home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cis.ohio-state.edu!rutgers!uwvax!gorgon!fullfeed!ruth!rat
- From: rat@ruth.UUCP (David Douthitt)
- Newsgroups: comp.org.eff.talk
- Subject: Beneficial Virus
- Message-ID: <m503wB4w165w@ruth.UUCP>
- Date: 9 Jan 93 02:58:45 GMT
- Organization: Network XXIII - +1 608 222 9253
- Lines: 103
-
-
- I thought I would give some thought about how a beneficial virus would
- be implemented, so that we're all arguing about the same points.
-
- One assumption I will be making is that the system(s) are single-user
- MSDOS sites - multiuser viruses present a whole new can of worms.
-
- First, the user installs the software on their system (System A). They
- run a program called INSTALL.EXE on the provided distribution diskette.
- With appropriate warnings and notifications, the install program places
- the marker on the system. It is possible that it would (with notice)
- load the compression virus into memory at that time, so that it would
- start to spread - or a choice could be given to place the virus on
- a file or files, to be activated later (during decompression).
-
- One note to be made - the virus being discussed has one purpose: to
- spread the decompression routine to as many executables as it sees fit
- to. The decompression routine and the viral infection must be kept
- SEPARATE when considering the ramifications of this system. The decompression
- routine could in fact be LZExe or some other, or could even be released
- separately from the virus.
-
- A file with the virus (and compression code) would operate in this manner:
-
- 1. The executable is loaded into memory and run.
-
- 2. The first code operates this way:
-
- a. Does viral marker exist?
- - NO: take your pick of ONE (and ONLY one) of these
- possibile actions to be taken:
-
- 1. remove virus from file (leaving compression routines
- in place to decompress file later).
- 2. skip to step 3 - virus remains to activate on other
- system containing marker.
- 3. remove virus AND COMPRESSION from file.
- 4. ask the user what to do (and choose one of the
- above or possibly a new install).
-
- - YES: load the virus (either from file or marker) into
- memory as a TSR program - with hooks into disk
- or file read/write/execute.
-
- 3. Having executed past the viral code, the decompression routines would
- take hold and decompress the program - virus or no virus.
-
- 4. The actual code would then begin to execute.
-
- My favorite option under 2. (NO) is to remove the virus entirely. A virus
- would cause too much unnecessary panic otherwise from a misguided and
- paranoid public who have heard too many stories which brand hackers as
- vicious evil malcontents intent on destroying the world by computer.
-
- If the virus was contained in the MARKER, then when a compressed file
- leaves the system (such as taking a compressed executable from System A
- to System B) there is no virus at ALL present in the file, and the virus
- does not spread.
-
- Since the virus is not resident in the file, there is not even a chance
- of this virus "escaping" a system because the only way the actual code
- gets there is by user installation. If there is an option to install
- the virus under 2. (NO) 4., then the viral code must be present in the
- compressed file and somebody would probably scream.
-
- One possible problem would be a malicious virus masquerading as a
- valid marker file. Then any file compressed with the beneficial
- virus would activate the malicious virus - some sort of check would
- have to be made that the virus was the *RIGHT* virus.
-
- Having considered the infected file, let's consider the virus in operation.
- It is assumed that the virus has only been placed in memory after it has
- been verified that the marker was present, and it is okay for it to run.
-
- The virus might operate this way:
-
- 1. The virus recognizes the fact that an executable file is being loaded
- for execution. How this is done is not particularly relevant, and
- could most likely be done a number of ways.
-
- 2. The virus checks for several signatures in the executable (after it has
- been loaded, of course). This would check whether the file already
- had been compressed (with who knows what), or whether the file had
- the required marker check routine.
-
- 3. If the file is not compressed, it would write out to the file on
- disk, in order:
- - the marker check routine
- - the decompression routine
- - the compressed executable
-
- 4. It would then delete the original and rename the temporary file to
- be the original.
-
- Hope this clears up the viral debate - but I still hope it rages on! :-)
-
-
- --
- UUCP: ....!fullfeed.com!ruth!rat | Network XXIII - +1 608 222 9253
- InterNet: rat@ruth.UUCP | 80Megs+ of Usenet/Fido programs!
- Fidonet: David Douthitt 1:121/23 | ... Files via Fidonet FREQ,
- "...because appealing to the masses has | anon uucp, first time login,
- never appealed to us." | mail to fileserver@ruth.uucp
-