Coverage (in German) of the event can be found at computerworld.ch and the Neue Zürcher Zeitung, and if you’ve got 55 minutes to kill, video of the event itself (also in German) is available at the website of the Parliamentary Group on Digital Sustainability.

Encouraging is that there seems to be agreement on what I’d consider the most important points: network neutrality, in the sense of fair and equal access for all to the Internet, is crucial to its survival, and regulation thereof at the level of technical detail is likely to be ineffective and is indeed dangerous. There seemed to be no desire on the point of the industry to split regulatory domains into wired and wireless (which, like the 1996 US telecommunications act which established last-mile competition, but only over copper, could have perverse effects). On the point of those existing violations of neutrality in Switzerland (largely blocking or limiting of competitive applications), Carsten Schloter remarked, “Whoever is [doing that], has already lost,” an encouraging sign for fair competition in the industry.

As to how to protect this, there is of course less agreement, with Mr. Grütter and Mr. Schloter arguing for self-regulation, and Mr. Glättli for legislation. I can see the former point; given the relative speeds of innovation in telecom and in legislation, the risk of getting legislation wrong and ending up either with an irrelevant regulation or, worse, a damaging one, is nonzero. On the other hand, if one takes self-regulation as the status quo, other operators in Switzerland apparently don’t agree with Mr. Schloter’s opinion, and would seem to have shown that it’s not working.

Therefore, I personally tend to come down on the side of legal protection for network neutrality, but defined in such a way as to constrain anticompetitive behavior in the market, to protect access to the network for smaller players in the industry as well as for consumers, and to explicitly maximize the freedom of innovation (both technical and business) within those constraints.

One first step in this direction could be a legal obligation to transparency, for which there seems to be limited agreement. In this case operators offering unequal terms or selectively blocking traffic would be required to advertise this fact; the theory is that this provides an incentive in the market for them not to do so. This could work in a small market like Switzerland with last-mile competition both in the fixed and mobile space, but does nothing for mostly-monopolized markets like the United States, and does nothing to prevent future consolidation, either.

In any case, the conversation continues, and I look forward to taking part.

More seriously, the Internet has largely replaced the public switched telephone network and the postal system as the basic communications infrastructure of our society; just as with these systems, the “last mile” is a natural monopoly, so guaranteeing equal access to it is important. However, the results that legislation of network neutrality will lead to may vary widely based on how, precisely, it is defined.

At its most basic, “network neutrality” means the network infrastructure treats all traffic equally, and is essentially an aspect of the end-to-end principle, which moves more complex and application-specific functionality to the network edge, and is one of the keys to the success of the Internet architecture both as a communications network and a platform for innovation.

Here’s where it gets complicated, though: what do we mean by “traffic”, and what do we mean by “equal”?

At the lowest level², traffic consists of packets. One definition of network neutrality would focus on this lowest level, and demand equal treatment for all packets. Aside from small absurdities — of course the network has to treat packets differently by sending each packet to its appropriate (i.e. different) destination, otherwise the Internet would be analogous to a post office that was good for nothing but letters to Santa, and that would clearly be ridiculous — this seems to make sense until you get into the fine print.

The problem here is that traffic looks different at the packet level depending on the service it’s providing. Most video applications, for instance, send lots of packets at a more or less constant rate; packets that arrive more than a few fractions of a second late are useless,  as they represent parts of video frames or audio signal that should already have been played. On the other hand, file sharing applications don’t really care when the packets arrive, as long as they arrive eventually. Interactive traffic like web browsing falls somewhere between these extremes: users don’t notice short delays on the order of milliseconds, but will perceive longer ones on the order of seconds.

So demanding that video packets get treated the same as file sharing packets is the worst of both worlds: file sharing gets meaningless guarantees about delay, while video traffic gets capacity it can’t use. A network that can fairly allocate appropriate service to each application would be desirable. This subtlety should be appreciated by any legal definition of network neutrality.

Indeed, there is current work in network optimization, such as the IETF’s Congestion Exposure working group, which promises to increase network efficiency, performance, and fairness, that would be made impracticable by such low-level definitions of neutrality.

When providing different network service to different applications, the question becomes: what do we mean by “application”, and here is where equal access becomes important. From a technical standpoint, there is very little difference between video from provider A and video from provider B. This is not true from a business standpoint. Running a network — the business of delivering bits from point A to point B — is a very low-margin business, and network operators have an incentive to bundle higher-margin services. This is the origin of the “triple play” offers ubiquitous throughout the developed world.

Obviously, the operators also have an incentive to make sure their high-margin services — video, telephony, etc. — run better than those from other service providers. There are two ways to do this, broadly: invest in service improvement, or reduce the performance of the services of their competitors by treating the packets delivering those services differently on their network.

This is the key of the open access argument, and why network neutrality is important. If incumbent operators use the tools and techniques of network management to stifle competition, the quality of service across the entire industry suffers, new applications don’t have room to emerge, and the Internet loses its power as an engine of innovation.

In legislating network neutrality, however, there are potential pitfalls to avoid:

• The tools and techniques used to deny fair and open access by providing different levels of service are key to the day-to-day management of networks — the difference lies in whether an application is getting priority to increase overall network performance, or whether it is purely at the expense of a competing application. Legislation that restricts the usage of these tools and techniques can increase the cost of running the network while decreasing overall performance. This would be bad.
• Different classes of service at different price points can give an important signal to operators about which traffic is more important to its users. Consider the difference between personal videoconferencing traffic and videoconferencing for e.g. telesurgery³: it is much more important for the second to work, the cost of ensuring it works is higher, and that cost can certainly be passed on to the customer. So approaches that prohibit tiered service agreements at different price points are also likely to have undesirable side effects for future innovation.
• The speed of innovation even in a relatively unexciting field like network management is much higher than the speed of legislation, so legislation that names specific tools or technologies is not likely to be effective, as new techniques would simply be developed to work around the law.

It will be interesting to see how the debate here develops.

1: for American readers, the National Council is the rough equivalent of the House of Representatives, but elected by a party-plurality list system, and without any of the built-in misfeatures that lead to it being a wretched hive of extremism, corruption, and idiocy.

2: the lowest level I’ll deal with in this article, anyway; my only experience on layer 1 is ripping fiber out of a convention center.

3: there are of course other barriers to life-critical services such as telesurgery over the open Internet, but network neutrality legislation enacted today will probably be in effect for at least half a century; I am optimistic that these barriers can be surmounted over such timescales.

I can’t explain this.

In simple terms of cause and effect, the tragedy in Newtown was the inevitable result of a set of interactions among situations on the ground: firearms are ubiquitous in private ownership (why?), identification and treatment of mental illness is severely deficient (how?), to the point it’s far easier for the mentally ill to get their hands on a gun than into a bed in a mental health clinic or hospital (ähm, what?). Add to this the perverse incentive that the swarm of media coverage virtually guarantees you’ll be the most famous person in the country for a week, and mass murder starts to look like an attractive option, if the twisted logic of your disease puts you in the market for going down in a blaze of evil glory.

The natural human reaction to such events is to ensure it never happens again, or at least to try to. Here’s where my powers of explanation fail me.

“Certainly this will lead to a gun ban.”

Nope.

“But Britain and Australia banned guns after the same thing. Australians are basically Americans, right?”

Not really, and doesn’t matter.

Philosophically, one of my core beliefs is that tools can’t be evil, only people can; therefore, the right to the manufacture or possession of any tool or technology should be guaranteed within reason unless there is a compelling social interest in restriction due to external dangers inherent to that tool or technology. This puts me pretty firmly in the independent-clause Second Amendment camp¹: my problem with it is that it only applies to weapons and not to everything.

One of the only things I believe more firmly than this is that policies have consequences, and should be evaluated and enacted based upon the expected and/or measured results thereof. Here is where the rubber meets the road in the weapons control debate I hold a few shreds of hope that America might finally start having.

We’ve been here before. After Columbine, America looked at the combination of ubiquitous weapons, untreated mental illness, and schools, and decided that schools were the problem: we continued locking down our schools, building fortresses out of them. I’m not sure how locked down Sandy Hook was, but I presume moreso than fifteen years ago, and whatever fortress-building they did do would not seem to have been particularly effective.

I guess the response went a little in the direction of mental illness with the idiocy that was and is zero-tolerance, threatening serious time in a grown-up jail for kids who draw exactly the same pictures of stuff blowing up in the margins of their notebooks that I did back in school. This is would seem to be more about making sure you’re not liable when it turns out the angry kid in the back of the class is really, truly disturbed and in need of serious help — and not just fifteen and deeply awkward as fifteen-year-olds are wont to be — than it is about actual safety.

The math² works like this: let p(w) be the probability that a given individual is in possession of massacre-capable weaponry, let p(x) be the probability that a given individual is sufficiently uninhibited (due most probably to a mental illness) to be capable of a massacre, and let p(z) be the probability that a given individual has access to a school. Ignoring dependencies, the probability of a Newtown is p(w) * p(x) * p(z).

If we take as given that preventing another Newtown is a policy aim, there is a trivial solution: set p(z) to zero. This is the post-Columbine fortress approach: “Close all the schools, there will be no more school shootings.” While this argument has actually been made by fringe advocates of homeschooling, I don’t think we can really take it seriously if we want to have a coherent society.

It would be great if we could put serious effort into reducing p(x), because untreated mental illness has social costs well above and beyond those of Newtown, Aurora, and Tucson. However, America is still a generation or two away at the least from comprehensive, means-independent, explicitly socially guaranteed medical care.

The preferred European solution, the one that everyone here seems to see as self-evident, is to set p(w) to zero or very close to it: if nobody has guns then there is no gun violence. Even Switzerland, which once bristled with army reserve assault rifles and to some extent still does, is taking guns out of reservists’ homes following a 2007 incident wherein a recruit shot a sixteen year old girl at a Zürich bus stop.

That isn’t to say there’s no violence: people get beaten and knifed here not-infrequently. But the massive difference in scale between blades and guns is a question that is considered well-settled by military historians. Killing once with a knife is difficult. Killing twenty-six times with a knife is not unthinkable but requires significant skill and luck.

Given the political strength of the weapons and ammunition manufacturers and their proxy lobbying organizations, backed up by a… I don’t even know what it is, a myth America has about its revolutionary self, a myth about its conquering of an unspoiled frontier, a red-blooded machismo that masks a deep-seated fear that someone will take everything you have if you cannot defend yourself, a history of pitched battles fought in our towns, from Stono to the Alamo, from Haymarket to Homestead, from Watts to Katrina… well, let’s just say, I’m not optimistic that you can separate America from its guns without a few generations of change.

In a different political climate, one in which anything that looks like government wasn’t automatically suspect³, it might be possible to reduce p(w|x) enough to have an impact. This is roughly “guns for everyone except those disturbed enough to be likely to use them for evil”. On the other hand, given how many people get convicted of having guns they’re not supposed to — I saw so many of these on a grand jury stint in Pittsburgh that I still have the relevant section memorized: 18 USC 922(g)(1) — this is not likely to be effective without significantly reducing p(w) as well: if guns are everywhere, they’ll be easy to get even if they’re controlled.

There seems to me to be enormous practical room for the restriction of firearms ownership, and a drastic reduction of the number of firearms in circulation, without impacting the fundamental freedoms of responsible hunters, sportsmen, and yes, straight-up I-like-to-see-stuff-go-boom gun enthusiasts. Three possible practical control measures right off the top of my head: (1) security clearances for firearm owners, (2) periodic proficiency-based licensing for each firearm, (3) periodic per-firearm taxation, to reduce stockpiling of weapons that aren’t actively used for sport or practice. I doubt these are politically feasible.

So I don’t have any real ideas, I don’t have any real explanation. And then we come to the reaction. I know I should’t pay attention to anything I read in the American news, but I get questions.

“Did some guy actually blame the massacre on not enough God in public?”

Yeah, and that guy was a semi-serious contender for President for a while this year. But didn’t Mörgeli just get caught comparing homosexuals to housepets? So I think this might just be yet another case of politicians saying stupid shit.

“Did some other guy say the solution was more guns in schools?”

Almost certainly. There’s so many of those guys, though, that I don’t even know which one you’re talking about.

It’s hard to take the murder of twenty children as anything other than completely senseless. I feel for the families shattered by the massacre, and grieve for the lives barely lived. But I hope, yet, that something good could come from a conversation in the aftermath about what kind of society America really wants to be.

1: I actually believe the dependent-clause interpretation was the intended one: we needed guns because we needed militias as we didn’t have the money for an army. However, I’m not a Supreme Court justice, so at this point my opinion on the matter is moot.

2: With profuse apologies to my mathematician readers.

3: Thank you, Grover Norquist.

The Internet has been marked by its openness since its earliest days. Expanding the end-to-end principle into a philosophy of operation led to the construction of a network that was open, neutral, and stateless, allowing a boom in innovation that has allowed the Internet to largely replace the public switched telephone network as “The Network”.

Today, this openness is under threat on economic, sociopolitical, and technical fronts. Closed platforms can be more easily monetized than open ones. As the Internet becomes a public utility, governments demand more control over it. Network address exhaustion, translation, and transition challenge the end-to-end nature of the network. This talk discusses these threats, as posed by current applications and services as well as situations beyond the network. It’s intended as the starting point for a discussion about what ISOC local chapters can do to ameliorate this situation, but should be interesting for anyone concerned about the intersection between the technical side of the Internet and the realities of policy and economics.

The location is to be determined, but will be near the main railway station in Bern. Details here. The talk will be open to non-members, but since membership in the Switzerland chapter is open to any member of the Internet Society with some connection to Switzerland, and ISOC Global Membership is free, if you’re interested in the open development of the Internet for the benefit of all people throughout the world, please sign up.

Yes, I was once so proud of being part of Obama’s election in ’08 that my Facebook profile picture had me throwing my absentee ballot into a mailbox in New York for about a year afterward. In retrospect I suppose it was naïve of me to hope he’d roll back the Bush-era extensions of executive power in the service of state security, to hope that any of the war criminals of 2003 or the straight-up Wall Street criminal-criminals of 2008 would see the inside of a prison, to hope that a vote for Obama was really a vote for a less divisive America, for an America built on cooperation and compromise instead of football-hooligan politics.

I would have been uneasy voting for Obama knowing that the state security apparatus is larger now than in 2008; a free society cannot exist in the face of ubiquitous surveillance, no matter how benign we may think the watchers to be today. And while I agree in principle that the most effective way to deal with the extremely low-density threat presented by international terrorism is through targeted attacks against individuals and organizations with proven ability and intent to carry out terrorist actions, I can’t reconcile the drone war as presently prosecuted with my respect for the rule of law.

It was this unease that made it possible for such a small thing to decide he wasn’t getting my vote, when I was publicly insulted by the Obama campaign’s press secretary spokesman Ben La Bolt^*, who accused me, as part of yet another cheap hit in the endless round of slap-fighting between Team Red and Team Blue, of being either a tax fraudster or — even worse — a forex guy, because I have a Swiss bank account. Okay, I’m not quite enough of a solipsist to believe this insult was personal, and let’s face it, so few Americans emigrate that I don’t even think it crossed the man’s mind that someone might have a Swiss bank account because they live in Switzerland¹. Americans abroad are safe to score cheap political points off of because, to a first approximation, we don’t even exist². But the Obama campaign did in this moment make it clear it didn’t want my vote, so do without it it will.

On the other hand, I couldn’t vote for Mitt Romney because he’s willing to be affiliated with the Republican Party in public, which shows, in my opinion, an astounding lack of judgement. One can certainly have a debate about whether “more government” or “less government” is a reasonable starting point from which to develop policy, and one can certainly discuss to what extent behavior not condoned by a particular interpretation of a particular religious text should be tolerated in a society. Having lived in a relatively socially conservative but deeply tolerant country with a generous social safety net for the past four years, and seen how well this works, I tend to find most Republican policies long on contempt and short on practicality, and the party’s platform³, with its praise for wealth inequality and sooty air⁴ and its open hatred of homosexuals, reads to me more like a paean to life in Victorian London than anything else. But I can see how people with different beliefs and priorities might see this differently.

However, a party that condones a contempt for the basic tenets of the Enlightenment among its members and within its policies is impossible to take seriously. It’s one thing to stick your fingers in your ears and go la la la la at any mention of carbon pricing⁵: you can make a lot of money burning coal, and it’s probably too late to avert a global climate disaster. It’s another thing entirely to hold that science “is lies from the pit of hell.”

The past three and a half centuries of scientific, technological, economic, and cultural progress in Western civilization are based on a philosophical foundation that knowledge is based on the empirical, the measurable, the observable. Rejection of this philosophy by an individual places them at a disadvantage; rejection of this philosophy by a society is potentially fatal. In short, science works, and it shall continue working whether you believe in it or not, whether you understand it or not, whether you find it politically expedient to ignore it in order to shore up your support among religious fundamentalists or not.

I won’t vote for Obama because of what he has, or has not, done. But I can’t vote for Romney because of what I believe.

¹: I will, however, celebrate one happy side effect of America’s late-spring Switzerland hate-on: we don’t have to deal with the embarrassment of Michele Bachmann having a red book.

²: A vitriolic rant against the unmitigated evil that is FATCA is the subject of a future post.

³: Extra irony points to the Republicans for quoting Benjamin Franklin on the first page of the platform. Franklin was a Deist, a scientist, and a noted lover of women, in at least the euphemistic sense. The esteemed Mr. Franklin would want even less to do with today’s Republicans than they with him, if they actually knew who he was.

⁴: Its first page on environmental policy contains just one word of body text in boldface: Coal. I really, really wish I were making this up. I’m not. (Hey, look! The GOP uses WordPress, too!)

⁵: The only mention of climate change in the Republican platform is to deride the Obama administration for saying the national security implications of climate change are on par with those of international terrorism. I agree that this is a laughable comparison: disruptions due to climate change in the coming century are far more important to national security than those due to international terrorism.

^*: Corrected source of Obama campaign quote, added link to tweet. Thanks, Tony!

That said, here are a few notes on observations that came to mind while I was there.

“The Taiwanese are Friendly.” Read this in the Lonely Planet. Agree. Okay, I didn’t meet that many Taiwanese that I wasn’t trying to buy something from or to convince to drive me somewhere, but those I did met the description. It’s actually the only place I can remember being as a tourist that random people coming up to me on the street or the subway never tripped my “this guy’s working an angle” detector.

The reality of nationhood defines a nation. Seriously. Cross-strait relations are complex enough that I feel like I’d have to do diploma work in international affairs on the subject before I could speak intelligently on them, but this point seems pretty straightforward. Following my belief in results, not ideology, it is very hard to stand in Beijing, then to stand in Taipei, and to take anyone who insists that Taiwan is part of China at all seriously. The Republic of China has effective control of its borders and appears to provide all the services of a government. The civil society and economy also seem distinctly Taiwanese. If it waddles like a duck, quacks like a duck, and has webbed feet like a duck, it might be a duck. China’s claim to Taiwan is exactly as silly as Taiwan’s (longstanding official but practically forgotten) claim to China.

Taiwan insists on its existence. Taiwan is understandably a bit defensive and twitchy about this, given that it’s officially an almost-unrecognized state (due to diplomatic coercion by the PRC) and lives under an openly-advertised military threat that seems even more pointless for having spent a week on each side of the divide. Taiwan is proud of its history, and perceived status as successor state to imperial China under the Qing – dates in Taiwan are even given as years of the “Republican dynasty”, if you will, founded in 1912 by Sun Yat-sen. Much of the National Palace Museum¹ seems devoted to or at least deeply influenced by this viewpoint: there’s a treaty room containing the ROC’s diplomatic archive, going back to Unequal Treaties signed by the Qing emperors. This room also contains a wall listing the 120 or so countries that de-facto recognize citizens of the ROC by issuing visas or landing permit waivers to them.

Globalization cuts both ways. This is probably colored by the fact that I spent a lot of time in Xinyi, which seems to be the finance-district-and-upscale-malls part of town. But the malls — and there were a lot of malls, basically nothing but malls — seemed to be largely full of Western imports: Swiss watches, Belgian chocolates, Italian suits, American-designed iPhones. Taiwanese with money seem to want to spend it on the same junk we do.

Taipei borrows my favorite bit of urban design from Bern. Bern? Well, the sidewalks in much of the city are covered by the first (American: second) story of the buildings, which is both useful in a rainy place and reminiscent of the Berner Altstadt (also a rainy place).

China minus central planning looks a bit like Los Angeles² with scooters. Facetious, maybe, but a hard impression to shake. Again probably colored by my stay in Xinyi. I’ve already mentioned the malls full of pricey European crap. Add the wide streets, the tall buildings, the hills and mountains in the near distance, the endlessly-under-construction metro, the ocean-moderated climate, and the fact that almost everyone seems to get around in something personal and motorized, be it car or scooter, and the LA impression holds more true.

Did I mention the scooters? Scooters themselves aren’t that loud, but multiplied by several hundred and I was woken up every morning by the drag-strip sound of rush hour out my 12th³ floor window.

All in all, enlightening. Next time I’ll have to see more than the square mile around Taipei 101.

¹The museum’s history itself is interesting – founded in 1925 in Beijing, it evacuated its collection in 1933 ahead of a the advancing Japanese army, and moved eventually to Taiwan in 1948 as the civil war between the Communists and the Nationalists intensified. It’s got the largest collection of Chinese art in the world, in large part because it escaped the Cultural Revolution. But that’s another story.

²I’ve actually spent very little time in Los Angeles. Glendale and Burbank, sure. So when I talk about Los Angeles, I’m talking about my impression of Los Angeles, which may or may not reflect its reality. I can however say with all honesty that in my experience, the air’s much, much better in Taipei, and the metro quite a bit more useful.

³My room was on a floor labeled the 15th, but the building lacked a 4th, 13th, or 14th floor for reasons of good luck.

Bagels boiling, Wallisellen, 7 November 2011

The problem with taking a week in New York and coming back to Zürich is that you miss the bagels. Bagels, however, are made by humans, and we are humans, so how hard can it be?

Turns out, quite. Basing our recipe on one we found on an American cookbook written in German (the author’s blurb assurs us that her husband is a real, actual American from Bedford, Pennsylvania, of all places), the results of our first attempt are shown here. Ariane gets most of the credit for these, she made the dough while I was still reading the Economist on the 751. I just did the photography, and formed a couple of rings.

Bagels finished, Wallisellen, 7 November 2011

What we got were a little too soft on the inside, a little too hard on the outside, and, while they were certainly among the best-tasting bread to come out of this kitchen, they weren’t really bagels. Things to try next time:

1. tweak the recipe (the Internet has a lot of opinions on how to make bagels, because the Internet has a lot of opinions on everything; sadly, most of these are wrong.)
2. boil longer (two minutes a side seems a bit short) with saltier water
3. let them rise a bit less (because they fell quite a lot in the oven)
4. buy bagels in New York like everyone else (the flight’s not free, but hey, the dollar’s cheap…)