Daniel,
by way of example, my PC has a firewall (Symantec) that does outbound
filtering. I appreciate the fact that when I launch an application for
which I have not previously provided authorization to access the
Internet (defined here as an IP range beyond my LAN subnet), the
firewall warns me before allowing the connection to take place and lets
me decide whether to block, allow this one time, or allow permanently
the access. With this kind of protection on devices such as the N800,
it is more likely that the outbound filter will also catch a silent
rogue app that, by some means, has gotten installed on the device.(these
days typically by a user being socially engineered to do something that
they should not do).
One of the challenges of any software developer group like this is to
perceive the product like the average end user and not a developer. With
respect to device security services, as opposed to, say a collaboration,
communications, or entertainment app, it is hard for an experienced
clueful developer to put him or herself into the "shoes" of the average,
not highly clueful, end user. Security is, whether we like it or not, an
essential part of the software/device/product usability mix even though
it is, for most developers, a boring aspect of computer systems/science.
Today's responsible sw/product companies and their software engineers
work hard to establish the right balance of security and useability.
Microsoft has been a major miscreant in this respect for many years, and
they have, painfully, learned their lesson and are now working hard to
correct past mistakes. Nokia clearly has the opportunity to do better
with its products of the class of the 770/N800 and set a standard for
others to follow.
Best Regards,
John Holmblad
Daniel Stone wrote:
On Tue, Feb 20, 2007 at 04:34:21PM -0600, ext Paul Klapperich wrote:
Nokia really doesn't have to do anything to "guarantee" that 3rd party apps
are safe, but I would certainly trust the integrity an official iptables
compiled by Nokia. They certainly have something to loose by somehow
subverting it, so I would trust it. And as it really wouldn't take anything
more than a checking the option in the kernel config before building, I
really don't think this is any additional burden to them.
Okay, so what are you going to do with this iptables package? As far as
I can tell, there are two options:
a) nothing;
b) iptables -P INPUT DROP; iptables -A INPUT -m state --state
ESTABLISHED,RELATED -j ACCEPT
a) provides no change over the status quo.
b) provides no change over the status quo, except that it encourages
people to have open ports. It would mean that people who wanted to
listen to the outside world have to explicitly punch a hole in the
firewall. Right now, people who want to listen to the outside world
have to explicitly open a socket on that interface, which to me is a
pretty clear statement of intentions.
I just don't see why you would want it. How would it make things more
secure, at all, except for the warm fuzzies that come from having a
firewall? It provides no practical benefit in any case which isn't
hopelessly contrived.
Cheers,
Daniel (not responsible for this decision, not speaking for N, etc etc)
------------------------------------------------------------------------
_______________________________________________
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers
_______________________________________________
maemo-developers mailing list
maemo-developers@maemo.org
https://maemo.org/mailman/listinfo/maemo-developers
