International Summit For Community Wireless Networks 2012.

21 October 2012

A little over two weeks ago Sitwon, Haxwithaxe and I made the trek to Barcelona, Spain for the International Summit For Community Wireless Networks, partially because we thought that we might get some useful things out of it for Project Byzantium, but also because Project Byzantium had been invited to attend and present some of our work and ideas for the community at large at the conference. So, arrangements were made in due course, and our journey took us from Baltimore to Philadelphia for a layover, and then an eight hour transatlantic flight carried us to Spain. Sitwon was traveling to Europe separately due to other arrangements having been made, so it was just Hax and I flying out of DC that day. TSA security was largely unremarkable for the outbound leg of our trip. BWI still has little in the way of wireless net.access so stick to using your cellphone or a cellular modem if you absolutely need to get online. Else, bring something to read to kill time.

Somehow, and I'm not entirely certain how (perhaps it might have been good karma paying off) Haxwithaxe and I were fortunate enough to have two empty seats in our row on the flight across the Atlantic so we had some room to spread out once the plane leveled off at cruising altitude. In hindsight, I'm strongly regretting not trying harder to get some sleep on the way to Spain. I tried elfnapping a few times but didn't manage to shut myself down well enough to recuperate. By the time I finally got any amount of sleep worth using the word I'd been awake for nearly 48 hours and feeling every second of it in my bones and wiring. On the other hand, I got a decent amount of reading and coding done on the way there, something I hadn't had time to do in the days leading up to the trip.

Passing through customs in Spain was trivial, much to my surprise. Clearing customs consisted of showing our passports and saying why we were there. "Academic conference" is considered a perfectly valid reason for travel so we were sent on our way in due course. We happened to meet someone whose papers I'd been reading in Philadelphia (and who knows more about core wars than I probably ever will). We picked up our luggage (well, I picked up my luggage, Haxwithaxe carried everything he needed with him, from computers to clothing) and immediately ran into a spot of trouble. None of the ATMs in the airline terminal were working and Hax ran into trouble with his bank almost immediately. For what it's worth, I was having problems getting the international SIM card I'd bought for my unlocked phone (a Sony Ericsson Xperia Ray) to register with the cellular net in Spain.

Long story short, read the instructions for both the SIM card and phone all the way through before messing with the hardware. The phone will be in roaming mode the whole time, so if you've developed an allergy to that little 'R' icon you'll have to deal with it. You'll also have to be careful with your data plan, so that you don't overcharge yourself into oblivion clearing your inbox. After hailing a cab and traveling to the hostel we'd made reservations at, we dropped off our luggage in storage because we'd arrived too early in Spain (local time) to check into our room. We hopped another cab and rode to the college that the summit was being held at. As we checked in at the front desk we also dropped off some HacDC and Project Byzantium stickers (as one does) and then got onto the campus wireless network to catch up on things. The summit opened for us with lightning talks by representatives of some of the projects represented at the conference. Sitwon spoke for Byzantium.

Things are a bit of a blur after that.

At some point that afternoon I headed downstairs to the summit's hacklab to do... something... and not much later faceplanted on my laptop and slept for a few hours. I probably looked rather strange, facedown on the keyboard and probably snoring like a buzzsaw. Some hours later I woke up with a wicked case of qwertyitis but was roused to sufficient consciousness to attend the keynote address of opening day, which was about wireless policy in the European Union. The presenter was Amelia Andersdotter of the Swedish branch of the Pirate Party and member of European Parliament. I'm sorry to say that I didn't find it terribly useful or helpful. Aside from the fact that we're not active in the European poltical community, Byzantium isn't working in the space of commercial wireless, either in cellular or data. We are a USian project and not European, so none of the policy discussed in the presentation applies to anything we're working on. I hope that the other projects got something more helpful out of the presentation.

After that day of the summit was over we opted to not go out to dinner with the rest of the attendees or visit the Guifi.lab that evening to hang out. Instead, Haxwithaxe, Sitwon, and I headed back to the hostel, went for pizza down the block after passing an unusually high concentration of sushi restaurants, took showers, and went to bed to try to catch up on sleep.

Now that I've had a few weeks to reflect on our time in Barcelona, Spain I think I can put together a reasonably coherent picture. For starters, if you hated living in the dorms while in college, don't stay in a hostel. While they are less expensive than hotels they are aimed more at teens and college students going walkabout in Europe. They generally seem to have less in the way of amenaties than hotels do (you'll have to supply your own everything, including towels and soap) and offer shared living space for limited periods of time - four to eight people stay in the same room in bunk beds. Each room has personal lockers, bathrooms on the tiny side, a common area in the lobby and a cafeteria. In case you find yourself staying in a hostel bear in mind that the water valve in the showers can be intermittant (to conserve water) so make sure it's at the right temperature for long enough before getting in. I assure you, your suitemates will not be pleased by your early morning wakeup call. For everyone out there who might think that staying in a youth hostel sounds kind of dodgy, I'd like to assure you that all of us felt safe staying there. We didn't have a whole lot of contact with many people staying there, but we never once got the feeling that we were in any danger, nor did we feel that the hostel was unclean or poorly cared for. Quite the opposite, in fact. If you'd like more information on hosteling, I'd suggest getting in touch with Hosteling International, which is a global organization of youth hostels in over 90 countries. They will be more than happy to help you find a hostel wherever you might be going.

Remarkably, Barcelona is a city which is favorable to unpowered travel. Many of the major roads we traveled had nine lanes - three one direction, three the other direction, and three down the middle set up expressly for people walking, riding bicycles, skateboarding, and roller blading. I was surprised to note a significant number of people of many ages ostensibly commuting to and from work (judging by their clothing) on skateboards, rollerblades, and bicycles. Automobile drivers did not give the people not driving a hard time, they waited patiently and moved on when it was time, very much unlike here in the United States. The people of Barcelona appeared much more fit and active than many of us in the United States, which runs counter to what one would think of people who seem to graze throughout the day, as opposed to planning to eat set meals at particular times of day. Paradoxically, what was doubly surprising was the number of people who smoked cigarettes while out and about. In fact, I saw many more people in Spain smoking than in the US these days. Aso, just as we have Zipcar service in the DC metroplex, it is similarly possible in Barcelona to rent bicycles, motor scooters, and even motorcycles for the day. While walking around the city racks of all three vehicles for rent were a common sight as were said vehicles driving around at certain times of day.

The Summit started remarkably in the day; we were able to get up at 0800 local time, arrive at the college at 0930 local time, and the conference started at 1000 local time. There was plenty of time through the day for coffee and snacks; tasty, strong coffee of a sort that even I can't drink a lot of. On Friday Haxwithaxe, Sitwon,and I caught Febrizio Sestini's keynote on collective awareness platforms for organizing people and activities in large areas (such as entire countries). Some of the ideas he put forth are remarkably similar to some of the things the Zero State is working on right now, and there is a paper that I'd like too get into Fabrizio's hands that he might find interesting.

The Approaches to Starting Mesh Networks panel (I think... I've got some storage errors around that date and time) I found a little disheartening. There are community wireless data networks in Canada and the European Union that have been around for literally years on end, with hundreds or thousands of nodes covering probably dozens of cities between them, and with thousands of users. Some of those projects have arrangements with existing regional telecommunications providers to provide uplinks to the rest of the global telecommunications network for their networks. Due to how strange and inconsistent US laws are, USian community networks keep getting shut down. In fact, sixteen states have specific laws outright preventing their construction, and some telecom providers like to force community networks to shut down (like Verizon did in Pennsylvania). On the other hand, Matt Rantanen's work setting up wireless for soverign tribal lands in San Diego county was truly inspirational. The list of things the infrastructure out there is built to withstand is amazing, covering everything from brush fires to the desert wind and dust.

The workshop on qMp, a variant of the OpenWRT firmware with implementations of some of the things that we built into Byzantium Linux was very interesting. The Quick Mesh Project firmware is designed to run on dedicated routers rather than general purpose equipment but it is designed to automate as much of the configuration and construction process as possible, from network topology to IP addressing to routing. qMp also takes the unusual step of using IPv6 for network addressing but tunnels IPv4 traffic over IPv6 because most wireless devices and not a few personal systems and applications are incapable of using the IPv6 protocol even after all these years. We got some excellent ideas from the qMp project and we're rethinking some bits of Byzantium Linux that we were uncertain of in our last release. We haven't started implementing any of it yet because a good implementation only follows from a good design, and we're still in the design phase.

I don't remember much about the OLSR presentation that afternoon because I'd started running flat by that hour of the afternoon, and no amount of coffee I drank was enough to keep my brain online. I'm sorry to say that I dozed through most of it, something I haven't done since high school, and I'm very embarassed that I don't remember if I went to any of the late afternoon presentations. I probably didn't because Project Byzantium had a presentation in the early evening that started late and wound up being cut off by the catered dinner in the lobby. We were invited to attend and speak at the Summit because one of our project goals is interoperability with as many wireless networking projects as possible. Byzantium Linux is uniquely situated to fill in gaps in the infrastructure because a laptop costs less than an industrial router, which may be overkill in some situations (such as covering the last three houses on a block, or running a service that would be useful on the local mesh). We talked about some of the problems we've identified and potential solutions to those problems. Among the things we discussed with the attendees were things like ways to detect the presence of another wireless network based upon its BSSID or the ESSID a particular project uses by default. We didn't seem to make a whole lot of progress during the presentation but Sitwon did on Sunday afternoon during the unconference track while talking to some of the leads of other projects at the conference. I hope they post their notes soon.

I should also mention that there was a small Commotion mesh running at the summit. A few nodes were scattered around the space to set up infrastructure, and there were a couple of applications running on it that we found useful during our time there. For instance, there was a Tidepools instance on which places of interest to attendees were mapped. There was also an Etherpad-Lite instance running for everyone to take notes during sessions and presentations, which hopefully will be published someplace soon. The IRC network didn't seem to get a lot of use, unfortunately, but that was probably due to the fact that we could just lean over and talk to one another.

On Friday night after the summit was over, there was a tapas dinner reception held downstairs from where the conference was held. I'm not sure what to say about it other than there was a lot of standing around and talking with attendees that we might not otherwise have had time to see during the day. A lot of catching up on developments with various projects was done, alongside a considerable amount of brainstorming solutions to problems.

Saturday morning's keynote was the epic fail talk - everything that community wireless projects did wrong, so that everyone there would not repeat the same mistakes. This turned into a roundtable as everyone in the audience shared the things that went wrong for them. For example, one develper accidentally reflashed his neighbor's router over the wireless network instead of the one sitting on the desk next to him. Another project talked about a colony of africanized bees setting up a nest in an antenna radome. Matt Ratanen talked about brush fires in San Diego county, California taking out his project's buildings that had backup power systems set up in them. Apparently, his deep discharge batteries burn, melt, and then leave an impressively toxic mess that fortunately can be scooped up in a single blob with a Bobcat and taken away for hazmat disposal. Taking the time to pour concrete slabs for the buildings was a good idea, it seems. There was also a story about Apple snails the size of softballs getting into the wiring of another project. Let it never be said that wireless hackers have boring lives...

That afternoon we attended the presentation by the Serval Project in Australia. Their project is peer to peer emergency communication software designed for Android devices which is, like Byzantium and qMp, self organizing and rapidly deployable, and includes applications such as text messaging, image and movie sharing among all nodes in the Serval network. To work atop as many platforms as possible (eventually) the wrote their own network stack from the ground up so that it would do exactly what they needed the way they need it. These are, incidentally, some of the things we want to build into Byzantium for community use but the Serval Project is on the client side, while we're on the server side of the network. I attended a presentation on bufferbloat by Jim Gettys of Alcatel-Lucent. In short, the current infrastructure of the Internet was never designed for the kind of traffic patterns we're running through it, and he gave an interesting view of what's happening out there right now and why. It was a good explanation of why relatively little traffic can disrupt a network at the most annoying times even when nothing else is happening. He also presented several possible solutions to the bufferbloat problem; hopefully one of them will be implemented soon because it's driving a lot of us crazy.

The festivities on Saturday night were not what any of us thought they would be, I don't think. The summit's organizers had reserved a room at Club Eclipse for dinner and drinks that evening. It's located on the top floor of the Hotel W, a four star hotel on the beachfront of Barcelona which can be seen for miles around due to its impressive construction and prime location. Sitwon, Haxwithaxe, and I took the metro and then hiked rather a long way to get there after dropping out stuff off at the hostel that evening. It reminded me of both of the New York and DC subway systems, which isn't actually a bad thing; if anything, I found it comforting. As we walking up to the hotel we began to suspect that we were very underdressed compared to everyone else we saw walking around. The Hotel W, made of chiseled stone and decorated with chrome and multicolored lights, looks like it should be a set in a James Bond movie. The lighting there was most impressive at night, and it looks as if somebody had a little too much fun with an Arduino when designing it. Club Eclipse could only be accessed by an elevator to the top floor which was guarded by a velvet rope, a hostess, and a bouncer, all of whom seemed nonplussed by we three representatives of the strange t-shirt brigade. Keeping our cool we talked our way past them and rode the express elevator to the top floor. That night we truly lived the cyberpunk life in Europe, if only for an evening. At the bar friends were made, shop was talked, and stories were told.

After departing Club Eclipse later that evening a group of us from the summit prowled the boardwalk for a couple of hours, talking and exploring the beachfront. Apparently, 3 Euros for a beer on the boardwalk is considered expensive, but not so much so that it's not worth it. I found the attitude toward alcohol in Europe a little jarring, having grown up in the United States. Most USians treat alcohol like a taboo, except in the cicumstances when it's not but the cognitive lines between the two are never drawn clearly. At 0200 in Spain picnics on the beach and a couple of beers on the sand are no big deal. You'd be hard pressed to find a beach in the States that allows even picnics on the beach after dark, let alone the open presence of alcohol. Here, we're conditioned to not be responsible but make every appearance that we are, which is reflected in many ways in the social world and constructs of the United States. Over there people seem to be conditioned to be responsible, and that's pretty much that. I was told by the folks we were hanging out with that they don't really have the problems with binge drinking as teens that we do in the States because drinking isn't a horrible-yet-open secret, it just is, the way that watching television is for us. Strangely, it cost us very little to take a cab back to the hostel that night; in hindsight we should just have cabbed it both and been done with it.

Sunday at the summit was the unconference track, which was deliberately unscheduled and set aside for breakouts and inspired talks that resulted from the other three days. That morning after talking with Paul from the Serval Project, we realized that each of us had some puzzle pieces that we needed to put together. We realized over coffee that Google seems to be going out of its way to keep ad-hoc support out of the main source tree of Android. In the bug tracker it's designated bug 82, and the pattern seems the same every time: Someone posts a comment to bug 82 asking why ad-hoc support isn't in Android. Someone working for Google (judging by their e-mail address and privileges in the bug tracker) who's never posted to that bug before (we've checked) says that it's a great idea, an obvious oversight, and asks for a patch. Even if a patch is sent back that someone at Google never, ever posts to that bug ticket again (we checked that, too). The implication and evidence we've seen suggests that someone at Google keeps spiking it but we don't know whom, why, or where in the Google power structure it's happening (or even if it's someone at Google). Paul and I did an improvised talk that Sunday on this topic, and afterward we set plans in motion to pool our information and efforts, and when we have something we'll tell everyone about it.

That afternoon Haxwithaxe and I had an interesting lunch at a nearby restaurant with some of the Freifunk, Serval, and OLSR developers, where we discussed everything from network simulators to the pros and cons of various mesh routing protocols and our stress tests thereof to tabletop RPGs. When lunch was over we returned to the building the unconference track was being held at. Sitwon was deep in discussion with people from other projects in a corner while Haxwithaxe and I spent the afternoon hacking. At the end of the day the three of us hiked back to the hostel to figure out what to do next. Somehow we got caught up in the crowd of a soccer game that afternoon. En route we'd noticed that all of the stores and many of the restaurants had closed for the afternoon, and then we saw more people on the street than I've ever seen anywhere, save perhaps in New York City on New Year's Eve. I can say pretty authoritatively, as a former resident of Pittsburgh, Pennsylvania, that the tailgate parties in Pittsburgh can go hide themselves. These people are trained professionals and Pittsburgh fans have a lot to learn.

Neither have I ever seen a home team crowd that posted spotters along the route the guest team's bus would take so that people could run alongside the bus and flag them off. There's a very strong national spirit there, evidenced by people wearing capes with the flags of Spain and Catalonia, scarves, and even regional and home team headscarves. We did the best we could to not seem like fans of the other team because we had no idea what might happen, and my high school Spanish (and inability to find a Catalan phrasebook) would probably not have been enough to get us out of any trouble. Additionally, Barcelona's finest lined the streets in riot gear, making a big deal out of being seen but not interacting very much with the fans. The amassed fans, for what it's worth, took notice of the riot cops and kept everything to a dull roar. We didn't see any violence but, to be honest, I didn't feel like sticking around to see for myself. Surprisingly, this was the only time we saw police in body armor, unlike at home where it's pretty common.

Sitwon will be staying in Europe for the next month or so for work, so we parted ways that night. Haxwithaxe and I got up very early the next morning so we could finish packing and get back to the airport. For what it's worth, security at the airport was thorough, polite, and extremely friendly with everyone. There was no "us-versus-them, and you're them" vibe there. Interestingly, their idea of security screening is far less invasive but much more thorough than any we've seen in the United states. Haxwithaxe's two laptops and laundry (compressed to roughly the density of neutronium) in his carry on luggage caused a deeper investigation than my pockets of random crap. In addition to searching his stuff they examined the scan further, applied filters to the images, and looked at the scans from different angles, things that I suspect the TSA may not be aware of. In the US, neither of those things would have happened. Again, there was no usable wireless so downloading a few last minute e-mails before our flight boarded wasn't possible.

Our flight back to the United States left the entire plane with frazzled nerves, thanks to the three and four year olds with ear infections reacting about as well as one would expect to the changes in air pressure. In fact, 7.5 hours of the nine hour flight featured their wailing and screaming just a few seats to our right. Several of the passengers sitting near us openly contemplated killing those children, and doubted that any court in the world would convict them for doing so. I'm somewhat surprised that federal air marshalls within earshot didn't make their presence known, but I also wouldn't be surprised if one of the upset passengers was an air marshall at wit's end. There was also other speculation as to the drugs their parents had to be on which allowed them to sleep through their children's discomfort, possibly thorazine or heroin. In other news, the people in the row immediately in front of Haxwithaxe and I put their seats back as far as they possibly could the first chance they got and kept them there until landing, which resulted in our typing awkwardly with our laptops balanced upon our stomachs. There was also the woman who smacked the top of my head for adjusting my seat when the numbness in my legs began to give way to throbbing pain.

After landing in the United States of America, things didn't improve appreciably. When we, the crowd of frazzled travelers got to Immigration Control, we were herded into a set of lines for people and had to prove our US citizenship, a task which took slightly over an hour for us, all told. There was inspection of our passports, verification of our tickets and itineraries, questions about why we'd flown to another country, what we did there, who we spoke with, what we did... there were more in depth questions at the US border than at EU Immigration Control. Conversely, the line for non-US citizens to present a passport (and visa, if required), get photographed, get fingerprinted, and go on their way was two minutes per person, tops. I timed it surreptitiously despite the very large signs and repeated announcements every four minutes that use of cellphones was expressly prohibited. We watched the ICE agents barely even look at the non-US passports, which were stamped and handed back with astonishing rapidity.

We then had to find and pick up our checked luggage, run to catch a shuttlebus to go from PHY terminal A (international) to PHY terminal F (CONUS flights, as well as the terminal our connecting flight would depart from), take our luggage through another baggage check, and then go through airport security all over again. That's right, without ever once leaving the "secured" area, we had to go through the security screening all over again. PHY doesn't have pornoscanners, just metal detectors, but they also have a much larger team of fully equipped TSA agents walking around and ostensibly doing their jobs. By this time some of our fellow travelers were understandably upset, frustrated, and concerned that they were going to miss their connecting flights given the delays encountered at Immigration Control. One of the TSA agents seemed to take offense at some of us grousing and told us, loudly and in no uncertain terms, that all of us were going to miss our flights.

Spoiler alert: We did.

They deliberately took their grand old time passing each and every one of us through the security checkpoint, even going so far as to prevent a few people from going through just to hold the line up. One of the TSA agents refused to let me go through because he was standing there reading the stickers on Haxwithaxe's laptop. My cellphone was in the bin so I wasn't able to bring up the stopwatch app to time how long this mess took to clear, but it was probably another half hour to forty-five minutes. After making it through security we sprinted to gate F1, which was as far from the security checkpoint as could be in that direction (thus sayeth the airport map). Hax and I not only missed our connecting flight to Baltimore, but we also missed the next one immediately following which happened to be at gate F32, diametrically opposed to the former gate and just as far from the security chokepoint that held everyone up. The airport information desk told us that we had two hours to kill until our next flight and have a nice day, but we chanced to meet up with some of the folks from the New America Foundation (who had also been at the conference). We fled to the nearest bar for a cold beer to soothe jangled nerves.

The final leg of our trip home was almost anticlimactic. We picked up our checked luggage, caught a bus, and then caught a ride with someone to our respective homes. It seems like every time I go on travel in October, I'll come back and autumn will have arrived in full force, and so it was that we stood cold and wet in a bus shelter that evening. I took the next day off from work and have spent the last two weeks catching up on everything that happened while I was overseas. It was only this weekend that I had the time and energy to write everything down in a more or less coherent fashioon. Again, I apologize for the lack of detail in places. As I've written this article I've come to realize that my long term memory was more disrupted than I'd originally thought due to sleep deprivation, jetlag, and stress. If I remember anything else I'll edit this article to add to it.