[olug] MPAA toolkit

Will Langford unfies at gmail.com
Wed Nov 28 00:59:52 UTC 2007


> As it stands right now most independant ISPs would not be able to say in
> court whether traffic for IP x.x.x.x belongs to user Y on date Z without
> a shadow of a doubt unless they were actively monitoring that IP address
> at the time. I know I could not and would not attest to it in a court of
> law. And I know first hand many ISPs that would be in the same boat.
> Heck I know of some that to this day could not be CALEA compliant even
> in spirit if a subpoena was issued.

As far as I know, ISP's do keep track of when a person was
connected... at least dial up ISP's would have been that way.  This
way account usage could be assessed and billed.  I had one small ISP
back in FL even go as far as to isolate me as a particular IP at a
particular time back when linux was my acting gateway and had some
'public servers' running on it (although I had no out-of-lan network
activity on those services... apparently they just port scanned me
since I had a high number of hours online).  Dial up ISP's logging
associated IP addresses is... in the realm of I'm unsure of heh.

Broadband ISP's... I doubt would do such a thing, other than perhaps
some kind of DHCP lease history perhaps.

> Asking any large ISP to archive netflow data (or even worse all packets
> which was originally proposed by the FBI) would be cost prohibitive, but
> to take it a step farther, that data would be useless unless the ISP
> could also guarantee the accuracy of that data in the archive it would
> be useless (wouldn't hold up in a court of law).

I didn't contest that network traffic monitoring would be... insane
:).  Accuracy of data would be interesting to attempt to prove,
however :).  I suspect that some random connections by a third party
could be used as a control for general 'it works' proof... dunno how
reliable such a thing would be in a gritty court environment tho.

> And even if this law was to pass it would be pointless in the P2P realm
> because so many sites are going to encrypted data to circumvent packet
> inspection by ISPs who are trying to shape the traffic.

I imagine this would be relatively simple (with a simple
circumvention).  If data isn't apart of a handful of recognized ports,
it gets throttled. Good luck on getting game/ip-telephony to work on
non standard ports... or if the service isn't recognized by your ISP
:).  Also, I guess, then, that alot of the leechification stuff would
move to those recognized ports :) (hence easy circumvention).

> Asking Universities to stem illegal activities on their networks is a
> good thing, holding all student aid and school funding hostage to the
> requests of a private business is something quite different. If a
> business is colluding in the wrong doing then it is liable, but if for
> instance if I was to use a work phone to call my bookie and place an
> illegal bet during March Madness I doubt you are going to find a court
> that will find my employer guilty of anything.

Regarding beginning of paragraph, I believe I said similar....

Latter half and using work phone to place a bet... is more
interesting.  Along similar lines, if an employee has pirated software
on their work system... does the business get any blame ?  As far as I
know, it does... even if it is unaware of the employee installing it.
Similarly... I wonder if an employee initiates a DoS attack from a
work based net connection... does the business have any liability ?
How does this kind of stuff apply to businesses that have and
advertise free wifi access (coffee houses, resturuants, etc ?  I think
with publicly accessible wifi stuff you probably have to agree to a
disclaimer when making the wifi connection ?  I've not actually used
one other than a hotel in NJ which required agreeing to a
disclaimer/waiver.

> for damages, I do however have a problem with the RIAA and the MPAA
> using the federal government as their enforcement arm via legislative

As such... you could easily say that alt.binaries.* is widely used for
all kinds of infringement on many things.  But, there's very little
effort done by ISP's to filter said groups.  Hm.  If I mail order my
infringement materials, the post office doesn't have any filters in
place... and it don't track regular postage stuff.

I dunno.  I know that businesses have some absolute liability in some
cases, and it'd be a good idea to at least have similar rules apply to
uni's...

Data logging stuff on any kind of non-dial-up network is simply
insane.  Packet sniffing / shaping is also not an appropriate
solution.  I imagine this type of discussion (ISP filtering content)
happened back in the 33.6 days, but... the insanity of it on modern
connections is a bit overwhelming.  Perhaps we need the TCP evil bit
defined.

-Will



More information about the OLUG mailing list