Hi!

Thanks for your reply!
> How big is that risk in reality?

I would say it is one of these situations where the chances of it happening are 
extremely low but the potential damage is very high -- both for users and the 
project. Kind of like all the emergency exits in an aircraft: the chances you 
will actually get to use them are low but they're here because it cannot hurt 
to know they might come in handy.

> Do you check the identity of your courier when he/she delivers a
> parcel to you? The answer is probably no for 99% of people. Should we
> be doing this?

As a matter of fact, I sometimes do. It all depends on which package is at hand 
and, in the case of file downloads, I would think downloading an executable 
(Camino) requires more precautions than downloading a simple image or a PDF 
file -- which can also raise issues but, as you so rightly point, there is a 
limit where one should trust the other party.
 
> In all honesty, even if I checked the courier's credentials, I have no
> way of knowing they are authentic. In the end it t depends on how
> paranoid we are. I usually grant trust to people every day, taking
> people (or businesses ) at face value is a reflection of how we view
> our world.

That is true.
 
> All a check sum would do is confirm that the correct one for that
> file, not its authenticity. Its very self-referential. Anyone capable
> of breaching the security of a server like these is just as capable of
> putting the correct check sum for the malicious file he just slipped
> onto the Camino servers.

Maybe -- although one of the mirror servers could, due to some administrative 
issue, become temporarily vulnerable in a way the Mozilla servers would never 
be, or the reverse. Also, it takes more time and energy to hack two servers 
than one, even for the most experienced of hackers.
 
> What is true is that I have faith in the people who have been working
> on this project   ... this extends to the choices they make about the
> security of their servers. It is part and parcel of the work of a a
> development team. There is a kind of chain of trust here which enables
> us to act in a somewhat less paranoid mode than some suggest.

It's not a matter of trusting the people working for the project here or their 
security choices. On the contrary: I trust them and this is why I use their 
product -- like million other people. It's more a question of ensuring all the 
chances are on their side should something bad and beyond their control happen.
 
> I've chosen trust in place of paranoia.

I'll leave it to a quote I especially like: "The difference between 
common-sense and paranoia is that common-sense is thinking everyone is out to 
get you.  That's normal -- they are.  Paranoia is thinking that they're 
conspiring." 

This being said, thank you for your input and sharing your opinion. I really 
appreciate it and fully respect it.

FJ
_______________________________________________
Camino mailing list
[email protected]
http://mozdev.org/mailman/listinfo/camino

Reply via email to