[Chapter-delegates] PIR and wikileaks.ORG

[Fred Baker]

On Dec 3, 2010, at 2:39 PM, Anya Chambers wrote:

> A ton of questions, that is in the realm of ISOC policy program. What's ISOC point of view while trying to stay away from the fight?

First: thanks for asking the ISOC community the question. 

My view:

Had a government organization served PIR or the registrar you mention with an everything-legal court order to take down a name, both of them being incorporated in the country that government runs, I think that they would be in violation of the local law had they not done so. That doesn't mean that there should not have been subsequent legal discussion; as I understand the relevant law, there are parties in the US that would *like* the ability to take down domain names under a bill (proposed law) called COICA, but there is not currently a law that authorizes such take-downs without due process.

You indicate that what in fact happened was that a stated registrar took down the name as a defensive move in response to a ddos attack. I think such an action is understandable, but lamentable. They at least owe the domain name holder an explanation of what happened.


On the larger topic of forcing content off the net, I think the ISOC community should understand that almost every country has some definition of what content is acceptable, and would like unacceptable content to be prevented from being delivered - if not everywhere, then within their borders. Muslim countries generally don't appreciate content that speaks ill of the prophet. China likes to control political information of several categories. More repressive regimes, such as Myanmar, have cut all links to the Internet to control content. In Europe, child pornography is unappreciated, and in the US, pornography as a class of material, and if appearances are to be believed, some forms of political material are unappreciated. The UN Al Qua'da Task Force would like all publications from Al Qua'da, but especially recruiting material exchanged in IRC chat and other media, blocked. And Oh By The Way, there is the never-ending discussion of the distribution of copyrighted material, which is the subject of COICA. It's not "those evil guys over there" that want to block content, it's everywhere.

There are a number of us that have been approached for "a solution that works". My comment to a colleague that was trying to figure out how to get out in front of the idiots:

On Dec 1, 2010, at 2:51 PM, Fred Baker wrote:
>> I do not know what the best path forward is. CAN "we" take back the initiative and propose something that actually works?
> 
> It starts out with defining "works". If "works" means "prevents <content> from being distributed", I don't think so, at least not in a manner consistent with EU and US concepts of freedom of expression. That was my comment to the UN Al Qua'da Task Force, which wanted service providers to prevent the distribution of Al Qua'da material by detecting in real time that a file being moved was Al Qua'da content and blocking it. It comes down to being able to determine that a stated bit of content is or is not OK to distribute and apply that rule. 
> 
> Comment from the WIkipedia on HADOPI:
> "On 10 June 2009, the Constitutional Council of France struck down the central, controversial, portion of HADOPI, that would have allowed sanctions against internet users accused of copyright violations (as opposed to being convicted for same), ruling that because "the Internet is a component of the freedom of expression" and "in French law the presumption of innocence prevails", only a judge can impose sanctions under the law.[9][10]"
> 
> The problem with DNS blocking is
>  - first that it is trivially defeated by getting an alternate name in a different *TLD,
>  - second that the fact that someone has accused the owner of a DNS name of distributing
>    one or more files (www.torrentfinder.com? What that they don't own do they distribute?
>    They provide pointers to such content, but they don't distribute the content), and
>  - third that if it ca be shown that there was one such file, that doesn't make the rest
>    of their stuff illegal to distribute.
> 
> That is illustrative of the issue. Cleanfeed uses as a rule the correlation between a URL in a GET with the set of URLs in a list. That skips some of the other issues, but it implies that someone knows that the file has a stated URL, and btw it is trivially defeated using https. So it doesn't detect/delete "bad stuff" reliably either.
> 
> Oh, you're from UN AQ TF and want me to watch the bits on the wire and realize that they're bad? Trivially defeated by zipping the file, perhaps in a directory or with a second file in the same zip file to further randomize things, even if we ignore the other REALLY OBVIOUS issues like "and how would you like me to do that?".
> 
> And then there are other issues. If kiddieporn is illegal in France, distribute from Lower SLobbovia where it is perfectly legal. If provider foo blocks you, there are clearly providers that won't.
> 
> No, I don't think there is a magic wand we can wave and make the distribution of content go away. We can, in various ways, make it harder to distribute. But the role of the courts doesn't go away, and distribution can't be 100% prevented.

It seems to me that there are at least two issues here, and maybe more. One is legal - the reason we have a thousand laws against obscenity in the US and still have large volumes of it is first that it is very hard to objectively define. If the issue was "the color red", we could objectively assess whether a content source was doing something illegal. But that doesn't work for most content folks would like to block. The other is technical - OK, so tell me how to block it that isn't trivially defeated?