Digital Karate

A while ago, a friend of mine mentioned that he was checking the list of donors to the Yes on Prop. 8 campaign for businesses. He would then leave a review on Yelp pointing this out. This really took my by surprise - I had never thought of local review sites as a channel for political activism.

Of course, I don’t actually think what my friend is doing will really advance gay rights. In fact, most likely it will just push businesses to remain neutral on political issues. It’s easy to make customers go somewhere else for what they need, but its hard to get them to consume more. By picking a side in a split issue, a bussiness will lose more newly informed consumers than it gains. That’s a simplification but I think it’s close enough.

I wonder if sites like Yelp will move to remove these kinds of reviews. While I don’t think regular users could be trained to flag these posts, I bet targeted businesses will be quick to.

>yone I know tuned in to analyze it. And perhaps to catch a major gaffe live. The company I keep? Perhaps - but how do you explain the news coverage immediatly jumping into post-game analysis where pundits weigh in on how good each candidate was at playing the debate game. News networks do it because that’s what people want to see, because people are viewing the debate from that angle. You don’t tune in to be swayed and convinced, you tune in to see how everyone else, the ‘average american’, might have been swayed and convinced.

My theory is that this is the only way these debates can be swallowed. They are completely vapid otherwise. Nothing new surfaces, no new posistions are expressed. We’re all lucky if the candidates answer the same question the moderator asks them, we’re lucky if they manage to respond to each other in any meaningful way. Their answered are dumbed down and rarely qualified. It would be intollerable for intelligent people to watch them without adding their own layer of analytic complexities.

It’s really a shame that we don’t demand a higher level of discourse.

So I went ahead and released the codebase I used in my undergrad research at UC Davis. I cleaned things up up and added some documentation. You can check it out here.

Here’s a tip I hope hits google, because I’ve seen this same question go unanswered in many places before I figured out my own solution.
If you’re trying to use JAXB in a situation where XML standards are used in a less than dogmatic fashion, you may run into problems with namespaces and namespace prefixes. There is no way tellSo egal wie beschäftigt mit Hauptchores oder Arbeit, casinos Anhänger nicht wurden erfahren sollen, scheisst der Streit des Reisens von einem Platz zu anderen gerade, zum ihrer Wahrscheinlichkeiten oder innen der Online Kasinopoker Aufstellungsorte zu nehmen.компютри JAXB to simply ignore them, you actually need to roll up your sleeves and write a SAX filter.

The general strategy is going to be to fulfill the ContentHandler interface with a new class which will intercept SAX events before JAXB can get them. Once you see the implementation it may seem sloppy, but the other methods I tried only ended in sorrow.

This code will make JAXB ignore all namespace prefixes and treat elements as if they are in the top level namespace. I’m sure you can think of other uses for this method.

Start by defining your own content handler:

The hack is getting JAXB to use it. You could just modify the library yourself. If you don’t want to do that, you can do the following. Define a XMLReader:

That reader will set the content handler. Now you just need to put it all together.

You can also create your own XMLEventReader and pass it to JAXB, but I encountered random NPE’s while parsing some XML. I think the JAXB error handling isn’t setup right when you pass in a XMLEventReader. There may well be some way to resolve that, probably making that the superior solution.

Try it here.

Pretty rough around the edges, but I’m constantly (albiet slowly) improving it.

And I got my 25 dollars back. The system really does work.

I could also title this post “How I used two YCombinator companies to raise money and awareness for some people who could use it.”

The first company was reddit. The first use of it was discoverying a story over at Reason which detailed a no-knock paramilitary raid on a seemingly innocent man which ended with him being shot 5 times (although he did survive). Without going into too many details in this post, I’ll just say that it was a pretty fucked up situation. The warrant was boiler plate. The force was uncalled for. The charges were on the verge of manufactured. The victim (Tracy) was interrogated while still very injured and medicated. The only non-police witness (a neighbor) was intimidated to silence, but confirms Tracy’s innocence in private with whomever asks him. It represented something seriously wrong with the local justice system and was another example of a much larger problem in America.

I noticed that in the article’s comments there was a post by someone claiming to be the victim’s sister. She gave out her mailing address in case someone wanted to help them out (how cute, right?). I emailed her to let her know she should setup a paypal account. After a short exchange of emails with her I realized she really could benefit from a website to support her cause. I didn’t want to run this website for her, but I was willing to set her up with a domain name and a WYSIWYG website host. I considered both Jottit and Weebly and decided Weebly was the better choice.

Weebly of course is the second YCombinator company I used. If you haven’theard of it, it is basically a super easy web site creation tool. Drag and drop. Not a lot of options. Perfect.

I bought the domain www.justicefortracy.com (the sister’s idea) and pointed it at the new Weebly site I had sketched out. I showed it to the sister and explained how to work it. I added a donate button and fielded some more questions from her and pretty soon we had a working site. The largest bit of confusion was that she was using Internet Explorer, which apparantly doesn’t play nice with Weebly’s website editor. I mistook her complaints to mean she was just having a hard time figuring out Weebly, not that her tools were holding her back. I think she made the same mistake. Once I set her up with Firefox she was uploading images and blogging away.

Once the site looked decent enough I used Reddit again, but this time I was submitting. The story hit number one in about an hour around noon, and stayed there until I went to bed. It generated 14,000 page views for that day (Friday). By 8 PM on saturday, the site reached another 16,000 page views. (side note: I wish Weebly told you were people were coming from.)

While I don’t know the exact donation totals, I bet the site has raised a thousand dollars in a couple days. I can assure you from my personal correspodance that this money was needed, and that his sister is who she claimed to be (and I am a skeptical person). More importantly, the story gained a lot of traction, raised awareness and triggered letters and comments of support for Tracy. Tracy is in a very shakey emotional state, and furthermore he feels no one believes him. Even some of his family refuse to believe the police could be so wrong, so you can immagine what some of his neighbors think. I can only hope this helps him carry on despite his severe injuries, criminal trial, and inability to work or drive.

I don’t think this kind of thing would be possible even 10 years ago, and it almost allows you a glimmer of hope for this society.

Dr Ben Goertzel was answering questions about his talk at the Singularity Summit. Someone came up to the microphone and asked whether he considered himself a nerd, transhumanist, humanist or technologist. Goetzel rescued an innane question with and awesome response, “I generally see myself as a monkey trying to get over the problem.”

This sums up the attitudes of a lot of people at the conference; their biological heritage is ugly and shackling, something to be overcome. It’s basically the opposite of 18th century romanticism. Interestingly the forfront of rational thought it starting to concede that human emotions have intelligence and rationality to them afterall. Personally I’m starting to concede the same thing, like it or not I am a monkey, and I should embrace it. 7 card stud strategiesfree online video pokerfree texas holdem poker play,free texas holdem poker,free online texas holdem pokerfree online multiplayer pokerplay texas holdem online free,online texas holdem,texas holdem online gameonline betting pokerpoker siteinternet poker,card game internet poker,bonus internet pokerpoker freefree internet pokercrazy game of pokerpoker gambling gameonline poker schoolfree poker softwareplay video pokeronline poker for fungambling pokerstrip poker gameonline poker bonusvideo poker bettingbest airline miles credit cardcard credit interest lowest ukcredit card debt consolidation loancard consolidation credit debt unsecuredaccount best buy card credit,best buy card com credit,best buy credit cardcard credit debt free helpbank card credit georgia monogramuk credit card companycard company credit uk? ?,credit card company,card company consolidation credit xxasdfinstant approval credit card,approval business card credit instant,approval card credit instant offercard credit mutual providian washingtonapplication canada card creditcredit card processing machine,card credit machine processingcard comparison consolidate credit debtchase credit card servicesmbna america credit cardsears credit card paymentcard credit debt negotiation settlement? ?,credit card debt negotiation,card credit debt negotiationcard credit credit free no online report,card credit credit free online report,free online credit cardsetup online credit card processinginstant online credit card applicationcredit card debt elimination,total credit card debt elimination,mbna credit card debt eliminationapplication canadian card credit visa,application card credit online visa,visa credit card applicationbank card credit,first premier bank platinum credit card,associate bank credit cardcard credit payment searsdirect merchant credit cardbank credit card application,application bank card credit orchard? ?,citi bank credit card application0 balance card credit transfercredit card debt relief,card credit debt debt relief stop,credit card debt relief nonprofitcard consolidation credit debt service

Nintendo’s sensor bar was a pretty neat idea, but it has it’s drawbacks. Your wii has to be somewhat near the screen and you can’t use the wiimote more than about 12-15 feet from the sensor bar. Both of these are a problem for me, as I have a moderately sized room with a projector setup. Also due to the gigantic nature of my screen, I wanted the effective sensor area of the wiimote to be closer to the size of the screen.

The natural solution was to build my own sensor bar (or IR sources, really). Judging from the interested in DIY and 3rd party sensor bars I’ve been seeing, I think other people want to be cool like me and make their own. This guide will show you how.

A note on wall power vs battery power: I’ll be focusing on a wired solution for this guide. Batteries seem like a poor choice to me as they will run out in about 24 hours. How many times will you accidentally leave it on overnight? Nonetheless, you can still use a lot of information below to help you on a battery based build.

1. Sensor Bar Basics

The wii sensor bar is nothing more than two points of infrared light. The wiimote uses these two points of light to decide where it is pointing on the screen. It is fairly strait forward to make a couple sources of IR light, but there are some considerations to take into account.

• Will it run off batteries or wall power
• How far will you be from the sensor bar

Those questions will determine what parts you need to buy to build your sensor bar (I’ll refer to the sensor bar as “IR sources” from now on, as when you will be making is neither a sensor nor a bar).
2. IR Source Positioning

There is a common misconception that if you space your IR sources farther apart, the effective sensor area of your wiimote expands. Much to my disappointment, this is false. The spacing of your wii IR sources has little or no effect on the sensor area. Empirically what does matter is how far away you are from the IR sources. This leads me to speculate that wiimote only cares about the center point between the IR sources, and does not attempt to decide how close it is to the screen based on the spacing of the IR sources. Furthermore, it seems like the wiimote uses the angle of this virtual center point relative to itself to to decide where it’s pointing at. This would explain why if you are farther back from the sensor bar, you will have a larger sensor area. In the graphic there are 2 wiimotes, A and B, pointed at the top IR source (the blue dot). While both wiimotes are pointed at the same location, their angle to the virtual center of the two IR sources is different. Wiimote A is cos(y/2x) while wiimote B is cos(y/x). This means that wiimote A must point farther away from the center IR sources than wiimote B to get the same angle from it, and thus wiimote A has a larger sensor area.

This means that if you double your distance from the IR sources, you should double the effective sensor area for your wiimote.

If you want to keep it simple, just plan on spacing your IR sensors one foot apart on some sort of rigid mounting. Don’t place them any closer together than Nintendo already does with the stock sensor bar, it seems to cause cursor jitter.

If you want to get fancy and have totally separate IR sources, keep in mind that you don’t want them too far apart. In general I found your wiimote must be a bit more than 3 times farther away from the IR sources as the IR sources are away from each other. This of course means people sitting at an angle to the bar can sit close than those who are directly facing them. The space doesn’t seem to affect the maximum range. I have heard you can’t be 12 times farther away from the sources than the sources are apart from each other, but this tested false for me.

3. IR Sources

We’ll be using LED’s for our IR source. The first thing to determine is how many LEDs you need, how bright they need to be, and how wide their cone of light should be. I went with 696-OED-EL-1L2 which are 100 milliamp LEDS with a 60 degree cone of light. I used Octopart to find the best price on these, which was 27 cents each on Mouser. They might have a minimum order, I’m not sure. I spend 20 bucks on parts and it was about 5 dollars to ship.

One of the reasons the IR LED I used is so great is its angle of emission. At 60 degrees, these LEDs have one of the best emissions angles I have seen. I’ve heard that people who bought LEDs with narrow emission angles had trouble unless they sat directly in front of their IR sources. This is why I didn’t just get LEDs at radio shack - they don’t publish these kind of characteristics. In actually testing, I got more than 60 degrees; I actually measured the effective angle of these LEDs at about 90 degrees (45 degrees in each direction from the way the IR sources face).

As far as intensity goes, at first I thought I was going to need at least 5 LED’s on each IR source, like Nintendo does. It turns out that even 1 LED on each side was so bright it worked for the max distance I could easily test: 20 feet. The Nintendo bar craps out at around 15 feet, so this means that means these LED’s are significantly brighter than the multiple LED’s that Nintendo uses in their bars. I decided to go with 2 LED’s on each source, to be safe.

4. The Power Source and Circuit

You will most likely only need 1-3 LEDs. The easiest thing to do will be to power them in series with a spare AC adapter. The first step is to find an old AC adapter you can cannibalize. I found one that went to an old wireless router. It is important to read the label to confirm it outputs at least 200ma (higher is OK too). You also have to make sure the adapter provides enough voltage for the circuit you will be building. When you find the right one, cut off the tip and strip the ends. Use a multimeter to determine positive and negative.

The circuit I’ll describe will have the left and right IR source in parallel with each other, but the LEDs on each source will be in series. Because both IR sources will be in parallel, they will have the same voltage, but of course they will split current between them.

Each of these LEDs takes 1.5 volts to power, so to run N LEDs on each side, you need at least N*1.5 volts from your power supply. You’ll also need a resistor in series with the LEDs, but its resistance depends on the number of LEDs you use and the voltage of your power source.

The math is simple. V = IR so R = V/I and so for N LEDs on each side the resistance of your resistors must be: (SourceVoltage-LEDVoltage) / 0.1 or (SourceVoltage-(N*1.5)) * 10

I had an adapter rated at 4.0-5.5 volts, and 2 LEDs. In my calculations I assumed the maximum voltage output to guard against overpowering the LEDs. This means I needed a resistance of (5.5-3.0) * 10 or 25 Ohms. I only had 47 ohm resistors, so I used two in parallel to create (effectively) a 23.5 Ohm resistor. In theory if the AC adapter put out maximum voltage it would fry my circuit, but in practice I was OK. Remember every resistor is rated to a certain wattage. The wattage going through a resistor is (voltage * current), or for the circuit described: (SourceVoltage-(N*1.5)) * 0.1. Note that resistors are commonly rated to 1/4 watt, but you can get higher wattage ones.

The circuit should look like:

I left it out for simplicity, but you should probably add a LED which emits in the visible spectrum so you know when this guy is on.

I suggest making your circuit on a breadboard first. I made mine on two so that I could test out the angles and distances before I built my first prototype. I used two different ones to play around with the spacing. Obviously most of the stuff in the picture is not related to the IR circuit. Note that IR LEDs are not visible to the human eye, but they are to cameras. If your cellphone can take pictures, use its view finder to see if the LEDs are powered. Remember LEDs can only be hooked up in one direction! The longer wire should be positive.

Once you have a good circuit, you can use some protoboard to build the left and right IR sources. Obviously after I took this picture I cut down the wires and soldered everything together in a somewhat clean fashion. This might not be the best protoboard size to pick, as there wasn’t any space on the board left to mount it to anything.

How you mount them is up to you. I used some cardboard to make some temporary stands. I’ll eventually put them into one discreet piece when I finish testing and find spacing I want.

You know things are getting good when you can use a website to organize your kids’ chores and pay their allowance. Talk about making peoples live easier. I can hardly imagine how this got done before the Internet.

Seriously, though, the landscape of ideas is quickly turning falling victim to a land grab. Most of it has been available all along, but only recently does the perceived cost of staking it out justify the perceived value of claiming it.

This is not a post complaining about how obvious or lame certain ideas are, its fairly well established that the cost is low enough to justify betting on a crazy idea. Paul Buchheit has a good post on how the seemingly bad ideas are generally the most promising for startups.

This post is about competition. Competition has always been a fact of business. In the traditional business models, competition helps get the best product out on top and only benefits the consumers. Automakers one up each other, competing for market share. Large Internet gateways offer increasingly powerful services while lowering their prices or simply making them free. costs or no cost at all. Local businesses keep their prices in check with other local offerings.

Furthermore, generally if there is too much competition, the strongest companies survive until there is only as much competition as the market can sustain.

Things don’t work this way for many community driven websites. Consider the rash of meet-up and even sites. There are at least tens of popular sites, and none of them is edging into a dominant position. If there was just one or two event sites, they would probably organically grow until they were immensely popular. Instead you see many many different sites with a small segment of the population. The problem is that none of the sites are getting the significant network effects they need to be truly useful, and the users themselves are getting jaded on the whole concept. Users won’t continue to hop from sinking ship to sinking ship until they find the best event site, they will simply abandon the concept.

If a website has the following characteristics, it is vulnerable to competition in the same way event sites are.

1) Dependant on network effects for value. If your website needs a large chunk of people to be useful, there will be a difficult hump to get over initially. Competition compacts this problem.

2) Low potential for revenue per a user. If a website is essentially paid for by advertisements, only a significant user base can make it profitable. This is similar to 1) except this is about the website being valuable to its owners, not its users.

3) Low barrier to entry. In traditional industries, weaker competitors go away over time. If starting a new website in your area is cheap, you may have an endless stream of competitors coming. Even if they quickly die off, they take users and even worse, make them jaded.

The third characteristic is particularly relevant to the new generation of web startups. If too many me-too’s come in, without a clear winner the whole pack can drown together. I don’t think I have seen this happen too much, but this is more of a prediction than an analysis. So far, things have worked out OK. Social networks have all 3 characteristics and yet the large ones are doing fabulous. I think partially this is because some managed to get a foothold before the floodgates opened.

You can imagine how things might have gone differently if 100 different versions of MySpace and Facebook launched at once. Every school or town would be on a different set of social networks. It’s quite possible the whole endeavor would have proved rather useless to both users and developers. With the space being tagged as lame by users, and unprofitable by developers, the whole pack might have drowned together.

There is such a thing as too much competition when it comes to web startups. Luckily, I don’t see many signs of “competition induced suffocation” yet. However, I would like the reserve the term for the 5-10 years ahead.