> As I stated yesterday, we already KNOW how akin to swiss cheese Windows 
> is.  How could any OS provide less security?

I don't think any of us wants to promote ANY approach to security with the
rallying cry of "better than Microsoft."  Fortunately, we don't need to.

I think there is a general misapprehension that simply being able to see
source code for a system is equivalent to being able to subvert a system
that is running that code.  This is not the same thing.  

Even widely-used security techniques such as PGP have long made the
underlying code publicly available for review (see, for example,
http://www.pgpi.org/cgi/download.cgi?filename=pgp50i-unix-src.tar.gz) so
that everyone may assure himself of the effectiveness of the safeguards and
the rigor of the techniques.  Knowing how PGP encryption works does not
permit anyone to "break" the encryption.  Likewise, knowing how an
open-source system prevents intrusion does not mean that this knowledge
enables anyone to circumvent that system's security.  

Moreover, an open-source system has the advantage that it does not require a
blind reliance on any single vendor's assertions about its security;
everyone may examine it and confirm for themselves that the code effectively
precludes unauthorized access.  Anyone's concerns with the code are quickly
and broadly communicated to the development community, and many hands work
on closing any cracks.  

However, because proposed changes to any aspect of any of the major "open"
systems undergo an extensive peer review before their inclusion in the
generally accepted system, any weaknesses are usually identified BEFORE the
system goes into general release.  In a closed-system environment, fewer
eyes review code before it is released, and those who do review it have an
allegiance to the owner of the code and an incentive to meet market
deadlines.

So there is no reason for open-source systems to be inherently insecure, and
in fact, the number of people scrutinizing the code contributes to its
stability and security and serves as a de facto certification of its
effectiveness.

        - Ed

- - - - - - - - - - -
Ed Ward [EMAIL PROTECTED]


_______________________________________________
DIGITALDIVIDE mailing list
DIGITALDIVIDE@mailman.edc.org
http://mailman.edc.org/mailman/listinfo/digitaldivide
To unsubscribe, send a message to [EMAIL PROTECTED] with the word UNSUBSCRIBE 
in the body of the message.

Reply via email to