstation at home that runs the
- >same executables as the one at the office? Some versions of the same
- >workstation won't even run each other's executables so for Joe to take an
- >executables file home with him is VERY rare.
-
- What about multi-user PCs? The convergence of workstation power and PC power
- is leading to more PC-type machines being used at home and at work, and
- Intel binaries are generally compatible.
-
- >>You are assuming that every user can (and should) have all his/her files
- >>locked up tighter than a drum, on a multi-user system. The first is
- >possible,
- >>but the second is debatable.
- >
- >Almost all multi-user systems I have been on have had write permission denied
- >as a default.. Even on systems that don't (considering a different version
- >would need to be done for each OS) the virus could be restricted so it looks
- >in YOUR root directory (I can't see why it would do otherwise) for the marker
- >file, and will not do anything unless the marker file AND the executable
- >you're running belong to you. That's incredibly easy.
-
- Write permission to where? System directories, certainly, on UNIX systems.
- But again, what about MS-DOS derived systems?
-
- This is part of the problem - these things are being mentioned well after
- the initial description of the BCV, and had Richard Powers gone ahead and
- written the BCV, it might not have all this validity checking, and would
- instead install in the system root directory, and infect any executable,
- regardless of ownership.
-
- >This way the sysadmin could install the virus for bin (who usually has root
- >privs, or very near) and it would just pack the files in /bin.. Considering a
- >user has absolutely no need to look at the executable then this would be a
- >100% legit way for the sysadmin to pack the files without being obligated to
- >tell you.
-
- Yep. But suppose the 'ownership check' was flawed, and merely checked if
- you could write to a file - then this would infect all executables if it
- was installed as root.
-
- >>It is the assumption that you (or the virus writer) know better than I what
- >>is beneficial to my system/files that is the problem.
- >
- >Where did you get this idea? If that was the thought of the virus writer then
- >they wouldn't make it dependant on a marker file.
-
- It is obvious from the context. If all the writer wants is to compress his
- own executables, without having to explicitly do it, he can write a batch
- job, or device driver, that will do so on his machine, (or his account), and
- none of this 'virus' discussion would be going on. But once it is made into
- a viral program, it can travel to other systems and infect them, and then
- it begins to affect the programs of others, and at that point the definition
- of 'beneficial' becomes relevant - I may considered it harmful to compress
- my executables.
-
-
-
- --
- Pete Hardie: phardie@nastar (voice) (404) 497-0101
- Digital Transmission Systems, Inc., Duluth GA
- Member, DTS Dart Team | cat * | egrep -v "signature virus|infection"
- Position: Goalie |
-