Archive for August, 2006

Tom Evslin recently spoke about the need for volunteer monitoring of broadband networks.  There's no question but that making network operation visible would help — we'd be able to tell whether non-minimal discrimination targeting particular applications was going on.

But it's a tough logistical problem, requiring adept coordination of decentralized data and people:

Each volunteer would download software that triggers their computer to send out test packets called pings to various websites. Because pings automatically trigger a return packet, they can be used to measure the speed of a connection between two computers. Each probe PC reports its results to a central server that can then work out from all the ping times whether packets from certain websites are being deprioritised, and if so by which broadband providers, says Evslin.

The providers will have all kinds of sensible reasons why they're managing their networks in particular ways, and plenty of room for arguments about why any data gathered is inherently inaccurate. But I like the SETI approach, and Tom's idealism is inspiring.

It would also be good to involve actual network researchers to do controlled experiments with the cooperation of the broadband providers.   That's the kind of thing that CAIDA does well.  But it needs funding and support from the right players. It's a big step to require “private” network operators to support research about the operational nuances of their networks. 

We make progress when we can see what we're doing, but on this particular front we'll probably be operating in the dark for a long time.

A Bill of Rights follows the U.S. Constitution to protect us from the depredations of a powerful government to which we have ceded authority.  The existence of a Bill of Rights assumes that there's a powerful entity against which we need protection.

In the draft Stevens bill, the Consumer Internet Bill of Rights assumes that the broadband network access providers are powerful — but it's not clear that the IBR provides much protection.

First, labeling:  users are “consumers” (not creators) or “subscribers” (think packaged content), and the IBR doesn't apply to video services “in which Internet service is not the primary service.”  Because the chief goal of this amendment is to put the incumbent telcos in a position to become broadband video service providers, this exception substantially lessens whatever protections the IBR creates.

The preamble re-uses language that leads into Section 230 — a section that shields interactive computer services like Yahoo! and eBay from liability for material created by others — to suggest that network providers should not be subject to regulation.  This is the call of the network companies:  protect us from regulation, and you'll be protecting the internet!  If there were true competition for broadband access, that call might make sense — as it is, it seems cynical.

But that's just the preamble.

The IBR itself “allows” “subscribers” to access applications of their choice.  That's fine as far as it goes, but it doesn't go nearly far enough.  As Ed Felten has made clear in his thoughtful explanation of the technical issues animating the network neutrality debate, “non-minimal discrimination” by network providers could make for a negative user experience of non-prioritized packets.  Users wouldn't be providing their own attention-feedback for those non-prioritized packets — rather, the network providers would be choosing winners and losers.  This could have (in my view, would have) negative effects on the evolution of the online ecosystem. 

Again, if there was competition in the market for broadband access, this kind of non-minimal discrimination wouldn't be such a problem.  As it is, however, we have gentle competition among like-minded giants – and very few choices.

The IBR states that “subscribers” can “connect any legal device of that subscriber's choosing to the Internet access equipment of that subscriber, if such device does not harm the network.”  Way back, AT&T used to claim that any foreign (not manufactured by them) device would harm their network — they kept that up for years, and the FCC went along.  AT&T even claimed that non-AT&T covers on telephone books might be included in this argument. Not until 1968, eight years after a court decision asking the Commission to re-examine its policy, did the FCC admit that these restrictions were “unreasonable, unlawful and discriminatorily applied.”  

There's no telling what the network providers could claim harmed their networks — and it would take years to litigate such a dispute.  In the meantime, the device manufacturer would have knuckled under.

Subscribers, the IBR says, will be free to exercise these “rights” as long as they fit with the limitations of the internet service the subscriber has bought.  Through their terms of service, the network providers will simply be able to say that they don't offer pure internet service — so some applications, some devices, some uses are off limits. Yes, to the extent a provider offers pure internet, they'll have to do that without requiring that the subscriber purchase other services from them — but they don't have to offer pure internet.

A network provider will be free to do whatever it wants to in the name of “maintenance” or “network management” or “unlawful” uses — and what those terms mean will be within the discretion of the network provider. 

The FCC will enforce the IBR, according to the draft bill, through its own litigation process.  The Commission is prohibited from issuing any rules or generally-applicable litigation results — everything will be one case at a time, one drawn-out piece of litigation after another.  This means that most mild irritations/latency/jitter that would cause a user to fall back into the comfortable services provided by the network operator itself will never be made public. 

The bottom line:  The IBR doesn't shift the current situation.  Network access providers have all the power and discretion they want — and, indeed, this bill if enacted would codify their right to packet-discriminate.

Without visibility as to what's going on (no neutral studies allowed!) or true competitive pressure, users will simply take what they can get, will buy the network providers' packages, and will settle back into their couches.  That's not ideal.

On the other hand, as I've said in the past, the only ex ante rule that will make unfettered internet access a reality is mandated structural separation.  We'd need to turn transmission into a utility in order to change the environment. 

There's a plain-spoken opinion out of the Eastern District of Michigan today. 

The bottom line:  There is no such thing as “inherent power” to violate the Constitution or the laws of Congress:

We must first note that the Office of the Chief Executive has itself been created, with its powers, by the Constitution. There are no hereditary Kings in America and no powers not created by the Constitution.

The court sees the public interest here — to uphold the Constitution — and has granted plaintiffs an injunction stopping the wiretap program.

A set of plaintiffs had the courage, led by the ACLU, to challenge the Bush Administration's warrantless wiretapping program.  A federal judge has had the courage to find that the “state secrets privilege” doesn't knock the case out, that plaintiffs have suffered concrete enough injury to allow them to have standing to sue, and that the National Security Agency has acted in contravention of FISA and the Constitution.

The opinion is squarely and simply written, and careful in its exegesis of “state secret” and standing cases.

This Administration has successfully used the “state secrets privilege” in several cases during our never-ending War on Terror. Their claim is that they can't defend this case without revealing state secrets. Courts have very frequently allowed this privilege to result in having entire cases thrown out. 

In this matter, however, plaintiffs are using only evidence that the Administration has stated publicly — that for the last five years (1) the President has authorized (and reauthorized) a wiretapping program that (2) operates without warrants, and (3) “targets communications where one party to the communication is outside the United States, and the government has a reasonable basis to conclude that one party to the communication is a member of al Qaeda, affiliated with al Qaeda, or a member of an organization affiliated with al Qaeda, or working in support of al Qaeda.”

This can't be secret — the Administration keeps confirming the existence of this program publicly, almost defiantly.

Although the plaintiffs can't establish their data-mining claims without the use of secret information, the court found that they can certainly establish their case questioning the legality of the wiretap program. And the defendants, our government, has arguments about the legality of the program that don't rely on secrets (but do rely on incorrect legal arguments).

The next big hurdle for the plaintiffs was standing — their ability to show that they've suffered injury that is concrete and particularized and results from what the defendants did. This could have been tricky — because no one knows who has been surveilled or how that information is being used.

But the court deals with this easily, because among the plaintiffs are lawyers whose foreign clients will no longer speak to them on the telephone or online for fear of wiretapping — and so the lawyers have to travel. Expensive, particularized, and concrete injury.  There are also scholars and journalists in the group who have been unable to communicate because of these same fears, and must travel to conduct research and interviews.

More fundamentally, the court sees beyond the formal standing requirements to this essential point:

Although this court is persuaded that Plaintiffs have alleged sufficient injury to establish standing, it is important to note that if the court were to deny standing based on the unsubstantiated minor distinctions drawn by Defendants, the President's actions in warrantless wiretapping, in contravention of FISA, Title III, and the First and Fourth Amendments, would be immunized from judicial scrutiny.

It was never the intent of the Framers to give the President such unfettered control, particularly where his actions blatantly disregard the parameters clearly enumerated in the Bill of Rights.

This court believes it has a real case or controversy before it.  Denying standing would render the plaintiffs' claims unreviewable and unreviewed.

The strongest part of the decision, and the call to remind all of us why the rule of law matters, comes near the end. FISA was set up to balance executive needs against the privacy rights of United States persons. We've established many special procedures to make it easy for the executive to get approval for wiretaps. To ignore this statute is to ignore Congress — and that's not allowed.

The wiretapping program here in litigation has undisputedly beeen continued for at least five years, it has undisputedly been implemented without regard to FISA and of course the more stringent standards of Title III, and obviously in violation of the Fourth Amendment.

Here, the President has ignored statutory instruction — and so his powers are at their lowest ebb. The Authorization for Use of Military Force on which the Administration relied said nothing about wiretapping.

This may be too plain and workmanlike an opinion to survive the intellectual legerdemain of the best hired-gun appellate lawyer — but it's right.  It's a good day for the rule of law. 

Today, some juicy and strange legislative provisions in the draft Stevens bill.

“Misleading words or images on the Internet.”  That's how the section that begins on p.226 starts.  It's a mighty big subject — but we do have a focus, and acts defying this language will trigger fines and imprisonment:

It is unlawful for any person knowingly to embed words, symbols, or digital images into the source code of a website with the intent [to deceive another person into viewing material that is obscene] [to deceive a minor into viewing material that is harmful to minors].

I'm guessing this is a repeat of the old meta-tag worries — “secret” codes designed to fool search engines into listing a site, much to the horror of the inadvertently clicking consumer.  This seems to be a legislative solution in search of a problem.  First of all, many search engines rank according to linkage, not according to those ol' secret codes.  And the engines will give a few lines of text — which won't be in the source code for a site but will instead be visible.  So what's the problem?

The “harmful to minors” definition is interesting — it attempts to create a nationwide “community standard” for this material:  

[Harmful to minors material is material that] is patently offensive to prevailing standards in the adult community as a whole with respect to what is suitable material for minors

COPA targeted speech that was harmful to minors according to “contemporary community standards.”  The government argued that that standard wasn't problematic online because most places in America have the same standard.  In the summer 2002 Supreme Court decision, Justice Thomas said that a web publisher had the burden of compliance with the most restrictive jurisdiction into which material was sent.  Justice Breyer didn't agree:

“To read the statute as adopting the community standards of every locality in the United States would provide the most puritan of communities with a heckler's Internet veto affecting the rest of the nation. . . The technical difficulties associated with efforts to confine Internet material to particular geographic areas make the problem particularly serious.”

Justice O'Connor also agreed in COPA that the use of “community standards” might be problematic online. 

Faced with the oddness of the Thomas position and the concerns raised by Breyer and O'Connor, the drafters of this Stevens bill decided to throw “communities” overboard completely.  The government must be assuming that “prevailing standards of the adult community as a whole” are obvious enough to avoid an overbreadth challenge.  It's easy to predict that there will be substantial and well-grounded complaints about this language.  I don't think the people in my neighborhood (Greenwich Village) have the same standards for what is “harmful to minors” as everyone else, and I don't think there's a “prevailing standard” nationwide.  Who prevails?  What on earth does that mean?

The next section prevents children's programming from presenting interactive commercial matter.  I know that children are big marketing targets, but why not let them interact with the cereal?  How can that hurt? If parents don't like the programming, can't they just turn it off? (ducking) 

The last bit suggests a study of “bus-casting” — broadcasting on school buses.  Now, I'm against being forced to watch anything — on planes or buses or in taxis.  So I'm all for looking closely at this.  But I still wouldn't mind interacting with the cereal (or licensed character) if I were six.  Heck, children are exposed to so much media they'll be totally jaded — they won't care if the licensed character can wave to them.  Stifling interactivity is likely to have negative economic effects on this programming — look what COPPA did to children's web sites.

Tomorrow:  the vaunted Internet Consumer Bill of Rights.

We are still in the middle of the mammoth Stevens bill.  The world is spinning around us — interesting things are happening, summer is ending in the U.S., back-to-school ads are everywhere — but those in Congress are still on vacation.

The next section of the bill (starting on p.206) requires televisions to have digital tuners after March 1, 2007 and begins the process of setting energy standards for the converter boxes that analog TVs will use. 

Starting on p.208, the bill requires cable operators to carry TV stations that are being transmitted in the cable system's locale.  Until 2014, the cable operator can take high definition TV and transmit it as “standard definition” television.  Small cable systems can offer digital TV programs in analog format.  Satellite carriers will have to do the same things that cable systems do.

Section 702 (p.216) reinstates the video description rules that were overruled by the D.C. Circuit in MPAA v. FCC (2002).  Take THAT!

A final section in this Title VII exempts Spanish-language TV stations close to the US-Mexico border from making the transition from analog to digital. 

Title VIII (beginning p. 221) is titled “Protecting Children.”  It requires video service providers to comply with future FCC regulations on child pornography. 

Preventing the abuse of children is one of the few almost-international norms we have.  But the next section requires a very careful look:  it requires “warning labels for websites depicting sexually explicit material.” 

This is an old, bad issue.  Congress keeps working on laws that will have the effect of limiting access by adults to speech that is legal for adults.  This is an incredibly broad new entry in this old, bad story.  The definition of “sexually explicit” is vague — it covers “material that depicts sexually explicit conduct.”  That could be almost anything in any popular movie today.  The FTC is supposed to develop labels that “will inform any person who accesses that website of the nature of the material and to facilitate the filtering of such pages or screens.” 

A recent letter by the Center for Democracy & Technology makes clear why this is such a bad idea.  The labeling requirement will hurt voluntary labeling efforts, won't prevent kids from accessing porn originating from outside the U.S., is tied to a disproportionate criminal sanction, and will prompt sites to self-censor.  Enacting such a provision will lead to years of litigation and won't help to protect children.

That's enough for today.  The next section, which targets “DECEPTIVE WEBSITE DEVICES [designed] TO TRICK INDIVIDUALS INTO ACCESSING MATTER THAT IS OBSCENE OR HARMFUL TO CHILDREN” looks too depressing to deal with until tomorrow.

I'm at the San Francisco airport.  Kudos to them — it took five minutes to get through security.  I'm not quite sure why this was so easy.  Maybe everyone gave up on flying today.  This gives me a moment to look at the next section of the Stevens bill, starting on p.195.

It's called the Wireless Innovation Act of 2006, but it doesn't seem very innovative.  It says that “certified unlicensed devices” (oxymoronic) may use “eligible broadcast television frequencies.”

The certified unlicensed devices have to submit to Commission testing (“to protect licensees from harmful interference from certified unlicensed devices”) and have to be remotely disable-able if the Commission decides they're dangerous.  That is a big deal — that means that these devices always have to be controllable.  A PC isn't remotely disable-able, but it could certainly be something that could use broadcast frequencies.  This requirement could have substantial chilling effects on whatever newfangled “certified unlicensed devices” the Commission has in mind.

It's a step in the right direction to open up spectrum for uses by wireless devices, but the testing and disabling requirements seem to be wrong-footed.  This means that all of these devices will be subject to Commission approval/testing, and that they be hardwired in some way to a central facility.  Not the most flexible approach.

The next section makes more sense — it requires manufacturers of analog TV sets to let buyers know that starting in Feb. 2009 we'll be cutting over to digital broadcasting and they'll need a converter.  Because a huge majority of Americans already get their television over cable facilities, this transition should work smoothly — and we have a couple of years to get the news out.  The bill requires both the Commission and broadcasters to help publicize the transition.  Yet another reminder that the Commission continues to mandate that broadcasters control content in particular ways (even though the justification for this is increasingly weak).  

There's an interesting provision requiring TVs to be able to block “programs with a common rating.”  As long as the TV owner gets to choose what and whether to block, having this capability makes sense (if the form this capability takes isn't mandated).

Yet another multi-stakeholder group is suggested by this section — the DTV Working Group.  This is like the group that's supposed to recommend “audio flag” rules.  Who's in this group, and who they claim to represent, will be a source of interest and concern.   

Tomorrow — p.205 — the “requirements for digital television sets and certain other equipment.”  Yes, this deregulatory bill gets complicated.

After a short break for an incredibly dense and interesting IP conference (thanks, Berkeley — 84 papers in two days!), we're back to the Stevens bill.

The next section (starting on p. 184) is titled the Community Broadband Act.  What it does is throw up roadblocks for communities that want to provide their own broadband services.

The section does preempt laws that prohibit cities from providing broadband of their own.  But then it goes on to make it quite difficult for the public entity to do this — taking away with one hand what the bill appears to give with the other.

The city/municipality can't give itself any preferential treatment with respect to rights of way or permits of any kind (even though the city presumably controls those access points). 

The city/municipality has to open up a bidding process to private companies, has to publish all the details of its plans, and has to act as if (as far as I can tell) it was merely an alternative bidder for the provision of these services.  As a final warning to cities, the bill provides that no federal funds will be available to bail them out if things go wrong.

The goal of all of this is to make it tricky for cities to avoid the costs and gatekeeping associated with incumbent local telephone companies when providing broadband services.  These incumbents want to make sure that cities find it very difficult to do for themselves what the local phone companies would like to do.

The Video and Audio Flag portions of this draft bill are among the most pernicious.  (The draft is here.)

The first part of the bill claims to be aimed at protecting digital broadcast video content, but is actually directed towards controlling devices that attach to the internet.  (Plug for article.)   The idea is that all devices will acknowledge and adhere to digital “flags” placed in content.  The flag itself is noncontroversial.  The problem comes when you force all devices to listen to the flag and ensure that flagged content can't be sent online.  So far, the FCC's own approval process has arguably been heavily controlled by the “content industry” — smothering in the cradle technologies for content protection that were viewed as insufficiently powerful.  The bill provides very few limits on the FCC's discretion to create rules and procedures — and the FCC's track record in this area so far is not a good one.

The digital audio broadcasting section is curious — earlier versions of the bill prohibited personal copying.  Now the draft vests complete discretion in the Commission to promulgate regulations about all kinds of digital transmissions.  There's a Digital Audio Review Board established that will advise the FCC and is supposed to work towards a proposed regulation.

When did the FCC become the internet lawmaker?  Why should the Commission be making rules about content online?  There seems to be no limit to the power granted to the Commission under this section:

[T]he Commision may promulgate regulations governing the distribution of audio content with respect to –

(1) digital radio broadcasts;

(2) satellite digital radio transmissions; and

(3) digital radios

What's not included in this description?  Anything and everything can be a “digital radio” – a bit is just a bit.   If the goal of this bill is to be “deregulatory,” it's hard to see how that goal is being served. Why create technical mandates in this area?

And, by the way, isn't this bill about telecommunications reform?  What's radio doing in here?

(Apparently the audio flag is essential to Sen. Frist.)

More tomorrow — plus auction news.

Let's take a look at the section of the draft Telecom Act of 2006 that starts on p.112.  It has to do with “phantom traffic.”

It requires that all “providers of voice communications services” (including any online application that's capable of connecting to the traditional phone system) preserve/present information — like the calling and called parties.  This is supposed to allow “traffic identification” by other providers that transport the traffic.

The traditional telephone incumbents are hoping to be able to charge for all voice traffic, don't want to permit encrypted traffic that they can't inspect, and want to be able to perfectly discriminate between different voice services.

What's odd about this approach?  Well, a bit is just a bit.  A voice bit ordinarily wouldn't look any different from another kind of bit.  There's nothing illegal (now) about encrypting traffic, and forcing voice bits to self-identify is a major change.  The original design of the internet was based on “layer independence” – the idea was that the transport layer wouldn't discriminate with respect to what it carried.  Reinstantiating the complicated system of telephone payments online certainly helps the incumbent telephone companies — but it's not clear who else it helps.

Traffic without a label isn't “phantom traffic.”  It's just traffic.

Then, on p. 122, we get to the video services section.  We've been told by the incumbent telephone companies that this is what they really want.  They want to avoid having to go town-to-town to get franchises to provide broadband video services.  They're seeking to standardize and simplify the process. 

The key question is whether the telcos should get this treatment without having to give up on something else.  Why should they have access to public rights-of-way without having to act like common carriers for these broadband services?  How much subsidization is appropriate?  Why wouldn't we want to wait for real competition to emerge in the broadband access market before giving this central concession to the telcos?  If everything becomes cable-like, we'll end up with a private communications world — is that what the country wants?

Tomorrow we'll continue — this time with the Video and Audio Flag subsection.

Recess has begun, so all the Senators can go off on little (or big) trips together.  They'll have with them a new draft Telecom Act of 2006 (enormous pdf).

To all the Senators who are carrying the draft (and the greater number of Congresspeople who just got the ad about the draft) — I just want to let you know that you can take the first 71 pages out of your notebooks.  That's a repeat of the House bill, which is replaced by the amended Stevens language.  Only 200 pages left!

Now you're ready to plunge into the Universal Service Fund section. (You may want to do some preparatory reading – Thomas Hazlett, What Does $7 Billion Buy?) 

The key take-away from the USF section:  it vests unlimited discretion in the Commission and the States to assess USF contributions based on any mechanism they choose (phone numbers, revenue, or something else, including assessments based on international portions of service).  Any entity providing broadband service to the public is covered — as is any online voice application that is CAPABLE OF connecting to the public telephone system, even if it doesn't. 

To keep the numbers from getting too visible, the bill says that the money flow associated with USF will be kept out of the USG budget.

Other provisions in the bill require voice services that are capable of connecting to the traditional telephone network to carry other people's traffic and comply with whatever disabilities-related regulations the Commission comes up with.  This is, of course, in addition to the E911 and CALEA obligations the Commission already said it would impose on VoIP services.  (Plug for recent academic paper on E911/CALEA.) 

So here we have an admittedly broken, bloated, graft-ridden funding system that supports traditional telephone services (largely) and won't help US broadband penetration statistics.  The money to support this system is going to come from free or lowcost or commodity applications of various kinds that happen to use IP and are capable of connecting to those old-fashioned telephone services.  That will lessen the economic viability of these new services, while failing to support the provision of broadband in the US (something we all thought was a national goal).  And all of this will happen under the radar — in impenetrable FCC proceedings addressed to services the traditional telcos can't stand – without facing up to our national, general need to fund internet access.

It's true that the Bill sets up a new separate program called the “Broadband for Unserved Areas Program,” to be administered by the USF, but that's for a small portion of the money collected for USF.  The vast bulk of the funding is going to go to propping up the existing broken system.

That brings us to p. 111.  On p. 112 there's a section called “network traffic identification accountability standards” that cries out for attention — but I'll let that wait for tomorrow.  You're on vacation, after all.

[Meanwhile, the D.C. Circuit has asked the FCC and the DOJ separately to file 15-page briefs responding to the recently-filed petition for reconsideration as to the court's affirmance of FCC's interpretation of CALEA.  This is (limited, be cautious here) good news -- it means that the court is taking the petition seriously.]