So you think it's futile to try to get software vendors to improve their
products. I suppose I can go along with that to a certain degree. But how
can you expect end-users to work around the brokenness in the software they
use? This seems both unfair and futile.
at my aforementioned sister's house, i did it by buying an off-the-shelf
$99 firewall and a $79 copy of suse-9 and spending an afternoon showing her
how to use them. i guess the general form of the answer is "tell people to
get some tech support rather than believing what their vendors say." i'm
not an expert on d-link firewalls, or on linux, but i know enough to know
that running MSIE and Outlook and not having a firewall was her problem.
Einstein taught as that even the simple act of observation influences our
surroundings. Wouldn't it make sense to try to leverage this influence such
that the future is shaped more to our liking, however small the change may
as sad as this is, the best way to accomplish that is by heaping public
scorn and ridicule on sean's and chris's employers every time they whine
about how folks are widely blackholing their customers. you won't
convince sbc or mci, but you might convince a lurker or two.
>> But the real issue is that this is even necessary. The biggest problem
>> we have with IP is that it doesn't provide for a way for a receiver to
>> avoid having to receiving unwanted packets. It would be extremely
>> useful if we could fix that.
> you realize that the virtual circuit X.25/TP4 people are laughing their
> asses off as they read those words, don't you?
It's easy to laugh if you don't have a world wide network to run.
i once had the honour of taking over a network dave rand had built, which
became an unprofitable dot-bomb on my watch. ouch! but it wasn't because
we refused to take money from spammers, or because we disconnected folks
pre-emptively when they violated their AUP. so, that's not what i meant.
if you want to put enough intelligence into the network so that "a receiver"
can "avoid having to receive unwanted packets" then you'll need to decide
how to throttle flow solicitations or else those flow solicitations will
become the new form of spam and ddos. this will require state, not just in
your hosts and upstream router and provider, but globally, end to end. and
if you do that you'll have bitten into the rotten apple of circuit switching
and x.25 and atm that the IP folks have been saying all these years wouldn't
scale and wasn't necessary. and so, the people on the other side (the losing
side, in my opinion) of that argument will laugh their asses off, whether
they have a world wide network to run, or not.