🚀 go-pugleaf

In article <bontchev.727978767@fbihh> bontchev@fbihh.informatik.uni-hamburg.de writes:
>
>bediger@teal.csn.org (Bruce Ediger) writes:
>
>> >What actually keeps viruses from spreading on such platforms is the
>> >different level of software sharing between the users.
>
>> Are there papers/books/other references that purport to demonstrate this?
>> I would be interested in reading such, if they exist.  Have any other 
>> virus-spreading experiments been publlished, other than Fred Cohen's 1984
>> work, and Tom Duff's "Computer Systems" article?  The Duff article was more-or
>
>If you have read Cohen's works, there it is demonstrated well enough
>that the only way to make viruses impossible for a Turing machine is
>to do either one of the following: limit sharing, limit transitivity,
>or limit functionality. Cohen also describes it in a more "readable"
>manner in his "Short Course on Computer Viruses".


This is not terrifically responsive.  The original assertion was
>What actually keeps viruses from spreading on such platforms is the
>different level of software sharing between the users.

(Sorry to not be able to retain original attribution.  It's not my
 statement, however.)


Yes, Cohen's papers describe what is _theoretically_ possible and
impossible about the spread of viruses.  The question being debated
is "Why doesn't a UNIX virus exist in the wild?"  A whole bunch of
straw men have been proposed and subsequently knocked over: file permissions,
address space separation between processes, more knowledgable users,
basic OS design that includes some small amount of security, lack of user
patchable interrupts, and greater morality/ethicality of UNIX programmers.
It's _theoretically_ possible to write a UNIX virus based on the info
contained in his papers, or info from a multitude of other books/papers. 

Sinced it is widely asserted/believed that publishing PC virus source code
is really conducive to more PC viruses (Ralf Burger's Vienna Virus source
is said to have led to major Vienna Virus infestations and variations,
the "Little Black Book Of Computer Viruses" was widely denigrated for
the same reasons), and everything short of compilable source code for
viruses has been published, there must be some reason why UNIX viruses
are not more widespread.

So, I ask again: are there papers/books/other references to studies of
the spread of viruses?  I would be interested in them because of the
dissonance explained above.  It also seems to me that Professional
Anti-Virus folks would be interested in such, since it seems that
vectors of infection would be a logical place to start preventing spread.

Yours,
Bruce Ediger