• P does peer-to-peer mesh networking in JavaScript.
• TowTruck allows you to add collaboration features (collaborative text editing, text chat, voice chat) to websites.
• PeerCDN forms a network from a site’s visitors, and uses it to offload serving up static content away from the web server and on to the networked peers.
• The Tor Project is interested in using WebRTC to enable volunteers with JavaScript-enabled web browsers to become on-ramps onto the Tor network for users under censorship, as part of the Flash Proxies project. The idea is that censoring organizations may block the public Tor relays directly, but they can’t easily block every random web browser who might route traffic for those relays over WebRTC, especially if each web browser’s proxy is short-lived.

All of this activity means that we might finally be close to solving — amongst other important world problems — the scourge of xkcd.com/949:

xkcd: File Transfer, used under CC-BY-NC 2.5.

I wanted to experiment with WebRTC and understand its datachannels better, and I also felt like the existing code examples I’ve seen are unsatisfying in a specific way: it’s a peer-to-peer protocol, but the first thing you do (for example, on sites like conversat.io) is have everyone go to the same web server to find each other (this is called “signaling” in WebRTC) and share connection information.

If we’re going to have a peer-to-peer protocol, can’t we use it without all visiting the same centralized website first? Could we instead make a WebRTC app that just runs out of a file:/// path on your local disk, even if it means you have to manually tell the person you’re trying to talk to how to connect to you?

It turns out that we can: I’ve created a serverless-webrtc project on GitHub that decouples the “signaling server” exchange of connection information from the WebRTC code itself. To run the app:

• download Firefox Nightly.
• git clone git://github.com/cjb/serverless-webrtc.git
• load file:///path/to/serverless-webrtc/serverless-webrtc.html

You’ll be asked whether you want to create or join a channel, and then you’re prompted to manually send the first party’s “WebRTC offer” to the second party (for example, over an instant message chat) and then to do the same thing with the second party’s “WebRTC answer” reply back. Once you’ve done that, the app provides text chat and file transfer between peers, all without any web server. (A STUN server is still used to find out your external IP for NAT-busting.)

There are open issues that I’d be particularly happy to receive pull requests for:

#1: The code doesn’t work on Chrome yet. Chrome is behind Firefox as far as DataChannels are concerned — Chrome doesn’t yet have support for binary transfers, or for “reliable” (TCP, not UDP) channels (Firefox does). These are both important for file transfers.

#2: Large file transfers often fail, or even hang the browser, but small transfers seem to work every time. I’m not sure whose code is at fault yet.

#3: File transfers should have a progress bar.

Thanks for reading this far! Here’s to the shared promise of actually being able to use the Internet to directly share files with each other some time soon.

The Belair 6-12 is an interesting new medium format film camera from Lomography. Here’s a mini-review, mixed in with photos from a roll of Velvia 50 that I shot with the camera at La Jolla Shores.

The camera looks amazing on paper — auto-exposure medium format cameras with interchangeable lenses are usually far more expensive, and the closest thing to the 6×12 panoramic format I can think of is the 35mm Hasselblad XPan II.

It’s much harder to get good results from the Belair than the XPan, though. The problems I’ve seen, starting with the most severe:

• “Infinity focus” isn’t infinity sometimes. This could be a lens calibration problem, or the bellows not extending far enough, or the film not being held with enough tension to stay flat against the film plane. This seems to make the camera useless for landscapes, which is what I’d want to be shooting with a panoramic medium format camera.
• When you take a shot, the shutter lever is on the front board that the lens is attached to via the bellows, rather than on the same side of the bellows as the camera’s body. This means that your sturdy tripod is keeping the camera’s body still while your finger is moving the lens board around and making your image blurry. This is probably the largest design flaw in the camera, and I thought the lack of reliable infinity was already pretty terrible. So, you actually shoot by putting the camera on the tripod, putting your outstretched index finger along the base of the front lens board to support it, and then using your thumb to activate the shutter extremely gently.
• The viewfinder isn’t coupled to the lens; you focus blindly by setting the distance between you and the subject and relying on depth of field. This makes the focus problems much worse, because you can’t even tell whether the lens is failing to reach infinity.
• The camera seems prone to “fat rolls”, where the spool doesn’t stay tight. This sometimes results in light leaks, and is what makes me think that the infinity problem might be about the spool not being flat against the film plane.
• The autoexposure is inscrutable (you can’t see what shutter speed it’s chosen) and sometimes makes bad decisions.
• There are two plastic lenses included, a 58mm (21mm at 35mm equivalent on 6×12) ultra wide angle and 90mm (32mm at 35mm equivalent on 6×12) wide angle. The lens quality is not good. Each lens can be used at f/8 or f/16, which results that I’d describe like this:

You can see that the interchangeable lenses don’t add much if you’re interested in sharpness; you’ll want to stay at 90mm f/16 almost all of the time. (I should point out that Lomography is selling glass lenses for the camera as an upgrade — if you’re willing to spend more money on those, they’ll probably have better performance.)

With all that griping aside, how is it to shoot? It’s pretty fun. I like these shots, although I’ve had other rolls come back with unusably poor focus or exposure. In the shot below, I actually like that the two sides of the image (the seagulls and the cliff) are very soft; they make the photo look more painterly and surreal, which works here.

I can’t recommend buying the Belair, at least at its current price. I took advantage of a 40% preorder discount, and I think some of my problems with focusing might be caused by getting an early production camera that way. While I’m excited to see Lomography trying hard to innovate and keep film alive, the Belair seems to be more of a toy camera than a serious one.

Over a million children in Peru have access to an offline Spanish Wikipedia snapshot on their OLPC laptop. The Wikimedia Foundation is e-mailing its supporters a link to a trailer of a documentary called Web that shows the effects of these laptops with Wikipedia on children in the remote Amazonas town of Palestina, Peru. I was involved in creating the Wikipedia snapshot, so it’s very rewarding to see the video.

I especially love that the film shows the children editing Wikipedia as well as browsing it, so that we’re involving new parts of the world in the Internet’s global conversation instead of merely giving our own knowledge to them.

Four of us (three OLPC volunteers and I) worked on this offline Wikipedia snapshot for less than a month in 2008, through ten releases and 190 Git commits, and then shipped it to Peru. It wasn’t something anyone asked us to work on — it just seemed like a good idea, and it remains one of the most important things I’ve worked on in my life. It’s a reminder to always be looking and ready for unexpected opportunities to make a large difference.

Takeaways from the talk for me:

• JavaScript will be getting many new features that make it more attractive to write in — e.g. block scoping, weakmaps, sets, rest and spread parameters, default parameters, fat arrow syntax, and many other uses of syntactic sugar that I recognize from CoffeeScript — but it’ll take a few years before we can use them in client-side code reliably.
• But if you’re writing for Node, you can start using them now with node --harmony.
• Traceur can compile ES6 to ES3.
• Continuum is a full ES6 VM written in ES3.

Loiturerei village, Kenya. Taken by UK DFID, CC-BY-SA.

Today is Giving Tuesday.  It’s a great idea. Here in the US, something feels odd about following our national day of giving thanks (Thanksgiving) with the consumerism of Black Friday, Small Business Saturday and Cyber Monday. As we shop to find gifts for those we love, we feel it’s also important to celebrate giving to those we don’t know, who need it most. We hope this post inspires others to give more and to celebrate giving.

For several years now we’ve celebrated our wedding anniversary by giving a percentage of our yearly pre-tax income to charity — a percentage determined by the number of years we’ve been married. This year that percentage is 7%. Our 7th anniversary was October 29th, but we’ve waited to hear from our favorite source for charity advice, GiveWell, to make their yearly recommendations. Luckily they did this yesterday, giving us the opportunity to post this today.

This year we are closely following GiveWell’s advice and giving 90% of the 7% to three charities: GiveDirectly, Against Malaria Foundation (AMF), and the Schistosomiasis Control Initiative (SCI). (The remaining 10% will be decided later, and will probably be advocacy and other nonprofits that may not be highly effective, but are close to our hearts.)

50% to GiveDirectly (3.5% of our annual income)

GiveDirectly is GiveWell’s only new recommendation this year, and we think it’s one of the most interesting charities out there. Its method is simply this: find the poorest people in Kenya (here’s how they do that) and give them money through the M-PESA money network.

There are all kinds of reasons why simply giving money to poor people directly might not be the best we can do (they might spend it on something we’d rather they didn’t, for example) but it does avoid the money’s impact being diluted by corruption or overhead.  More importantly, GiveDirectly will be quantifying how much it helps. They will follow up with the recipients over the next year — using a randomized control trial for which they’ve pre-published the survey and analysis plan.

We’re hopeful that better interventions exist than GiveDirectly.  But we want their project to succeed because it shares the commitment to measuring outcomes that we think is vital, and it can serve as a baseline to compare other charities to in the future (i.e. “Can you do something that creates more improvement to lives than GiveDirectly? Prove it.”).

30% to Against Malaria Foundation (2.1% of our annual income)

AMF distributes insecticide-treated nets for protecting against malaria infection. GiveWell estimates the cost per life saved is just under $2,500. Malaria is not usually fatal, so there is also a fair amount of disability due to illness is also being prevented.

10% to Schistosomiasis Control Initiative (0.7% of our annual income)

GiveWell thinks that this charity — which concentrates on the “Neglected Tropical Diseases” (usually worms/parasites) — offers an extremely effective intervention at improving DALYs (see below). This is because the infections they focus on are readily treatable using very inexpensive drugs, yet often come with debilitating symptoms that don’t quite kill the “host”.

For You! Taken by Nomadic Lass, CC-BY-SA.

Donating effectively

It’s hard to list all the reasons people choose to give, or do not. One issue we’ve seen raised is the belief that “charity doesn’t work”. We believe that simply isn’t true. It may be true for some — many — perhaps most! Government-managed foreign aid especially so: it’s only around 1% of the US budget and mainly goes to political allies. But there are non-governmental charities that demonstrate real improvements, and GiveWell supports these. Giving can work, but it’s important to find effective giving opportunities.

And for that reason, we waited for GiveWell’s latest recommendations. GiveWell looks for organizations that maximize the improvement to lives caused by each dollar you’re giving.  This seems like it should be uncontroversial, but it’s not yet common to think about giving this way.  Perhaps one reason for this is that it requires a way to measure outcomes and compare them against each other, and that’s very difficult.  GiveWell is doing a fantastic job trying to do this all the same, though, using tools like the Disability-Adjusted Life Year (which is a measure of health that’s better than just measuring how long people live), randomized control trials, and the kind of statistics knowledge you have when you’re a charity review organization that was founded by a bunch of ex-quants.  (A Businessweek article referred to GiveWell as Hedge Fund Analytics for Nonprofits.)

A second reason people are sometimes reluctant to think about donating effectively in this way is that for most of us, it’s going to involve donating to people far away instead of in our local communities.  The price of living here in Boston, MA is very high, both for rent and food — in contrast, more than a third of the people in the world live on less than USD $2/day (most people don’t realize that this number is adjusted for the purchasing power of goods and services in the US!).  When trying to decide whether to donate locally or globally, it’s clear that our money can do much more good in other countries than here in the US.

A third reason that people are reluctant to give to maximize outcomes is that we don’t have the same emotional connection to people across the world as we do to an individual call from help from someone that we can see — counter-intuitively, studies such as this one show that people have a strong bias towards giving more money to help a single identifiable victim than to help many “statistical” victims.  The Internet has helped to reduce the effects of this emotional bias, with sites like Kiva giving a name and face to the global poor. Perhaps GiveDirectly could benefit from adopting a Kiva-style interface itself.

Closing thoughts

Each year we ratchet up the amount we give, and this year has brought us a new financial development: our first child. When people learn about our annual tradition they wonder how it will scale — will we be doing this on our 20th? Our 50th? Our 101st? (We hope to have that last problem!) As Yogi Berra said, “It’s tough to make predictions, especially about the future.” We know the responsibilities of parenthood will demand more of our finances, and balancing that with wanting to help others will be a lifetime project. Tithing (10%) is a very common tradition, and we want to at least reach that. Maybe we can go beyond it. For now we’ll take it one step at a time, and try to give a little more each year.

The team’s far larger than just OLPC staff — we’ve been fortunate to work on the project with Maryanne Wolf and her team at the Tufts University Center for Reading and Language Research, Cynthia Breazeal and team at the MIT Media Lab, and Sugata Mitra at Newcastle University.

(There’s also a Technology Review article with slightly more information, and an OLPC SF conference talk video that goes more in-depth.)

One of the reasons I look forward to these trips is the chance to get to hang out with my Taiwanese coworker Gary Chiang, who’s both an extremely talented engineer and an amazing photographer.

Before we got to Shanghai, I tried to look online for places to go to take photos during the downtime on our trip, but I mostly just found pointers to the standard guide book places. Gary took me to a bunch of excellent places that weren’t in the guide book; I’m writing about them here so that maybe you can go too. (All photos are available under CC-BY-SA 3.0.)

1933 Shanghai Slaughterhouse (虹口区沙泾路10号)

This is an Art Deco concrete slaughterhouse for cows, designed by the British architects Balfours and built in 1933. It’s been restored and converted into a space for creative companies as well as cafes. Lots of people come here to take photographs; pre-wedding shoots, cosplayers, everything. It’s got a strong M. C. Escher feel, with narrow “skyway” walkways going between the different floors.

Gary has an interesting theory that architects do their best/riskiest work in foreign countries, where they’re mostly immune from criticism by their own country’s media; no-one cares if your wacky art deco concrete idea doesn’t work as long as it’s not at home.

More photos in this set.

Shanghai Film Park (松江区车墩镇北松公路4915号)

This is one of China’s largest active outdoor movie studios, with impressively open access — you can walk around previously-used sets as well as people on shoots. There are all kinds of “fake” buildings: a church, prison, boat, train station, an “old Shanghai” area, big western houses, and lots of very old cars.

In the final shot below there are extras eating lunch on a break from their shoot.

More photos in this set.

Tianzifang (田子坊 泰康路艺术街)

Tianzifang’s a preserved shikumen residential area, with high terraced houses and narrow alleyways, now used as a mix of housing and art/crafts/tourist shops. It’s very touristy in recent years, but still worth seeing.

More photos in this set.

Zhujiajiao (朱家角)

Zhujiajiao is an ancient (about 1700 years old) water town: a town built around the intersection of different rivers. It took me a while to figure out that some of the best things to see in Shanghai are small trips (40 minutes by car) away. Zhujiajiao is one of the smaller water towns, which has the advantage that it’s not packed with tourists.

More photos in this set.

50 Moganshan Road (莫干山路50号)

The center of contemporary Chinese art in Shanghai, it’s a warehouse complex converted into studio and gallery space. Lots of street art nearby, too.

More photos in this set.