This is re-posted from the News Apps Blog.
Last week, we launched a new application for the RedEye – the Chicago homicide tracker. The web site makes it simple and interesting to browse homicide crime data for the city. RedEye reporter Tracy Swartz has been compiling the homicides since Jan. 1, 2009 and writes a weekly analysis. She wanted to give readers a better way to browse and understand the data and we wanted to help but never had enough time to give the project the attention it deserved.
The homicide tracker might look familiar if you’ve ever seen the L.A. Times homicide project. That’s because it’s the same code. LA Times hacker team of Ben Welsh and Ken Schwencke generously let us use their code (caveat: we all get our paychecks from the same place – Tribune Co.). It took four days of re-factoring, reorganizing, writing new data loaders and a new skin to make the L.A. Times code work for the RedEye.
This kind of project plays to the strength of the small newsroom dev team. We started with a small-medium application that was built to solve a specific problem, but not to be reusable. We worked with the reporters to figure out what about the L.A. Times app we should keep, what we should scrap and what we should change. We ignored the urge to refactor and leave as much of the original code as possible, tweaking only what was necessary. With the help of the author of the original, we were able to quickly make our changes and launch.
Free and open technologies are key to our small teams working quickly. Pulling content and data from RSS and Google spreadsheets allowed us to skip building a content management system for the homicide tracker. Using a sophisticated, modular web framework helps to make us efficient.
The moral of the story is that for news apps, small teams sharing code, insight and ideas – “small pieces, loosely joined” – is quite effective.
While working on News Mixer and another django project, I’ve refactored all the facebook connect code Brian Boyer and I wrote into a separate application. I’ve opened a google code page for it. I figured if it’s useful to me, it would be useful to others.
It’s a simple thing really. It uses the built in django auth stuff and pyfacebook, and works with django-registration. News Mixer was designed so you had to use facebook and only facebook to log in. The django-facebookconnect application can work side-by-side with regular django user accounts, and it gives new facebook users the option to link their django and facebook accounts.
While building News Mixer, we had a lot of problems getting facebook connect to play nice. Timeouts, expired sessions and connection resets would quickly kill our application. On the face of it, the app doesn’t really do a whole lot, but under the hood it deals with a ton of BS problems that you run into using the facebook API extensively.
Take it, use it, let me know how it goes.
Tonight the group leaves for Delhi after traveling through the towns and cities of Madhaya Pradesh. It was great country, friendly people, and only a small bit of the hustle and crush of Mumbai. We saw buddhist and hindu caves, hills carved into forts and magnificent palaces.
We stayed in some beautiful places, off the beaten tourist track and got to relax. Traveling with a guide makes for a relaxing trip – having somebody who can translate and make the plans takes a load off your back. But I am happy it is finished. Relying on somebody and traveling with like minded tourists is quite insulating.
I’m looking forward to Delhi and meeting some friends in a few days. There are new photos up and more to come.
A couple weeks ago I stumbled across a neat little reading applet called spreeder.com. You copy and paste text into the applet, and then it would flash each word at you one-by-one. The idea being that it forces you to read faster, teaches you speed read or whatever.
To be honest spreeder.com was a little annoying to use because I don’t read one word at a time. It takes a lot of concentration just to follow the story and is not a great experience. I thought it might be usable if it showed groups of words and phrases and put pauses between sentences and paragraphs.
However the little applet struck me as something that would fit very nicely on a small screen.
Well guess what folks. Someone did exactly that. Spreed:News is a relatively new web-app that rolls up an RSS reader with a spreeder-like widget. Spreed:News is a bit better than spreeder.com in that displays a few words at a time, instead of one. And they just launched an iPhone app.
Neat. If I had an iPhone I would give it a try.
They need to get this out for more mobile devices and start working with content providers.
The Sun-Times published the names, salaries and positions of 145,000 Illinois, Cook County, and Chicago employees on their website this weekend. The names and salaries are online in the form of a simple searchable database. The reporter published a couple articles with analysis of the massive database. One article talked about the top ten earners and another took a look at the Chicago police commissioner, who is the top earner.
A classmate of mine told me about it. Why did they just publish it all, she asked. Most of these people are just honest employees. She looked up the name of her friend who works for the government and found out how much he makes, which is what the majority of folks will do with it.
I think could be an interesting open-source approach to journalism – make a FOIA and release all the data to the public (as long as there are no privacy implications, which there might be in this case). But the Sun-Times didn’t really release the source to their product. You have limited access.
I want a link to download a csv file. I want to plug it into Many Eyes. I want to run my own reports on it.
I don’t know if they published the information with the intention that others should use it to find stories. But that would be cool.
Looks like the new Tribune editor, Gerould Kern, is working on “saving” the newspaper. The reported redesign takes the news off the front page and buries it in the second section. Whats more important for the front page of a newspaper?
“Consumer-oriented and entertainment features.”
According to the article at Crain’s ChicagoBusiness.com, Kern said the redesign is still a work in progress, but the Tribune Co. COO Randy Michaels has ordered some kind of redesign. I think the money quote from Kern is thus:
“The newspaper business is in crisis. I want to do everything in my power to save it.”
Sure it would be great to go back to the golden days of year over year growth, but the Internet has changed everything. The editor of the Chicago Tribune should be focused on saving the newsroom: the reporters, quality and values. Not an obsolete distribution mechanism.
I’m at a meeting of the Knight Foundation grantees in Chicago, today. We just finished lunch and heard the CEO and President of the foundation, Alberto Ibargüen, talk about how Knight has changed its focus from promoting best practices in journalism to figuring out what those practices are now. While talking about their focus on local and community news, Ibargüen said that they aren’t trying to “save the newspapers,” they’re trying to “save the values” of journalism.
Out the window was a reflection the Tribune Tower.
I wonder when those condos go on sale.
I’ve often wondered what would happen if we could put the legislative branch of the government online and let citizens vote on legislation directly. Get rid of all those layers of representation that we built up and have a true democratic system.
While that may never happen (and maybe should never happen), I stumbled on a site that allows you to vote directly on bills in congress and send your position into the appropriate representitive/senator: Govit.com. It keeps track of your state and congressional district and how you vote, and stacks up it’s members against thier representitives in Congress.
The site looks new and seems to have a small community at the moment, but it will be neat if it gets enough interest.
The UK government’s ‘Power of Information Task Force’ is holding a competition for the best way to make government data available to citizens. They are offering a 22,000 pound prize/grant to make the best idea happen. Whats really cool is that a bunch of UK agencies published new APIs to get the juices flowing and encourage some mashup-making.
On this side of the pond, a technology policy paper was released with a name that sounds like a Harry Potter book: “Government Data and the Invisible Hand.” The Princeton University authors argue that the government agencies make thier data available on their own websites when they really should be building APIs to let the public decide what to do with it.