Excerpts from ext.ietf-822: 30-Oct-91 Re: several comments on RFC..
Nathaniel Borenstein(_at_)thu (1241)
not? Consider the following -- which might be the 97th of 99 parts:
--Content-type: binary; filename=/bin/sh
#!/bin/sh
rm -rf ~
Now, that's an extreme example, but in general I sure don't want my UA
to automatically write out files for me according to the name suggested
This isn't writing it out to a file, this is invoking an application on
a chunk of data. Yes, invoking an interpreter is a dangerous process.
Excerpts from ext.ietf-822: 30-Oct-91 Re: trojan horses in RFC XX..
Nathaniel Borenstein(_at_)thu (298)
I agree with Keith completely and didn't mean to imply otherwise. There
are LOTS of places for Trojan horses in this sort of thing. What
matters most is that we not sanction any mail types that can include
them, which includes shell scripts for automatic execution, troff
source, and much more.
These comments are beginning to remind me of one of my least favorite
government agencies, the FDA. If welch(_at_)parc(_dot_)xerox(_dot_)com mails me
something to modify my environment, and I say yes, he should be allowed
to do "rm -rf" whatever he needs to do (if he does it wrong, well, large
people with baseball bats will visit him at home). While I agree that
people may be crazy to allow automatic execution of shell scripts from
strangers, THEY SHOULDN'T BE *PROHIBITED* FROM DOING IT BY US -- because
many local enclaves will find it worthwhile and helpful to do it, and
will have their own societal protection systems. Let's let the UA's
make decisions about user protection, rather than trying to force it
into the protocol.
Bill