Internet Week in Washington
Congress is back in session with a bang, and the Internet is smack dab in the crosshairs of the Congress in a big way. Just this week, there have been hearings on rogue sites legislation in the Senate Judiciary Committee (Video,) a hearing on net neutrality and competition in the House Judiciary’s Subcommittee on Intellectual Property, Competition and the Internet, and a full grilling of the FCC commissioners in the House Energy and Commerce Subcommittee on Communications and Technology (Video.)
I took part in a staff briefing last week to help bring new staffers up to speed on the Internet, as did many others I’m sure. The level of the dialog in the Senate and House E & C committees was pretty good, and while I haven’t heard the Judiciary hearing yet, the witness list suggests it was good as well; Larry Downes was on it, and he’s one the best-informed all-around scholars in the country on this subject.
The technical subjects around the rogue sites issue pertain to whether the proposed measure will “break the Internet” as opponents charge, and to a lesser extent, whether such actions as domain seizures will be at all effective. “Breaking the Internet” sounds like a dire prediction until you realize that it’s a charge that has been made rather routinely against every significant change in received Internet practice, from TCP/IP itself on down to the anti-spam blacklists, and it generally goes hand-in-hand with the charge that people will simply circumvent the measure. Astute readers will note that these two charges cancel each other out: If we assume that a given technical measure will be widely circumvented, it’s not particularly consistent to claim that it’s going to “break the Internet.” How does that happen if everyone’s avoiding the measure?
The “breaking the Internet” claim depend on the assumption that people who want to download pirated content – which is apparently everyone, according to the critics – will switch to some as yet non-existent rogue DNS that provides addresses for blacklisted web stores that deal in unlawful content. It’s not really necessary to go that far, of course: Editing the right local file will do to supplement the standard DNS, but people who’ve been tricked into purchasing unlawful content aren’t going to do that, much less go to a rogue DNS that redirects them to bogus sites instead of the ones they want. What may not be apparent to the Congress – although the truth is certainly starting to dawn on them – is that the so-called “Internet engineers” who signed a letter written for them by the attorneys at the EFF are not the most articulate people in the world, so when they say “this is going to break the Internet” they actually mean something like “I don’t like this because it feels strange.” Any number of IETF standards have been criticized for their potential to “break end-to-end” as well; engineers aren’t known for their facility with the English language.
The Senate committee seemed to have an epiphany during the hearing on the question of blocking access to overseas sites that don’t rely on a US registrar who willingly complies with the ICE enforcement measures today, namely that search engine regulations can do the trick. If Google and Bing are willing to delete blacklisted domains from their search results, the rogue sites problem is greatly reduced. This is one case where the Internet’s tendency toward consolidation can be leveraged to address a problem. Witnesses seemed to suggest that the DNS blacklist is not an essential enforcement tool provided search blacklists can be adopted in its place, or in addition to it. If I could only have one tool, it would be stopping the credit card transactions, but multiple tools are better.
The House Communications and Technology subcommittee hearing was a marathon affair that made me happy to have three monitors on my desktop so I could let it play while working on some other things, pausing for the dramatic parts. Chairman Walden is extremely well-briefed on the subject, and was all over every aspect of the issue that matters. He has the distinction of being a former FCC license holder who’s very familiar with the powers of persuasion the Commission holds over similar firms. Walden was especially effective when he brought out the language of the Comcast/NBCU merger conditions that require the merged firm to abide by the Open Internet rules even if they’re struck down in court. That’s not a condition that suggests great confidence in the agency’s authority.
The key issue on the technical side of the debate, although it’s minor compared the question of FCC authority, is whether the Internet demands an “all packets are equal” rule that bans the practice that Free Press and friends labels “paid prioritization.” The term is misleading because prioritization is one means of providing Quality of Service for premium or very low cost transit services, but not the only one. So the terminology itself displays a “fallacy of composition” on the parts of the neutralists (confusing the part with the whole.) Chairman Genachowski said that the Internet has never used QoS. which is sorta true and sorta not, depending on which part you examine and during what period of time. QoS is certainly part of the architecture of the Internet and the subject of a number of RFCs. Genachowski’s response indicates that he’s not nearly as well briefed as Walden is on the subject of Internet norms and standards, which is unfortunate. The FCC is supposed to be the technical expert, after all, while the Congress is supposed to be the policy arbiter. The apparent issue is that these two bodies need to get their roles realigned in terms of their actual statutory relationship. The best way to do that is for the Congress to update the Communications Act, which is going to take a while.
One thing that is clear is that the passage of the FCC’s Open Internet order didn’t clear the air regarding net neutrality; the plague of the greatest public policy boondoggle of the 21st century is still very much with us.
Too much policing will eventually harm the millions of people who look upon internet for information at the click of a mouse.
To some extent, the spam-wars DID break the Internet, including collateral damage as spammers tried to evade them – which is completely logically consistent. Now, it wasn’t a world-shattering breakage, but it did happen, and it did hurt people. I actually thought it was going to get much worse than it did. There’s an interesting case-study to be done, roughly on how and why the border skirmishes did not evolve into an all-out World War. But the people raising that concern were quite right – and some of the answer was (gasp) g-g-g-g-government action. I also recall the Libertarian-type ideologues hated that government laws played a part, until their funders made clear there was no money in ranting about it.
So life’s more complicated than you would have it 1/2 :-).
Care to elaborate on how the spam wars nearly broke the Internet, Seth? I didn’t see anything like that happen myself.
Don’t take a view of If-I-didn’t-see-it,-it-didn’t-happen. By analogy, there’s plenty of violence along the US/Mexico border that you haven’t seen either, but it’s quite real for the people affected. Anyway, are you familiar with Paul Vixie’s efforts then, “MAPS”, and what happened with connectivity (not just email!) being broken at times to *websites* it blacklisted? There were lawsuits over the project.
http://www.digitaltransactions.info/classes/computer-crimes/media3-technologies-llc-v.-mail-abuse-prevention-system-llc-2001-wl-92389-d.mass.-2001
Is your argument so weak that you can only press forward by putting words in my mouth, Seth? I only said I didn’t see the Internet stop working as a result of the RBLs, not that it didn’t stop working.
I’m asking for evidence that I’m wrong, and you’ve provided me with a lawsuit filed by a spammer against MAPS. Not only is a lawsuit insufficient evidence of the Internet failing, the one you cite was quickly settled in favor of the defendant: http://files.grimmelmann.net/cases/Media3.pdf
My quest for evidence is no further along than it was when I posted the entry, but now I’ve lost five minutes of my life by checking Google for the outcome of a frivolous lawsuit.
I was attempting to pre-emptively deal with an anticipated objection about scope. I have provided you evidence of Internet breakage due to spam blacklists. That’s what the lawsuit is about. The lawsuit was not filed by a spammer, _per se_, but by an ISP – which was involved with spammers, and hence the blacklisting and lawsuit against it. The point of the lawsuit as evidence is not that the ISP-plaintiff was correct, but rather that these events did happen (i.e not some lobbying press release), and reached the level of litigation (where, nontrivially, perjury penalties apply).
The important part to check there was not the outcome, but the events, e.g.:
“If Media3 did not comply, MAPS informed Media3 that it would place on the blackhole list not only the ten contested websites but also any other websites that were on the same Class C network as the contested websites.
This prospect was of some concern to Media3 because, as a hosting company, one of the primary services that it provides to its customers is ensuring that their websites are freely accessible and can easily access the Internet. Inclusion on MAPS’s blackhole list would threaten Media3’s ability to deliver good access to the Internet. After some exchange back and forth via e-mail and telephone between MAPS, in California, and Media3, in Massachusetts, Media3 refused to comply with MAPS’s requests. MAPS then listed the disputed websites and any other websites on the same Class C network on the blackhole list.”
Yes, this sort of thing still happens and has happened to me personally. There are more than 100 anti-spam blacklists these days, and each one has different criteria for inclusion and removal; some charge a fee for “expedited removal” from the list. This is troublesome for Apache virtual domains like bennett.com, which shares an IP address (not just a network prefix) with a number of other domains. If one of the domains is blacklisted, then bennett.com is blacklisted and that’s a pain. I don’t sue anybody when it happens, I notify the ops and switch to a different mail server until it’s straightened out.
The point is that even when blacklisting goes wrong, it doesn’t “break the Internet” or put anybody out of business. Spam is the root cause of these issues, not the measures that aim to deal with it in the limited way that’s possible short of authenticated e-mail.
[…] This post was mentioned on Twitter by Steve Crowley, High Tech Forum. High Tech Forum said: Richard Bennett talks about hearings, briefings during "Internet Week" in Washington – on High Tech Forum http://shar.es/361Az […]
This is a funny line – “engineers aren’t known for their facility with the English language.”
I’m not an engineer, but isn’t the “best” way to “break the Internet” via BGP hijack?
Attended a federal security briefing last month, wrote it up for CircleID here: http://www.circleid.com/posts/20110124_the_dark_internet/
Interesting account, Chris, and you raise an interesting question as to which of the several ways to break the Internet are best. In Egypt, switching off the power to the central Ramses Internet Exchange was the preferred method.
Richard, when you say “Yes, this sort of thing still happens and has happened to me personally”, I think you’re again not reading what the lawsuit actually concerned, since you then say you “switch to a different mail server until it’s straightened out.”. The Media3 vs MAPS lawsuit did not involve just mail servers, but all Internet access. Hence, it’s true it did “break the Internet” (again, properly qualified to some extent). Whether it put anybody out of business is not really the claim. Nor does where to assign moral culpability (“Spam is the root cause of these issues”) change the fact that these events occurred.
I would submit that the spam came closer to breaking the Internet than the anti-spam blacklists did, Seth, properly qualified to some extent. The last time I checked, the Internet was working about as well as it ever has.