pragmati.st - Home

reCAPTCHA Thinks Its Funny

2009-12-29T18:11:00Z

Over Memorial Day Weekend (yes, it has clearly been a while since I’ve posted) I was working feverishly to launch a new product. I was stressed out, I was tired, I pretty much worked every waking minute that weekend.

On Monday I was doing some final testing on the user registration flow and got this from reCAPTCHA:

Smart ass.

Rails Error: "no such file or directory - getcwd"

2009-12-29T17:32:00Z

If you see this then you probably deleted the directory that your server is running in using rm.

If you’re on your dev machine you should kill the server, cd to the proper directory, and restart the server.

Boxee, Hulu, and the Future of Video Advertising

2009-02-20T22:53:00Z

Boxee is a fork of the excellent XBMC media player that runs on Macs, AppleTVs and Linux. It provides a great remote control based interface for accessing your stored video, music, and photo libraries on your TV, as well as providing site specific browsers for viewing media from the internet sources like YouTube, Comedy Central, the BBC, CNN, Last.fm, Flickr, Picassa, and, at least for a few more days, Hulu. For a more in depth introduction check out this New York Times article.

Hulu recently announced that they were being asked by their content providers to cut off access to Boxee. So as of about 20 minutes ago, you can’t access Hulu via boxee. A lot of people, myself included, are really annoyed at this. I think that Marc Hedlund pretty much nailed it.

Basically, the studios are worried about anything that makes it easy for the general public to view video content on their TVs over the Internet. Never mind that this is what people want, or that there I can and will still watch Hulu on my TV (see Fred Wilson’s take). The studios are worried about the cannibalization of their television audience. Hulu has ads, but television ads are much more lucrative. The studios are going to protect that market, at all costs, for as long as they can.

I think that the studios are being short sighted and incredibly stupid, but I’m going to choose to focus on another question: Why are TV ads more profitable than video ads in a Hulu stream on a per impression basis?

Television Advertising

Television is fundamentally a mass media. Everyone watching a TV show gets the same ads (well, mostly, but not exactly as local affiliates and cable providers do insert certain local ads) But there are fundamental limitation on how well they know their audience. Simply put, there is no way to know who is watching the ad, so the best they can do is target the demographic average. Never are the limitations of this approach more clear than when you watch a show for which you are clearly not the normal viewer. Let’s call it the “Top Chef Problem”.

I watch Top Chef. There. I said it. I find the reality show portions pretty annoying, but I love food, love cooking, and damn if that show isn’t entertaining. However, watching the ads makes it really clear that I’m not their target audience—women with children are. So the vast majority of ads are a complete waste on me. I’m not buying beauty products or minivans. There is no way that I’m going to watch the Real Housewives of Orange County, so they might as well stop advertising it to me.

But, since with TV ads there is no way to know that I’m a 27 year old straight man who happens to love cooking and drinking, they keep showing me ads I have no interest in, and Dove Body Wash keeps paying for ads that have no chance of turning me into a consumer.

Internet Video Advertising

What about Internet Video Ads. Unfortunately, the state of the art seems in Internet Video Ads seems to be roughly equivalent to that of TV advertising. Actually, it could be even a little worse.

Last night, I watched a few streaming episodes of South Park last (from Comedy Central via Boxee). I saw the same ad for Jack In The Box over and over. I live in Maryland. The closest Jack In The Box is in North Carolina. Jack In The Box paid for an ad that has absolutely no chance of converting me to a customer.

This should never happen. Advertising on an Internet Video Stream isn’t a mass media. Its a targeted media. They’re streaming the show to me. And they know a ton about me. Lets look at the bare minimum that they know:

My location. Geolocation isn’t that tough. Looking up my IP gives my location within 2 miles. No more Jack In The Box ads.
My ISP. At my present location, Comcast. Why then does Hulu regularly show me ads for Comcast High Speed Internet when I watch Hulu over Comcast internet.
They know I use a Mac. Probably not a ton of use in showing me ads for the Zune. Sorry, Microsoft.
What I watch. Based on my cookie they know what other shows I watch on Hulu.

The Future

This isn’t even taking into account the truly scary amount of data that Double Click / Google / Facebook have accumulated about me on my travels across the web.

Lets go back to the “Top Chef Problem”. Imagine what they could do with all that information. No more ads for minivans. No more ads for makeup. No more ads for easy to prepare meals for kids.

Instead, they show me ads for local restaurants, wine and beer.

Now, lets put all the stuff DoubleClick knows about me to work. They know I’m a Software Engineer, so they show me job listings. They know I’m hip hop fan, so they advertise an upcoming Mos Def concert. They know I play soccer, so they show me a Nike soccer ad. They know I read aveceric.com religiously, so show me an ad for On The Line

Ads this targeted have to be worth much more, per view, to advertisers. So it would reason that advertisers would pay more per view for them. So even though internet video will need to include fewer ads than traditional television (to effectively compete with bittorrent / other sites that don’t have any ads) the studios could still make just as much money, or more, per view than by selling more expensive ads.

Wouldn’t it be more productive for media companies to spend their time and money investing in advertising technology then engaging in a Sisyphian struggle to protect the staus quo?

The future of video entertainment is on demand. The media companies aren’t going to be able to continue to extract rents (meant in the economic sense) from controlling distribution. In a digital world, once you release something people are going to be able to get it on the device they want. The only question is how the media companies figure out a way to make money off of that.

Teachers and Quarterbacks

2008-12-18T05:49:00Z

- Quarterbacks by Ableman / Teacher by Editor B

Most Likely to Succeed, the most recent Malcom Gladwell piece in The New Yorker, is a really interesting look at the similarities between trying to draft an NFL quarterback and trying to hire a good public school teacher. The main similarity being that we are incredibly bad at predicting success in either field.

For all the effort on scouting, combines, and testing of college quarterbacks it turns out that there is no demonstrated correlation between wonderlic test scores or draft position with an NFL quarterback’s career performance. Similarly, teachers who earn a Masters Degree or an advanced teaching certificate don’t perform any better than those with a vanilla bachelor’s degree. So NFL teams end up selecting Ryan Leaf with the number one pick and giving him an 11 million dollar signing bonus while passing on Tom Brady until the 199th pick. Similarly, and more important for society, we end up with teachers who have Masters Degrees, but couldn’t teach you how to make a peanut butter and jelly sandwich.

I’ve always found the way we select teachers strange. You don’t need a master’s degree in math to teach high school geometry. The skills needed to earn an advanced degree and the skills needed to be a good teacher are completely different. Sure, sometimes those skills develop in the same person, but I’m willing to wager that is the exceptional case.

Let me make a quick conterfactual argument: If education and intelligence correlated with teaching performance, then college TA’s (generally Masters or PhD candidates who are incredibly intelligent and know the material inside out) would be great teachers. I think anyone who ever had a Calculus, Physics or Comp Sci class in college will immediately agree that this most certainly is not the case.

In fact, I think you can make a good argument that the skills and attributes you need to earn a Master’s Degree in math (to take one example) are a significant negative predictor of success in teaching.

Let’s face it, there is no subject taught in high school that requires a graduate level understanding of the material. The most difficult subject taught in high school is probably Calc AB and that is something that we expect college freshmen to master. The skills needed to be a great teacher are almost entirely social: you need to be able to engage your students, explain the material in multiple ways, control a classroom, and have a remarkable amount of patience with your students. These are not the skills that graduate level academics select for.

I know brilliant people who became teachers and failed miserably. I’m also fairly certain that some of the best teachers I had were actually pretty low on the IQ scale. If we as a nation are serious about improving education we need to look at ways to find and train better teachers.

There is a lot more I could say about this: about how we face the same problem in hiring programmers, or about what the implications of admitting we are terrible at predicting success are. But it is late, and I am tired, so maybe another day.

Alexa Web Crawler - TamperedWithCookie Error

2008-12-02T19:37:00Z

I’ve recently started receiving a ton of CGI::Session::CookieStore::TamperedWithCookie errors on an application I manage. All of the errors are coming from the Alexa Web Crawler.

The application is using the Cookie Session Store (Railscast) introduced in Rails 2. Upon inspection it appears that the problem is related to the newline characters rails inserts every 60 characters. Rails encodes these newlines as %0A. The alexa web crawler seems to convert these to \n. This is the only difference between the two cookies that I can see.

Rails Cookie:


_application_session=BAh7CToMY3NyZl9pZCIlZDA1ODAzY2MwZGZjNzJkN2I4NGFjZTE5OTcxNzZh%0ANjI6FnNlYXJjaF9jYXRlZ29yaWVzWxRvOg1DYXRlZ29yeQc6FkBhdHRyaWJ1%0AdGVzX2NhY2hlewA6EEBhdHRyaWJ1dGVzewciCW5hbWUiFUF1dG8gLyBUcmFu%0Ac3BvcnQiB2lkIgYxbzsHBzsIewA7CXsHIgluYW1lIhNCZWF1dHkgLyBJbWFn%0AZSIHaWQiBjVvOwcHOwh7ADsJewciCW5hbWUiEkJvYXQgLyBNYXJpbmUiB2lk%0AIggxMTVvOwcHOwh7ADsJewciCW5hbWUiFUJ1c2luZXNzIC8gQWRtaW4iB2lk%0AIgY5bzsHBzsIewA7CXsHIgluYW1lIhRDb21wdXRlciAvIFRlY2giB2lkIgcx%0AMW87Bwc7CHsAOwl7ByIJbmFtZSIVQ3JlYXRpdmUgLyBNZWRpYSIHaWQiBzEy%0AbzsHBzs;

Alexa Crawler Request Cookie:


_application_session=BAh7CToMY3NyZl9pZCIlOGJlNGQ2ZGYzYWZjODRhZGI4YmNlMWUxZTkwNmNl\nYjA6FnNlYXJjaF9jYXRlZ29yaWVzWxRvOg1DYXRlZ29yeQc6FkBhdHRyaWJ1\ndGVzX2NhY2hlewA6EEBhdHRyaWJ1dGVzewciCW5hbWUiFUF1dG8gLyBUcmFu\nc3BvcnQiB2lkIgYxbzsHBzsIewA7CXsHIgluYW1lIhNCZWF1dHkgLyBJbWFn\nZSIHaWQiBjVvOwcHOwh7ADsJewciCW5hbWUiEkJvYXQgLyBNYXJpbmUiB2lk\nIggxMTVvOwcHOwh7ADsJewciCW5hbWUiFUJ1c2luZXNzIC8gQWRtaW4iB2lk\nIgY5bzsHBzsIewA7CXsHIgluYW1lIhRDb21wdXRlciAvIFRlY2giB2lkIgcx\nMW87Bwc7CHsAOwl7ByIJbmFtZSIVQ3JlYXRpdmUgLyBNZWRpYSIHaWQiBzEy\nbzsHBzs

I’m not the only person having this problem. Any ideas?

Interesting Things to Read

2008-11-21T17:13:00Z

I’ve recently read some excellent long form magazine articles.

Michael Lewis

The End

Michael Lewis is the author of Liar’s Poker, Moneyball, and The Blind Side (all excellent books). Before becoming an author he worked at Salomon Brothers in the 1980’s when they were the first investment bank to go public and one of his immediate co-workers invented the mortgage backed security. It is an excellent examination of just how badly fucked the system was.

Also to the “How did professionals so badly misjudge risk and fuck us all?” category are two excellent episodes of This American Life from earlier this year: The Giant Pool of Money and Another Frightening Show About the Economy.

David Foster Wallace

I’ve also recently read two excellent David Foster Wallace essays. I actually came to both organically as they were linked to from other things I was reading. Still, I feel a bit guilty to only now be reading more Wallace. (I’d only read the Federer piece and Consider the Lobster while he was alive)

Host

About talk radio host John Ziegler (Who recently got involved in some controversy with FiveThirtyEight.com and a generally ridiculous interview about an upcoming “documentary” he is releasing) By the way, the quotes around “documentary” are because he has a clear agenda in making the film, not because of its ideological content. That, in my opinion, renders it editorial or argument, but not a documentary. I would also use quotes if referring to Michael Moore’s “documentaries”

The String Theory

About Michael Joyce, at the time the 89th ranked professional tennis player in the world. However, to say that it is about a tennis player does not even remotely do the piece justice. It is about specialization, sacrifice, grotesqueness, desire, and talent. I think it was even better than the Federer piece which was amazing.

Error - rake ultrasphinx:configure

2008-11-07T21:08:00Z

A new Web Designer I work with was running into an error when attempting to configure ultrasphinx for his local Rails development environment.


rake ultrasphinx:configure

rake aborted!
Anonymous modules have no name to be referenced by
/var/lib/gems/1.8/gems/activesupport-2.0.2/lib/active_support/dependencies.rb:402:in `to_constant_name'
/var/lib/gems/1.8/gems/activesupport-2.0.2/lib/active_support/dependencies.rb:214:in `qualified_name_for'
/var/lib/gems/1.8/gems/activesupport-2.0.2/lib/active_support/dependencies.rb:477:in `const_missing'
/home/deploy/apps/stylespotted/releases/20080131024018/vendor/plugins/ultrasphinx/lib/ultrasphinx/configure.rb:33:in `load_constants'

I was baffled for this for a few days. We finally figured it out this afternoon.

He is a UI developer, and new to Rails, so hadn’t created his DB. The fix was simple:


rake db:create
rake db:migrate
rake ultrasphinx:configure

And he had a working development.conf for Ultrasphinx

Election 2008

2008-11-05T22:19:00Z

- flickr user sanjaysuchak

Anyone who knows me knows that I am very happy with the outcome, but regardless of your politics you have to admit that we witnessed something special.

A quick story

My mom worked as an election judge in Montgomery County. She was at the polls from 6 AM to after 10 PM and said she’d never seen anything like it. 2 hour lines, people genuinely excited, 65% turnout. She shared a story that I think illustrates why yesterday was special.

When the polls opened at 7AM there was already a line. While waiting, a 74 year old black man fainted. An ambulance was called, and while they were wheeling him out on a stretcher ~~he asked if he could vote before he left~~ (Edit 11/13/2008: Apparently I misheard my mother’s telling of the story. She asked the EMT about letting the man vote. Still a good story in my opinion). The EMT said it wasn’t worth risking his life over. My mom promised him that if he made it back he’d go straight to the front of the line.

A few hours later he came back in a wheelchair and cast his vote.

The Me Meme

2008-10-15T02:35:00Z

Taking a break from work to participate in a silly blog meme.

Take a picture of yourself right now.
Don’t change your clothes, don’t fix your hair…just take a picture. (should be super-easy with Photobooth)
Post that picture with NO editing.
Post these instructions with your picture.

via raganwald

Freezing to Rails 2.0.4

2008-09-22T03:11:00Z

One of our applications is still running Rails 2.0. I’m planning to upgrade to 2.1 soon, but in the meantime I wanted to upgrade to 2.0.4 to get the patch for the REXML DoS vulnerability. We freeze our version of Rails (and so should you) but I couldn’t freeze in the normal way:

rake rails:freeze:edge

This resulted in a message that the release was not found so it checked out the head instead.

I managed to freeze by running the following two commands:

sudo gem install rails --version 2.0.4
rake environment VERSION=2.0.4 rails:freeze:gems

Explanatory Text is a UI Design Smell

2008-07-26T16:03:00Z

I’ve had to switch the phone on my Verizon account a few times in the last couple of weeks. Verizon has an online tool that makes it pretty easy. You login to your account, click on “My Phone”, click “Activate a Phone” and then enter your ESN or MEID. The ESN or MEID is the unique identifier of the handset and is typically found in very small text underneath the battery.

All in all it is a very good experience and makes a fairly arcane task pretty easy. However, there is one bit of explanatory text (highlighted in red) that offers an opportunity for an improved user experience.

Since ESN or MEID is either in decimal or hexidecimal there never will be the letter O it will always be the number 0. However, my mother shouldn’t have to think about hexadecimal base numbers if she is changing her phone. A better solution would be to just treat the letter O as the number 0. Then the user doesn’t have to worry about it, and you can get rid of the explanatory text.

This reminds me of code smells. A code smell is a symptom that there may be deeper issues with the code. For instance, long methods are a code smell. They don’t mean that there is necessarily a problem with the code, but they do suggest that there is a good chance there is.

I think that long blocks of explanatory text are a User Experience smell. Sometimes there isn’t a better option, and you really need that text, but in most cases a little extra thought about the UI can eliminate the need for the explanatory text.

Smartphones

2008-07-01T18:42:00Z

I think that the release last year of the iPhone represented the biggest jump in computing since the introduction of the personal computer in the 1980’s. The iPhone is not a “smartphone” in the sense of its predecessors. It is a handheld computer that happens to do voice well.

The designers clearly were willing to sacrifice to optimize for how mobile devices are actually used. The biggest complaint about the iPhone is the lack of a physical keyboard. The touch screen keyboard is cumbersome. It is sufficient for typing short messages, but is woefully inadequate for long passages of text. However, the decision to abandon a physical keyboard allows for a much larger screen. Without the large screen surfing the web, reading emails, and watching video would not work as well. A larger screen makes it much easier to recieve information. And that is the crux of the mobile experience. A mobile device is primarily used to pull information; to look up directions, check movie times, read an email. If you need to write more than 2 or 3 lines for an email it can probably wait until you get back to a full computer. The willingness to sacrifice the uncommon to make the common delightful is what makes the iPhone great.

This is not to say that I think the iPhone is the end all and be all of mobile interfaces. As much as I love it, it is not perfect. There have to be other interesting approaches to mobile computing. Unfortunately, all the existing phone producers have done is to produce cheap imitations. Just google for “iphone killer” to see a sad array of products without an original idea.

What the mobile space most needs is legitimate competition (If you think that Windows Mobile or RIM count, then I just don’t know what to tell you) Android could be very interesting, but I’m not sure if Google is going to be able to pull it off. Apple managed to produce the iPhone by completely controlling the hardware and software, and by pistol whipping one of the weaker carriers into allowing it. Google is trying to build a platform with multiple hardware manufacturers and networks. I don’t see how they’re going to be able to get them all on one page. I’m also not sure how one is supposed to effectively develop applications for the various phones with wildly different hardware.

That said, the first round of winners in the Android Developer Challenge include some very interesting applications. I’m particularly excited about Android Scan an application that lets you scan the barcode of a book, movie, or CD using the camera on your phone and instantly get reviews, samples and comparison shop online. I think that this is going to be really useful. Imagine being able to scan the barcode of a bottle of wine and instantly get reviews as well as finding out that it’s half price down the street. This is the future of shopping.

A lot of the other winners are underwhelming. There are a lot of location based social networks that will never get to a critical mass and a lot of emergency notification apps. Still there is a ton of potential, and an open mobile platform is really exciting. I love the idea of being able to replace the email client on my phone if I don’t like it. We’re in the infancy of designing for mobile devices, and the lower the barrier to entry the better. We need more people trying new things.

In the end I think that the iPhone is going to remain the pinnacle of mobile experience for some time. The seamless experience Apple has created is going to be hard to beat. If Android comes out before the iPhone becomes available on all carriers I think it will find a decent market. I would strongly consider an Android phone before switching away from Verizon’s network.

At least I am hopeful about the future of mobile devices, and that is something I’ve never been able to say be before.

Cool Shopping Cart Interface

2008-05-02T15:57:00Z

Baggu sells reusable shopping bags. Their shop has a simple interface that makes it really easy to visualize what you’re purchasing.

Reusable shopping bags are a good way to cut down on our consumption. I bought the beautiful gray one from BagTheHabit for my girlfriend for Christmas.

Web Content Mavens Meeting: Tuning Up Your Web Analytics

2008-04-24T03:21:00Z

I attended the Web Content Mavens Meeting tonight. Phil Kemelor gave a presentation on “Tuning Up Your Web Analytics”

I didn’t think that the presentation itself provided much value. Everything that Phil said boiled down to the following: Make sure everything you do with web analytics helps the business.

In other words, think about what questions you want your web analytics to help you answer and always try to tie that back to something of business value. This is a very important point, but one that I feel is self-evident and certainly one that can be made in less than an hour.

I was disappointed that he didn’t offer specific suggestions for how to improve your usage of analytics. Instead he stayed in the land of platitudes with suggestions like “Reach out and touch a stake holder” That said, web analytics helps to answer questions that are business specific, so coming up with general advice is clearly a difficult task.

To be fair, I think that I am a bit more technical that the target audience of the talk, so maybe others did benefit.

In any case, the meeting had a good crowd and was well organized and the sponsors paid for free beer which was cool. Thanks to the organizers.

SandwichBoard Launch

2008-03-19T14:25:00Z

In September I left Lockheed Martin to start a company with my very talented colleague Ian Lotinsky. Today is a very proud day for us, as after several months of private beta we have just publicly launched SandwichBoard.

What is SandwichBoard?

SandwichBoard is the best way to create and maintain a restaurant’s website. We’ve worked really hard to make managing a restaurant’s website with SandwichBoard as easy as writing an email, updating a Facebook profile, or writing a blog entry. We let restaurant owners and managers update their menus, events, locations, and news from a web browser without having to learn anything about HTML or web design.

The best way to describe SandwichBoard is to show it in action, so please go check out our introductory video.

Why Restaurants?

I started working in restaurants when I was still in high school and paid my way through college as a busboy, waiter, bartender, and manager. I love restaurants. And I saw how painful it is for most restaurants to manage their web sites.

Every restaurant needs a web site, but getting a good one is currently an expensive and time consuming proposition. First, you have to hire a graphic designer, then you have to find, set-up, and pay for web hosting. Most of the time this involves signing a multi-year contract. One restaurant I know just spent upwards of $20,000 on a site redesign.

And after all that you can’t easily update your site. Every time you change your menu or want to make a change to a page you have to call up the designer and pay them more to make the change. This leads to restaurant web sites being out of date. And the only thing worse than not having a web site is having customers show up to find out that you now close an hour earlier than your 2 year old web site says.

SandwichBoard solves this problem. When a restaurant signs up for SandwichBoard they can start building their site immediately. In minutes they can create a basic site, and in about an hour they can create about a complete site. We give them a choice of several pre-built themes. They can use the theme as is, customize the theme themselves, or higher a designer to build them a custom theme. Once the theme is set up they can easily manage their entire site without the need to involve (and pay) the designer.

We’re so confident that restaurants are going to love SandwichBoard that we let them try it free for a month. We also don’t lock them into a long term contract, so they can cancel at any time.

Please Help Us…

Making it to launch is a big accomplishment, and is cause for celebration (And believe me, I will be celebrating tonight)

But now comes the truly important part: getting customers. So please help us spread the word. If you know people in the restaurant business please let us know. If you’re going out to eat please check and see if they have a web site. If they don’t, let us know. If they do and it’s bad, let us know.

If you know graphic designers who are interested in some work let us know. We would love to refer restaurants to them. We manage all the hosting and software, so designers can focus on design. We’re not looking to make money off the designers, they keep whatever fee they negotiate with the restaurants, we just want to get good designers building beautiful web sites on SandwichBoard.

Thank You

Thanks to everyone who has put up with me for the last several months while we built this application. I flaked on plans, was generally a bad friend, and for that I’m sorry. For those of you I’ve been bad about keeping up with please know that you are in my thoughts and I hope to catch up with you soon.