|
Forbes just published this article that’s being shared all over my facebook friends feed.
I left a comment on the article that I’ve copied and pasted here, as it is just about long enough to qualify as a CircleID post by itself.
* * *
The problem is that peering isn’t always settlement free—and even if it is, if and only if there’s an equitable amount of traffic exchanged between two ISPs. And then there’s transit, where you pay another network to carry your packets for you.
Peering and transit economics are rather more complex than settlement free handshake deals, and at least these days, they’re more often than not based on contracts with every i dotted and t crossed, rather than handshake deals (though those too aren’t unknown).
This generally means that networks that have much more eyeballs than content will inevitably spend a lot of money, on international bandwidth capacity over undersea cables, transit etc, just so that their customers can reach all the content provider sites (google, fb etc), which are hosted elsewhere. This is mitigated to some extent by the use of CDNs like Akamai, local mirrors of popular sites etc, but isn’t exactly cheap.
The content providers of course are entirely against any such measure as it’d dent their profits if they were to actually pay a lot more for connectivity than they already do—and so spend quite a lot of time and money to lobby internationally, as well as hire respected names in the internet industry who, [possibly in several cases convinced that these measures are an attack on the essential freedoms of the Internet] blog, tweet, give interviews and influence articles about how pernicious these measures are, and how they’ll lead to the ultimate destruction of the Internet as we know it.
Meanwhile, the largest content providers also buy (and in some cases, even build) substantial internet bandwidth capacity between different countries and set up datacenters around the world—effectively becoming their own ISP to cut costs from what it’d cost them to buy large quantities of connectivity at retail prices.
An excellent (though rather old) paper on this was written by Geoff Huston, formerly from Telstra and now at APNIC, published in the respected Internet Protocol Journal.
Another elephant in the room is that the internet’s coordinating bodies are, at least on paper, under the control of a US government agency—the department of commerce.
While there has been a largely hands off approach on behalf of the USG so far, more than one country that doesn’t torture and censor its citizens is not quite comfortable with relying heavily on and treating as a critical resource something that is actually under firm control by the USG, should they choose to exercise it.
There is of course the traditionally utopian idea of a free and wide open internet “frontier” (as in space, the final frontier, as in that fine old time when men were men and shot their own meat, sewed their own clothes on the american frontier, as in the Electronic Frontier Foundation and its declaration of independence of cyberspace), with no rules, regulations or anything other than the “Code Is Law” concept that Lawrence Lessig once proposed… an article of faith among true believers in internet freedom.
This faith unfortunately tends to collide with reality on several occasions.
There is the extremely broad crossover between international civil and criminal law and the Internet. Someone in russia or brazil is just as capable of emptying out your bank account with a banking trojan as is a hood with a shotgun sticking up your local branch.
The virus writer (quite often with organized crime backing these days instead of just being a pimply kid hunched over an old PC in his parents’ basement) is far far less likely to get caught than the hood is, and electronic money transfers are far less easy to trace and claw back than actual banknotes are.
That’s just one case—where the concept of dual criminality (illegal in both the criminal’s and the victim’s jurisdictions) makes it relatively easy to prosecute [if there happens to be a MLAT and an extradiation treaty in place, without which prosecutors have been known to lure criminals out of their country with a phone call that says the criminal won a paid holiday in a casino resort]
And in civil law—peering and transit are, as I said, business agreements between companies, and there is a huge mass of litigation in front of various courts and regulators around the world. The same thing with say libel and defamation, tort law.
There have been several other frameworks before—such as the Council of Europe’s Budapest convention on cybercrime, which the USA and several other countries are already signatories to (though not, I think, Russia or China, or any of the Arab states). You would also find that a lot of cybercrime happens because of vacuums in law enforcement—criminals can operate best where they fear arrest and jail and/or extradition at the least.
Back to the current governance (or shall we say coordination?) model, there’s always ICANN, not particularly a poster child for good governance. Its governance has generally been driven primarily by commercial interests, with sort of nominal participation from its Government Advisory Committee (GAC) and its Non Commercial Users Constituency (NCUC)—and there has been enough criticism, within ICANN itself, and in the media (including in previous Forbes articles).
Proposals floated in ICANN, just like proposals in the WCIT or before other ITU groups in the past, or like weird and wonderful bills tabled in Congress, have been open to quite a lot of criticism in the past. Some get through and some never do see the light of day after they are tabled.
So it is unfortunately not just about rapacious telcos, repressive regimes and ITU apologists. There’s a much more complex backstory that has been completely glossed over (though to be fair it isn’t quite easy to compress over a decade’s worth of politics into a three or four page article).
In fact, that NY Times article you linked to as “by an ITU apologist”, is hardly what I’d describe as an apologia. For example, in the several years that I’ve known Milton Mueller from Syracuse University, quoted in the article, he has been a longstanding critic of ICANN, and of the ITU—and a longtime member of ICANN’s NCUC—and his views don’t tend to make him the favorite of several participants in ICANN—nor is he all that popular with several stakeholders in the ITU process.
BTW full disclosures in case they’re necessary:
1) I’m speaking entirely for myself and not for my employer.
2) I am good friends and in some cases, peers / colleagues with several people on the “other side”—registrars, content providers, civil society activists
3) I’ve written a paper on botnet mitigation for the ITU in 2007 and spoken at some of their workshops on cybercrime, the last time in I think early 2009.
Sponsored byCSC
Sponsored byIPv4.Global
Sponsored byVerisign
Sponsored byDNIB.com
Sponsored byVerisign
Sponsored byWhoisXML API
Sponsored byRadix
Yours is one of the better efforts to write informatively on this. Thanks for taking the time and effort.
... based on his replies to my comments on the forbes article. http://www.forbes.com/sites/larrydownes/2012/08/09/why-the-un-is-trying-to-take-over-the-internet/ Some operational input into the comments might be a good idea. He does cite ISOC etc studies about a lot of peering agreements still being settled over a beer and a handshake (which seems kind of strange given the very detailed requirements the big players - even the largely open peering types - have)