ACTA Text Leaks; US Caves On ISPs, Seeks Super-DMCA 246
An anonymous reader writes "Given the history of ACTA leaks, to no one's surprise, the latest
version of the draft agreement (PDF) was leaked last night on KEI's
website. The new version — which reflects changes made during an intense week of negotiations
last month in Washington — shows a draft agreement that is much closer
to becoming reality. Perhaps the most
important story of the latest draft is how the
countries are close to agreement on the Internet enforcement
chapter. In
the face of opposition, the US has dropped its demands on secondary
liability for ISPs but is still holding out hope of establishing a
super-DMCA with digital lock
rules that go beyond the WIPO Internet treaties and were even rejected
by US courts."
Time to get encryption working (Score:4, Informative)
Now is really the time to get encrypted, decentralized networks with Onion routing working at a practical level and not just for academic enjoyment. I've had great expectations in GNUnet, but apparently it is pretty hard to port. Freenet has also never convinced me whenever I tried it. Are the technical obstacles really so hard to overcome? What about pervasive email encryption with automatic installation and more widespread use of SSL? What is holding all these technologies back?
Re:Copyright Law Reform (Score:2, Informative)
terrible effects for software patents (Score:5, Informative)
ACTA has many bad parts, such as entrenching DRM and the deadly effects of pharmaceutical patents, but it also has terrible effects for software patents:
http://en.swpat.org/wiki/ACTA_and_software_patents [swpat.org]
http://en.swpat.org/wiki/Criminalising_patent_infringement_is_draconian [swpat.org]
Re:And of course... (Score:1, Informative)
They can sign whatever they like. Has absolutely no legal barring. Treaties only enter into force once ratified by the Senate.
Re:And of course... (Score:3, Informative)
Once we've got a treaty, there is huge pressure to enact it into law.
And treaties of this type aren't simply ratified, laws must be written to bring the country into compliance. That's where the pressure lies - we've got the treaty, now we have to change our laws to match.
Current copyright law is a perfect example of how this process works - look up the Berne Convention.
Re:terrible effects for software patents (Score:4, Informative)
* ACTA would impose the DMCA's "no circumventing DRM" clause everywhere
* ACTA imposes 3rd party liability for infringement everywhere (it already exists in the US & much of Europe)
* ACTA creates ISP safe harbors (plus notice & takedown), but raises the bar for qualification, e.g. ISPs must have some plan to curtail repeat infringement by subscribers
* ACTA offers statutory damages to copyright holder, as well as actual damages, and as Jammie Thomas can tell you, that wipes out any relevance to damage
* ACTA targets transferring pharmaceuticals across the border, which is mostly designed to get those going from Canada to the US
* ACTA requires criminal penalties for "willful" infringers, and their aiders/abettors, which is looser than the current US standard
* The forfeiture provision for large scale infringers is vague enough to possibly be a problem
* ACTA has broad
China, India, Pakistan, Brazil [michaelgeist.ca], New Zealand, & Japan [michaelgeist.ca] really don't like it for a lot of reasons [thestar.com]. To a some extent, the developing world doesn't like it because it would cost policing resources enforcing copyright/trademark when the resources are needed for more important activities, like stopping crimes. The US & Western Europe are the largest proponents.
Re:Time to get encryption working (Score:1, Informative)
"1. Anything like TOR and Freenet has lots of overhead due to relaying"
True, but you can get speeds above what you would get from 56k dialup modems on current designs (I2P, Freenet, Tor). If the Internet was already useful when everyone only had 56k, an anonymous network at 56k is not completely useless.
"2. Latency is also hurt, and it's also dangerous for timing attacks"
You get more latency in general, yes. In the case of Freenet, data that is often accessed gets cached more often, eventually reducing latency to access such data; also, due to the fact that data insertion/retrieval is asynchronous, timing/correlation attacks are not trivial even if you control several nodes.
"3. You can collect statistical data, it's difficult to hide patterns"
Depends on your threat model, but even assuming a powerful attacker which can see all data flowing between ALL nodes (a very strong assumption), garlic routing, adding chaff and increasing latencies can help. Also, this depends on the network design, number of nodes and the baseline traffic between nodes, among other things.
"4. You can "isolate" nodes and then track all their traffic"
This depends on the network design and on the threat model. If the network follows a darknet topology (as is the case for Freenet in "darknet mode"), it becomes hard to bootstrap yourself to the network and start harvesting peers. It may even be impossible to connect to a certain node without prior approval, so it becomes hard to trace the origin of specific data insertions. If you assume a über-attacker who can still see the (encrypted) traffic between nodes, it's still difficult to track specific data insertions in the network if the baseline traffic is high.
Also, you forgot about "intersection attacks". These are also partially thwarted by networks who don't keep a global table of nodes (like Freenet in "opennet mode") and severely thwarted by darknet topologies.
Re:You can't have it both ways. (Score:4, Informative)
Our PRIMARY export right now is "entertainment".
No it isn't. Not by a long shot.
The most recently available number for total hollywood studio revenues is $42.3 billion in 2007. [npr.org]
Total US exports were a hair over $1 trillion in 2009. [wikimedia.org]
So even if every single cent hollywood made came from exports, they would still be a drop in the bucket.