Archive for January, 2007

I'm about to leave Los Angeles and return to New York, hoping to be back in time for breakfast. I can hear the planes outside the window of this hotel.

When I arrived at midnight after the trip out, the car rental place was in complete disarray.  The computers were down.  The people were down.  Nothing was working.  Someone behind the counter decided he wanted to help me, and so he went out into the cold night and proudly brought me this:

I drove away with an (unintentional) screech of wheels.  The next morning, I started my LA driving adventures:  manuvering through terrible traffic, finding freeways, getting lost.  This afternoon I convinced my niece to do the driving for me, which was a huge relief.  She got a big kick out of the car.

It's always sweet and sour to come back here for me.  I went to high school here, but I don't know much about the territory.  I do recognize everything about how it feels to be here, and I always notice the pitiless sun and the solitary palm trees — and also the sweetness of the air near the beach.  I notice the campiness and the ugliness of things.  It's a harsh place when you visit, and I think it gets easier once you've been here for a while.  I'm from here, but I don't feel a close connection to this place.

I visited the ICANN offices today in Marina del Rey, to meet the new Chief Operating Officer, Doug Brent, and to talk to several people.  ICANN is also “from” here, as a matter of history — its offices are in the same building as ISI.  Should ICANN always be here?  Does it make sense?  This city is more about media than technology or governance.

But, as I say, it's about time for me to leave, so no grand pronouncements about anything much.  I sadly have to return the car.

In addition to the white
spaces and 700
MHz issues I've mentioned recently, there's another spectrum
matter that I keep bumping into and wondering about:  M2Z.

This
is a bid to offer wireless broadband to all, if the venture can get the
spectrum for free:

The company in a filing
with FCC says if it was to get the soon to be vacated 2155 MHz
to 2175 MHz spectrum for free for 15 years, it would offer universal
broadband access nationwide, and in exchange it would hand over 5% of
its gross revenues from its premium broadband service. In addition, it
will give free 512 kilobits per second wireless access to one and all,
which will be supported by advertising.

I can't tell how
their petition
is being received at the FCC.  For political reasons, they're
promising an “automatic, default blocking of access to pornographic,
obscene, and/or indecent material” and a secondary network for public safety.  But they're also
promising a new nationwide wireless broadband system, and rumor has it
that M2Z has $400 million in venture capital to implement their plans.
M2Z stands for “move the cost of data transport to
zero.”

Daily Wireless says that the block of
spectrum that M2Z wanted (20MHz of bandwidth) has been taken out of an
FCC auction — because “The FCC has decided that “free” nationwide
broadband wireless (and
exclusive use) is not an idea whose time has come.”  I don't have confirmation of this — expert advice welcome.

Big day at USC's Annenberg Center today.  Very interesting people — many projects going on — very glad I went.

Can we cast aside cynicism for just a moment, and say that having a session about the preservation of the internet is a good idea?

I don't know what's “behind” the State Department suggestion that they hold a global internet freedom day. Maybe it's just a dig at Google for doing business in China (everything has something to do with Google these days).  But I'm glad that the topic of an open internet is something this Administration is willing to discuss.

Law enforcement, the content industry, and network providers aren't that interested in an open internet.  They all have strong reasons to prefer something much more constrained.  But [law enforcement] doesn't necessarily = [State Department].

We'll see, but for today I'm optimistic.  Maybe they'll want to hear about OneWebDay.

I've been on leave this past term — playing to my strengths, planning my own schedule — and now school is starting tomorrow.  So I don't have a blog post to offer.  I'm glad to be going back to the classroom.

See you on the other side of tomorrow.

About 18 months ago I wrote a little post called “Up With Up” — about the importance of uploading.

There are big clashes coming when users discover they can't upload predictably. 

1.  BitTorrent.  John Waclawsky pointed me to this Wikipedia chart listing ISPs that block BitTorrent. How do the ISPs do it?  And how do they decide between “good” and “bad” BitTorrent protocol uses?  James Enck and others point out that there are many studios and virtual world companies using it for distribution — a legit, legally Torrent-able Harry Potter trailer, anyone?  Huge headaches coming, and only getting worse with high definition files. (Top ten Torrent-ed files listed here.)  Here's a paper about symmetric broadband use in Japan — where a small percentage of users is accounting for most of the upload traffic.

To block it, you need to sniff out the Torrent seeds and headers.  Fascinating discussion here between admins trading tips on blocking a computer-literate Torrent user.

2. Not providing symmetrical upload.  Carriers assume that people will download, not upload.  Dave Burstein provided me with the following (paraphrasing, all mistakes are mine):  For DSL, there's also a lot of interference caused by the strong user-side transmitter.  So ADSL was originally designed for 6 Mbps down, 768K up. And, besides, telcos want to be able to hang onto enough bandwidth to sell videos. For the telcos, it would be simple (and incredibly inexpensive — pennies — to provide symmetric 768K up and down, but they want to upsell people to 3 Mbps down, 768K up. Same with cable — it would be easy to provide symmetrical upload, but DOCSIS 3.0 is designed for 1 Gbps down, 100 Mbps up because cable carriers don't make money on upload “services.”

In other countries where there's more competition, carriers tend to provide symmetric upload/download.

3.  The Venice Project. From the people who brought us Kazaa and Skype.  Here's a description:

“The Venice Project is a streaming video application, and so uses a relatively high amount of bandwidth per hour. One hour of viewing is 320MB downloaded and 105 Megabytes uploaded, which means that it will exhaust a 1 Gigabyte cap in 10 hours. Also, the application continues to run in the background after you close the main window.”

“For this reason, if you pay for your bandwidth usage per megabyte or have your usage capped by your ISP, you should be careful to always exit the Venice Project client completely when you are finished watching it.”

Here are some network operators mulling over the implications.

So — we've only just begun the upload battles.

[Heartfelt thanks to the members of Gordon Cook's list who sent along this information.]

Two different people can see entirely different landscapes and end up solving problems differently.

So I'm working through how different perspectives on just why an open internet is important.  My tentative conclusion:  if you talk about competing applications prompting innovation, you'll lose.  If the only economic and cultural justifications you have for the need for a layered approach to internet regulation (an approach that treats transport differently from applications) are (1) the explosive innovation that competition among applications would produce and (2) the appropriate mapping between the “actual” architecture of the internet and the regulatory approach to be taken to it, you'll lose. 

Why?  Because it's so easy for the carriers to use the layered approach against you.  It's so easy for them to complain that it is unfair for rich companies like Google to be riding on their pipes without paying.  They can transform the competition argument into one about fairness and equity.  They can also point out that unchanging network architecture, if frozen into place, may discriminate against their applications in the future. And finally (and perhaps most importantly) they can say “okay, we won't discriminate based on the type of application that seeks to take advantage of our pipes.” 

This last part is the heart of the problem — if network providers get to decide that they can charge all video packets, and that X new thingie is “really” a video packet in disguise, that gives them a tremendous amount of discretion.  Who's to say what new thing falls into what category?  And who's to say that blocking an entire category of stuff (say, P2P applications) is discrimination?  “They're all being treated alike,” the carriers can say.

So playing around with the importance of competing applications isn't necessarily an effective perspective.  Even the layers perspective as a whole may end up not being effective unless you can clearly separate out transport.  It's so easy for the layers perspective to be flipped into an argument about the enormous benefits of vertical integration.  “Innovation”-based arguments can so easily be transformed into “they want to free-ride” responses.

(By the way, watch out for people who want to say that “access” (the last mile) is separate from “transport” (the backbone), or that there's a “control plane” that is separate from “applications.”  All of these distinctions are dangerous.  That's another post, and another risk of an overly-religious adherence to a detailed layers-based argument.)

So we need another perspective.  Here's my suggestion:  Of course applications are important, but they're secondary to what they facilitate.  What they facilitate, for more and more people, is human relationships.  These groups/affiliations/whatever may be cross-application; they may be invisible; they may be hard to draw lines around.  But they're what's actually interesting and complex/evolving about the net.  Content-delivery isn't the game; humans are.

Once you use this new perspective, you gain a new appreciation for facilitating communications rather than the profits of infrastructure providers.  And it turns out that facilitating these communications will lead to new ideas and economic growth.  Rather than providing economic success just for the carriers, we'll be providing economic success for everyone.

Writing, writing — not much time left.

If you know someone who would like to operate a new unrestricted gTLD — an open one, not a “sponsored” one — could you let me know? 

I don't want to know what the string is (so please don't tell me).  I'm just trying to get a sense. 

This looks interesting — Scott Page, “The Difference: How the Power of Diversity Creates Better Groups, Firms, Schools, and Societies.” From the introduction:

If we sample widely, we're more likely to find the one person who can solve the problem or who can make the key breakthrough.  We did not get the theory of relativ[ity] from a crowd.  We got it from a diverse, novel thinker in a patent office.

I'll let you know how it turns out.

Sen. Sununu's announcement a couple of days ago that he was drafting legislation calculated to keep the FCC from creating audio or video flags was welcome.  Sen. Sununu has been deeply suspicious of the broadcast flag for some time.  Here's an EFF report from a year ago:

Sununu, an MIT grad, interrupted to ask the question so far unconsidered by his colleagues: Do we need this mandate at all?

He pointed out that “we have a whole history of similar technological innovation that has shown us that the market can respond with its own protection to the needs of the artists.” And he concluded with one of the most damning depictions of the ahistorical nature of the flag (clip from Congressional RealVideo) you'll hear on the Hill:

“The suggestion is that if we don't do this, it will stifle creativity. Well…we have now an unprecedented wave of creativity and product and content development…new business models, and new methodologies for distributing this content. The history of government mandates is that it always restricts innovation…why would we think that this one special time, we're going to impose a statutory government mandate on technology, and it will actually encourage innovation?”

Something like this argument shows up in an article I wrote a few years ago now called The Biology of the Broadcast Flag.  Where there is ample competition for new ways of dealing with bits (including many forms of private DRM), it makes no sense to impose a tech mandate – all that will happen is that the evolution of those new ways of doing things will be stifled.

The reason the net neutrality battle is going on in this county is that there isn't that kind of competition.  We have a system of regional duopolies that are tightly controlling access and have every incentive to reinstantiate their old, clearly-defined services online. So in this case government intervention is appropriate.  There's no reason why the single sidewalk available should be allowed to monetize the interesting conversations taking place above it.

Nice to see Sununu taking such a brave step — now we need to find a Senator who has a similar understanding of the internet access issues.

The DTV transition keeps creating more and more issues.  Among other things, the idea was that last year's DTV bill would clear 24 MHz of spectrum for interoperable public safety communications and allow the auctioning off of other spectrum — potentially bringing in billions of dollars. 

This spectrum is in the 700 MHz range.  So Congress is targeting funding to pay for emergency radios in that range.  And the FCC is looking for comments (PDF) on their planned “Implementation of a Nationwide, Broadband, Interoperable Public Safety Network in the 700MHz Band.”

But there are many interesting wrinkles here.  According to the Associated Press, Mayor Bloomberg told Congress on Monday that they shouldn't be targeting emergency radio funding at the 700 MHz range.  He's pointing out that NYC has been using 400 MHz (penetrates building and tunnels better than 700 MHz) and has invested a lot — and should be supported.

Here's a good LightReading background piece — note particularly the petition by Cyren Call (more details here) seeking to have only commercial wireless operators and public safety entities sharing the whole swath of spectrum that is now scheduled to be auctioned off.  It looks as if the wireless carriers want to get their hooks into this spectrum by arguing for a privatized public safety network.

Lots going on here.