Archive for April, 2007

The second Access to Knowledge conference at Yale is going on this weekend.  (Conference site is here.)

Here's the wiki page for the Internationalized Domain Names panel (it has notes on it).  It was a distinguished panel, with Dr. Wei Mao, Ram Mohan, Hong Xue, Peter Yu, and Milton Mueller speaking.

Ram Mohan's talk about the need for a sustainable policy framework for IDNs was particularly interesting (see the wiki for more detail).  He talked about India as a case study — 22 official languages, newspapers offered in 87 languages, and only one ascii-based DNS.  His view is that adequate technical and protocol standards already exist for IDNs, and that what's incomplete is the policy to govern the rollout of IDNs.  He had several suggestions for what principles should guide that policy, including avoiding user confusion, adopting the UDRP, and consulting with governments.

Hong Xue pointed out that the ICANN Board passed a resolution in Sept. 2000 saying that ICANN “recognizes that it is important
that the Internet evolve to be more accessible to those who do
not use the ASCII character set.”

During yesterday's Ofcom-immersion, it was clear to me that Ed Richards wasn't interested in having some form of government-internet access-involvement story unfold in his country.

Ken Zita of Network Dynamics (who was just terrific) then got up and pointed out that in Asia they are unafraid to talk about telecommunications as part of “industrial policy.”  (For me, that phrase connotes smokestacks and Margaret Bourke White photos, so I'd probably say “economic policy.”)  For example, Zita continued, in S. Korea the government said where they wanted to go, invested in research and development, [and invested money and made micro loans], and they're now seeing 70% of adults (not just kids) involved in online social networks.  Very high speeds, very low cost.  Hong Kong is also not embarrassed to talk about economic policy and telecom.  Zita noted that opening up spectrum, as Ofcom plans to do, is a traditional telecom regulatory move.

In response, Richards said that, yes, it's true that some countries in Asia have this policy approach to telecommunications, and a couple of countries are doing this in Europe.  [Paraphrasing]:  “It's an overarching political choice.  It's not something we feel is a good idea in the UK.”

[Update -- see snarky article here about Ofcom's failures.  Thanks to Dirk van der Woude.]

Today, we're reading that the Indian government “proposes to offer all citizens of India free, high-speed broadband connectivity by 2009.”

Well, so far in the US we also haven't linked economic policy explicitly to universal highspeed internet access (although we should), so the best we can do is hope for more enlightened spectrum policy and assume that someone will show up to help.  The Commission is making noises about possibly being interested in “open access” requirements for part of the 700 MHz auction.  (Commr. Copps:  “I .. am pleased that we ask general questions about how the 'open-access' proposals made by some might encourage the development of a sorely-needed wireless 'third pipe' in the broadband market.”)

According to Harold Feld, “this condition would prohibit the licensee from engaging in retail wireless service sales (on this spectrum), and require the licensee to lease wholesale spectrum on a non-discriminatory basis.  If adopted, it would make 30 MHz of 'beachfront' spectrum available for lease in every market in the U.S.  This proposal is in addition to the 'Frontline' proposal, which would create a public-private partnership with public safety to make up to 22 MHz of spectrum available under an 'open access' condition, but such commercial use would be 'preemptable' by public safety.”

So it's not economic policy, it's tinkering with spectrum policy, but it's better to be asking questions that might make wireless highspeed access a realistic alternative here in the US.

Ed Richards, Chief Executive of Ofcom, was at Columbia today. 

He reminds us that Ofcom was formed in December 2003 “as a response to convergence.”  They have authority over broadcasting, telecoms, spectrum management; they also have antitrust authority within those sectors, although Ofcom is “entirely dependent” of the UK government. 

Richards says that 95% of spectrum in the UK has been subject to command and control regulation.  So his vision is to move by 2010 to a “predominantly market-led” scheme to cover 70% of spectrum, “available for use by any technology,” and subject only to avoiding harmful interference.  I believe Richards said that Ofcom is encouraging the release of spectrum by law enforcement — something like a total of 350 MHz below 3GHz.

On net neutrality:  Ofcom sees NN like the Janet Jackson wardrobe malfunction issue — when the issue is understood, “we in Europe wonder what the fuss is about.”  But Richards does think that every country will have to consider this issue, although outcomes will vary.  He thinks there may be some advantages to consumers in treating different applications differently.  His major point is that he has there is a different view in the UK.  “If you have network operators who are dominant or have market power, charging for prioritization may be anticompetitive, then the regulator should intervene.  But if a network operator has no market power and is charging for prioritization, then it's a different question; their activities shouldn't be automatically construed as anticompetitive or necessarily regulated.”

Richards continues, saying that the NN debate does give us insight into importance of disclosure to consumers — consumers should be able to switch providers, and they should know which ISPs are making prioritization decisions.  This should be an obligation of suppliers to communicate this information to consumers.  In particular, he says that Ofcom is actively exploring whether network operators whose traffic shaping activities change materially should have to tell consumers — and if these changes are significant consumers should be allowed to break their contracts with the provider without penalty.

Dave Burstein asked a strong question:  BT isn't giving 24 Mbps to anyone (really) — it's 1 Mbps up and 8 down.  This means that London is behind Paris in terms of broadband.  How do you change that?  Give incentives to BT? but that might be a wasted subsidy.  Or do you provide a stick instead — the regulator can point out that the existing copper is obsolete and lower the base rate that BT is allowed to charge.

Richards responded (paraphrase):  Yes, your facts are right, but “there will be change”.  Why — because people deserve it? because we're behind Paris?  I hope that we are not as concerned about politicians are about Paris.  I know that broadband is a big yardstick of national machismo, but you have to think about these things carefully.  You shouldn't sell your soul in the short term.  You could end up losing all the longer-term dynamic benefits of competition. True, we may be entering a time when the economies of scale/barriers to entry for new providers are even more profound, but you shouldn't leap ahead with a policy response out of a national sense of pride.

And here's where I heard it all come together. 

Richards said (paraphrase mine):  We won't give network providers money — instead, we want to let the market make the decision.  Are consumers willing to pay for a higher-speed broadband network?  It has to be be funded by consumers.  I see no case for funding broadband by the government.  A national response of government funding would likely waste taxpayers money, preempt the market, and re-create a state monopoly. So we have to encourage consumers to pay more — they need to like the service proposition that they get.  This can include content rights, bundles of services, etc.

Someone sent me a copy of Seed magazine today.  I have to say that I'm really grateful — I'm interested in everything in this May/June 2007 number.  Right at the top of the cover is the headline:  How Complexity Arises.  And, hey, there's a drawing of Lee Smolin on p. 41, and a few words from him about whether the laws of nature evolve: [T]here is never perfection, but always time and change.

But the piece that grabs me is a conversation between David Byrne and Daniel Levitin about music, language, and memory.  Oh, it's great.  Levitin suggests that “music might be evolutionarily older than language.”  Byrne talks about the emotional effects he can intentionally pull off when performing.  They both understand that “we use art and music to communicate so many things that language won't.”

Musical experiences take us out of ourselves, Levitin says, and induce a state of half sleep, half wakefulness:

We don't really have the ability to explain how it happens or why.  But it does seem to have something to do. . with this balance between seeking order and predictability and violating that order and predictability.  And when you have a complex pattern of rhythm or pitch, which is what music is, you relinquish some of your control.

Someone asked me today to use this post to describe what it's like to play music.  I think David Byrne and Daniel Levitin are better people to ask, by far (so buy the magazine).

The instrument is always greater than you are, and in a sense there's no magic to it — it seems as if you practice and have the guidance of a good teacher, you'll be a better player. I'm beginning to understand that strength comes from not using that strength, but from instead having an understanding of what you're doing and why, in great (but humble) detail. The Art of Practicing is a good guide to the meditative focus that musicians have. Sometimes I wonder what people do with their time who don't have to practice. 

But the real point is to play with other people.  The best description I've read of what it's like is in An Equal Music by Vikram Seth.  To listen completely to what's going on around you, and to react without time to react, is a tremendous joy.  When things are going well, and real music is emerging, there's a collective neurological cascade that can't be described but certainly exists. 

Right now I'm working with a singer and a pianist on a program of only delightful music.  Not kidding — a bunch of tangos, little Kreisler pieces, Schubert's Shepherd on the Rock, Mozart and Bach duos, songs by Weill and Faure.  (Not one of these pieces was originally written for the viola, but I am shamelessly, enthusiastically bending them to my will.)  Playing these things with other people, once you're flying along and you've transcended all petty technical difficulties — such as, for me, what clef I'm playing in – is pure happiness.  That's what it's like.

Anthony Lewis is here at Cardozo this evening, speaking about whether to protect the confidentiality of the journalist-source relationship by statute. 

He has several core messages.  First, journalists shouldn't always be entitled to this protection.  What if a story has been planted with journalists by secret police or other bad actors?  What if a journalist has defamed someone by passing on the statements of confidential sources?  Should we provide no remedy to those whose reputations are ruined by journalists? Second, who is a journalist?  If there are 40 million bloggers, and they are witnessing the world around us, should they be kept from their citizen-duty of testifying?  Third, can't we trust courts to balance the benefit of the confidential information to society against the harm that would be caused by disclosure of the confidential source?  Lewis is 80 this year, but boy is he acute and well-spoken.  He doesn't think it's possible or appropriate to write a shield law that will work, particularly given that the government will demand a “national security” exception that will have the effect of swallowing the stated rule.  He wants reporters to look to courts for protection of their confidential relationships with sources on a case-by-case basis.

Max Frankel then gets up to respond.  His is the battlefield view.  We live in a garrison state, he says, and we have since the beginning of the Cold War.  Our presidents can classify millions of documents by executive order, and reporters can't work with any of this information unless they can have access to confidential sources.  Of course reporters must write, to protect our nation and reveal the bad acts perpetrated in the name of national security.  They must “publish and be damned.”  And if a few prosecutors want to seek disclosure of confidential sources, well, let 'em try.  Reporters should make few promises about confidentiality, but when they make them they must stick to these commitments.  Don't let judges weigh the value of particular snippets of information to the public — after all, these bits are woven together by reporters to build stories, and it's impossible to say what one disclosure may have contributed to the ecosystem (my word, not his) of a story.

Victor Kovner is next.  He optimistically and with great precision tries to clarify the discussion in a lawyerly way.  Reporters have an absolute privilege in many states not to serve as witnesses.  We have this law on the books in New York, and law enforcement has been able to continue its work.  And reporters have a qualified privilege not to reveal their sources.  This privilege stems from the First Amendment, from statutes in many states, and from the federal common law.   The Department of Justice's own internal guidelines on this subject, written under John Mitchell in the early 1970s, embody this qualified privilege:  only go after a reporter's confidential source when the information is material, when it is closely mapped to the underlying claims, and when there is no other way to obtain it.  (“Materiality, criticality, and exhaustion.”) 

Kovner points out that Special Counsel Patrick Fitzgerald wasn't bound by these DOJ guidelines, and so he went after Matthew Cooper and Judith Miller for their sources.  As it turns out, the quality of their information wouldn't have overcome the qualified privilege, had the privilege been applied (or so Kovner believes).

Kovner finds recent opinions on the subject of the existence of a federal common law qualified privilege by Judge Tatel (DC Cir) (concurring) and Judge Sack (2d Cir) (dissenting) to be particularly well-written.  Counting noses, he's suggesting that there are judges who believe in the existence of this qualified privilege, legislators (at least one former journalist) who understand its importance, and prosecutors who already feel themselves bound by it — and he even thinks that President Bush wouldn't veto a federal shield bill if it were presented to him.  The law would have to have a national security carveout, he has to concede, but he seems to think that would be better than no law at all.

But, as of the time I had to leave the room, Kovner hadn't explained how we'd decide who is a journalist.  Reporting from Cardozo Law School, I'm…

I remember going to an FTC workshop in November 1999 about “online profiling.”  At that workshop, DoubleClick was getting a lot of attention.  People were concerned that DoubleClick was going to combine extensive personal offline information from Abacus (including the catalog-buying habits of tens of millions of households) with its own databases of online profiling information.

It was quite dramatic, as I recall.  Not too many people knew what a “network advertiser” was at that point, and DoubleClick seemed surprised at the focus on its plans.  (People were struggling to understand “cookies.”)  I remember thinking that offline direct-mail companies had an awful lot of information to work with and no one ever seemed to tell them what databases they could use.  But I dimly recall that a DoubleClick executive promised at that meeting that they wouldn't merge these offline/online databases.

A while later the Network Advertising Initiative issued self-regulatory principles that required sites using network advertisers (sites allowing third parties to place and read cookies on visitors' hard drives) to disclose that cookies were being used and to make it possible for users to opt out of those third-party ads.

And then, as far as I could tell, the issue went to sleep for a while.  There are always concerns about identity theft, but these concerns aren't linked to online profiling by advertisers.  There are concerns about governmental over-collection of data — but, again, it's not the private advertisers who are the problem.  Back in 2002 there were concerns about tracking in connection with medical and financial data, but the lawsuits seem to have boiled away. Companies routinely disclosed that third parties were cookie-ing on their sites.  And users got used to targeted ads.

Now, with Google's acquisition of DoubleClick, beaconing and cookie-ing and tracking are newsworthy again.  The Electronic Privacy Information Center has filed a complaint urging the FTC to keep Google from completing the deal. 

Google, for its part, says that it doesn't plan to combine the kind of cross-site information DoubleClick has with its own information.  Google also says that users should have the ability to opt out of cookies.  (EPIC's complaint is that Google matches queries to IP addresses and doesn't allow users to request that this linkage not be made.)  And Google has recently undertaken to anonymize its query data  (by changing some of the bits in the IP addresses and by changing cookie information) after 18-24 months.  (FAQ here.)  Google resisted the DOJ's over-reaching COPA-related query request last year and has joined with other companies in asking for baseline US privacy legislation.

EPIC's complaint doesn't actually say that Google has lied about its practices or that it is violating existing US statutes.  Its complaint is that Google doesn't give adequate notice of its data collection practices (collecting user search terms with IP addresses), but Google's home page is notoriously and intentionally simple.  People interested in finding out what Google is doing can navigate their way to find out.  I remember being presented with a good deal of privacy-related information when I downloaded Google Desktop, and choosing (I believe) not to have it “phone home” to Google. “Fair information practices” and the OECD Privacy Guidelines are not part of US law (yet).

Of course, the legal details don't matter that much here.  Google's services are so popular and it has such a big presence that its acquisition of DoubleClick raises suspicions all on its own.  There's something about cross-site, invisible-to-most-people network advertising that gives people the willies — even if Google can't access DoubleClick's clients' data. 

This has the feeling of November 1999.  An acquisition of a complementary company in the advertising business — an event that would go unnoticed in the offline world — is making people worry about what might happen to the privacy of their machines.  So the lawsuit is being reported on (InformationWeek article here), and we'll probably have several workshops and more legislative efforts.

Another fine report from Pew, this one about teenage online social networking practices.  There's a lot to work with here — from how many teenage kids are posting online profiles (61%, most of them 16-17, vast majority on MySpace), to how careful they are about how much personal information they post online (more careful than you might think), to how much their parents know about what they're doing.

It's this last point that is particularly encouraging.  The big split in the most recent Supreme Court COPA decision is between Kennedy and Breyer, with Kennedy saying that there are plenty of choices of relatively-effective (and certainly less-restrictive) filtering tools out there for parents to use, and Breyer essentially saying that parents are helpless so mandated shields of various kinds should be put in place to protect kids.   It turns out that, in fact, parents are knowledgeable and are giving advice to their children about what to do online.

In comparison to television and video games, the internet is a much more parent-regulated piece of technology.  Take a look at this:

More than eight in ten parents (85%) of online teens said that they had rules about internet sites their child could or could not visit, and a similar number (85%) said they had established rules about the kinds of personal information their child could share with people they talk to on the internet.

Most kid-used computers are in public places in homes; most parents are checking up on what their kids are doing online; and more than half of parents with online teens use filters.

This is a serious survey — I hope it comes in handy when COPA comes back around again.

If you want to get access to a Congressional Research Service report, you have to hope that someone has made it available to OpenCRS.  You may be out of luck.  And that's not a good thing.  Or, you may never have heard of OpenCRS, and you'll be missing out.

Here's an op-ed by Ari Schwartz of CDT on the subject.  We — all of us — paid for these reports, and they should be freely available online.  (They're often really great reports.)

In a little more than a year, members of the public have downloaded
more than 3.5 million CRS reports from OpenCRS.com. Making the full
catalog of these reports readily available over the Internet will sate
those demands and help produce a better-informed electorate.

I really like the candor of this quote:

The . . . telling defense for the decrepit CRS policy [keeping the reports offline] came
from former Rep. Bob Ney, R-Ohio, who was quoted as saying: “Let’s say
that I’m working on an issue and I’m trying to look for some research
that helps me to get my point across and, all of a sudden, the
Congressional Research Service sends me over something and I read it
and I say, ‘Oh, no, that’s not going to help.’ Let someone else do the
research. Why give your opposition free research?”

Hah.

=== in other news, welcome to the blogosphere, Rob Frieden!

The FCC is asking whether it could be collecting better data on broadband deployment:

Notwithstanding the robust statistics and the more granular broadband data that have been reported on FCC Form 477 beginning September 1, 2005, the Commission continues to consider the need to improve its data collection, particularly regarding data reflecting broadband deployment and availability in rural and other hard-to-serve areas, and also regarding subscribership to new broadband-enabled services such as interconnected VoIP service.

Th Commission has come in for a lot of heat (from the GAO and others) for always having good news on broadband deployment.  Penetration is increasing! because if broadband is somehow available in a zip code, we count everyone in the zip code as covered –no matter how many actual subscriber there are.  We're doing better and better! because we count wireless carriers as an option, even though they're owned by the same old telcos and they're providing very-high-walled gardens.  So now's your chance to help them out in the data collection techniques department.

And why you're at it, you could file comments by May 16 about what's actually going on in the highspeed access world.  Here's the Notice of Inquiry.  Just may take a few years.  But, as I reported a while ago when the Commissioners met to consider this not very dramatic step, they're saying that they have zero evidence that there is any blocking or interference taking place.  Of course, no one really knows, because outside researchers aren't allowed in.  And our expectations are so low, and there are alway a lot of explanations as to why X or Y might be happening.  (E.g., Why couldn't I upload using my Verizon connection last night?  could be something having to do with a hop far away.  Who knows.)

So They Have a Lot of Questions.  Should keep everyone busy for a while.  We can chat about all this with our friends in the U.K.  Meanwhile, in Japan, Korea, Amsterdam, and a host of other places, they've got other things to think about.

I've been asked to write 5-10 pages answering the question “what's broadband good for?”  Although I'm tempted to say “and why do we have light?” I realize that it's meant as a serious question.  So I'll try to frame the answer in a transformative thought-provoking way.  I was going to post a substantial set of paragraphs on this, but I've just had a long session dedicated to uploading new pages to my site (uploading was just not working for me tonight). 

We used to think that the “content” “on” the internet was as diverse as human thought.  (That's a reference to the CDA case, now ten years old.)  Now we know that it's not “content on” – it's “communication using.”  In fact, pre-packaged content is lessening in importance.  We're still at the very beginning of all this. The screens we have could still be used in more interesting and complex ways, and higher-speed connections will make that complexity more approachable for everyone.

And we'd like to be able to upload.  More tomorrow.