Copyediting and Computer Code

Every now and again I’ll find myself in a conversation with copyeditors about the future of their craft. One point I often bring up is that a big part of the job in online newsrooms needs to be overall QA of the site. And one of the most challenging workflows to support that is the copyediting of computer code. The example I always use to illustrate the point is the AP style on state abbreviations. If the Web developers define the abbreviation for California as “CA” instead of “Calif.” … well that’s something that should stick in the craw of every copyeditor until the code gets changed.

And now I have an actual piece of code to illustrate the example. (This comes from the code that runs OpenBlock — the much awaited open-source version of Adrian Holovaty’s EveryBlock. This isn’t meant to pick on that community. They’re doing difficult and needed work. And this could happen anywhere… which makes it a good anecdote.)

What’s the workflow in your newsroom for making sure that this gets changed to “Reporting Officers’ Names” before launch? Should the designers give editors a mock-up of all the static text elements (including words-as-graphics) on the page? Should the developers give editors printouts of all the tables that contain datafields that might get on the live site? Or do you just publish and come up with some sort of sampling scenario?

How does it work in your newsroom? How should it?

Ryan Thornburg is the author of the new online journalism textbook and newsroom manual, Producing Online News, available from

Article Comments Are Alienated Experience

Jaron Lanier, one of the pioneers of virtual reality, once kindly said — I guess — that I often use when thinking about or speaking about online journalism: “Information is alienated experience.” A blog post from one of my students at UNC has done a nice job recording an anecdote from the 2010 Online News Association conference that I think brings into focus the role of comments as form of alienated shared experiences.

Michelle Cerulli, a second-year MA student, told me this story and I encouraged her to blog about it. The short version is this: While attending a session about article comments, she watched a mild-mannered man use Twitter to quietly excoriate one of the speakers. This man didn’t stand up and confront or question the speaker in person. Instead he used this virtual soapbox to disagree with her — in what Michelle described to me as incredibly rude terms — about the role of comments on online news articles.

What was his beef with NPR ombudsman Alicia Shepard? She was saying that online comments tended to be more vitriolic than you hear in “the real world.” His words on Twitter said that Shepard was wrong. But his behavior said that she was dead on. And, according to Michelle, he appeared to be oblivious to the irony.

And while this story so far might seem to some a perfect set-up for a conclusion in which I rail against online comments, that’s not where I’m heading. Online comments are important because it is there that our collective id gets revealed. Many of us reveal in anonymous or pseudonymous comments our fears and hopes n ways that most of us would deny if we were ever confronted with them. Online comments show how us — or at least some non-representative sample of us — experience the world in a way that we alienated from ourselves and the polite company around us.

And that unfiltered id — that alienated experience — is a happy hunting ground for a reporter who hopes to more clearly explain to his readers our increasingly complicated and interconnected world. The problem with comments is not that they are mean. The problem is that there are too few people mining them for hidden hopes and fears and too few people willing to patiently ask probing questions of the crowd.

More and more news organizations are hiring “social media producers.” I hope they’re given the challenge of not just distributing the news to the crowd, but also diving into it and finding individuals who are able to articulate why they’re much more scared, angry or jealous than they are willing to admit in a room full of their peers.

Audience Engagement Starts With Audience Tracking

In the match-making game that is the summer internship and job hunt now getting underway at J-schools across America, I always warn students to never take a job working for an editor who talks about how many “hits” her site gets. And I train my students so that they’ll never be the person whose resume gets tossed for doing the same.

Chapter 3 of Producing Online News and its related tipsheet provide a good overview of the who, what, when, where and why of online news audiences. (And that’s something that’s always changing, Pew reported yesterday that 4 percent of online adults — and 10 percent of Hispanic online adults — use geosocial tools such as Gowalla or Foursquare.)

But students can begin to learn both mass communication research concepts as well as skills if they have the chance to use Google Analytics (or the high-priced and industry dominating Omniture service) on a real, live news site. That will prove to be one of the strength’s of UNC’s new Reese Felts Digital Newsroom, an opportunity that is still pretty rare for journalism students.

The good news is that the Web is full of good free guides to using Google Analytics on a news site. Here are four good guides to get you started:

  • Tracking Your Users (
  • The Journalists’ Guide to Analytics (Mark S. Luckie)
  • Google Analytics – Adding Tracking Code(Brett Atwood)
  • Installing Google Analytics
  • Lessons From ONA ’10: What It Takes, Part 2

    Aggregation continued to be one of the online news community’s big buzzwords at the 2010 Online News Association conference last week. The idea behind aggregation is that individual news organizations can achieve comparative advantages and that the entire information economy can function more efficiently if the news organization links to reliable information from bloggers, sources and other news organizations rather than replicating the information with its own take.

    But aggregation isn’t free. You can either automate it, which might cost a newsroom $25,000 to $100,000 in up-front costs, plus constant tweaking of the algorithms and processes that gather, organize and automatically publish news stories from external sources. Or, you can put humans and their infinitely superior cognitive flexibility on the task.

    But what does that cost? Based on some estimates I’ve put together based on conversations at ONA:

    * It takes an average of 8 minutes for a news producer to read a blog post or news story, write a summary and categorize it by location and subject.
    * Based on a VERY limited sample that desperately needs further research, you can estimate pulling in one blog post per week for every 4,500 people in your market. (Please send me any data you have that would help me solidify this number.)

    In my home market of Raleigh-Durham, which has about 1.5 million people, aggregating local content might take about one full-time position and cost a news organization maybe $35,000 a year plus benefits.

    How does that match your experience with aggregation? What am I missing?

    Lessons From ONA ’10: What It Takes, Part 1

    At least three national news organizations approached me at last weekend’s Online News Association conference to see whether I could recommend any students with great news judgment and programming skills. That’s what news organizations are desperate to hire today. Why? Well, as former president George W. Bush will tell you some things — like learning how to program — are just hard work.

    Lunch with a friend last week helped me put some numbers on just how hard it is. I was meeting with him so that he could show me the server he set up and the computational journalism he had been doing since we last had a chance to catch up. At heart, he is a writer and a reporter, yearning during our conversation for the chance to do more long-form narrative text stories. But in his newsroom, he is the resident programmer/journalist and has asked by his editors to hire more people like him.

    Here’s what it took for him to become “tech savvy.”
    * In high school, he took one computer programming class. He didn’t study or use computer programming at all in college. He wrote and edited stories at the campus paper. After graduation, he was hired in jobs as a researcher or blogger.
    * During the last two years, he taught himself how to code. He set up his own Ubuntu server, with PHP and MySQL. He learned some ActionScript, JavaScript and XML. He uses Excel, Visual Basic and to report stories and build interactive editorial Web applications.
    * He works 60 to 75 hours per week.
    * He spends 90 percent of his time working with and learning about computer coding.
    * It took him two years to get to this point of technical proficiency.
    * That is a total of 5,500 hours.

    He was not born with the IT chromosome. He did not wish himself to state of savvy. He has clearly been blessed with an incredible brain that was nurtured in an environment that valued education and intellectual curiosity. But that didn’t get him his job. He got his job because. He. Worked. Hard.

    Let’s point out how difficult it is to get 5,500 hours of computer time under your belt.
    * College students spend about 15 hours a week in class. Good ones will spend another 25 hours reading and working outside of class. That’s 480 hours a semester, 560 hours a year. At that rate, taking ONLY coding classes, you’ll get to 5,500 hours in just under 10 years. Which makes you this guy. Nobody wants to be that guy, so it’s time to accept that editorial programmers are committed to life-long learning.

    * Let’s say you knock out a few coding classes in school — 500 hours worth — enough to get hired by a big news organization as a developer. That leaves you with just 5,000 hours to go. Working a standard 40-hour week, you’ll burn through those in 125 weeks. That’s about 2.5 years, after various and sundry holidays, illnesses and vacations.

    * Or, maybe you were a good liberal arts student and didn’t blow any of your tuition on coding classes. But your smarts and broad-based knowledge land you a job at one of a very few news organizations that commit seriously to career development. Google spurs innovation with its famous “20 percent time,” which allows its developers to spend a day a week working on projects that are not part of their job descriptions. So, your boss lets you play with computers for one day a week. You’ve got 5,500 hours to make up. And by the time you’re celebrating your 35 birthday you’ll probably be at the point where you can start developing your own editorial applications.

    What the conversation with my friend made me realize is why it irks me so much when people come to me saying that they can’t perform some computing taks because they are “technically illiterate” or “not a computer person.” My friend isn’t a computer person. I’m not a computer person either. But we try. We hack our ways through incredibly frustrating failures by simply doing this. And so can you. If you want.