Hacker News Reader

Comment by albertsun | Hacker News Reader

It's really depressing how cynical that is. (Developer working at the NYTimes here).

The actual reason links to studies are often left off articles (particularly in cases like this with sharp deadline pressure to publish fast before the competition) is that the software powering the editing and publishing workflow really badly needs improvement. An incredible amount of work and knowledge goes into a story like this.

Versioning rich text through many different software tools designed for writing/editing and publishing across many platforms is hard. Sometimes people copy/paste by hand and in doing so a link can go missing. The news industry needs more technologists to work on these problems. We're hiring for people to do that, by the way: http://developers.nytimes.com/careers/

(Also, the link to the study is now in.)

View on HN

It's cynical because that's what we, regular readers of news, have to deal with. NYT may be top of the line, but sadly it's not above the line.

Now I don't want it to sound dismissive or personal in any way, but tell me - if say, few of great software devs now drop everything they do and come to NYT to help, sit down for months and develop the most awesome software package the world of press has ever seen, will it actually solve the quality issues articles have? And more importantly, if sold to other papers, will it suddenly solve their problems?

Will it make journalism honest and trustworthy instead of lies and clickbait bullshit?

I'm not sure how much blame to put on broken publishing workflow, a lot of this seems really to be about broken incentives - "deadline pressure to publish fast before the competition" that leads to the "many articles, as sensationalist as possible, truth be damned" mentality, especially in the management layer.

But you did give me a pause here. Only recently I had a chance to peek at internals of a tiny part of manufacturing industry, and oh boy how much money they waste on badly designed software, which is badly designed because of deadline pressure and top management pressuring to iterate over a broken software package (and then messing with the process) instead of scrapping it altogether and doing it right. Maybe software is more to blame than I thought.

> Now I don't want it to sound dismissive or personal in any way, but tell me - if say, few of great software devs now drop everything they do and come to NYT to help, sit down for months and develop the most awesome software package the world of press has ever seen, will it actually solve the quality issues articles have?

This actually happened and that's how we got d3.js. It didn't fix journalism though.

Oh, didn't know the origin story; I only thought they just like it over there. D3 is absolutely awesome!

>Will it make journalism honest and trustworthy instead of lies and clickbait bullshit?

You get what you pay for, which online tends toward nothing.

It turns out that paying people to talk to you, although it can work, tends to get them to tell you what you want to hear. Often, this is "lies and clickbait bullshit." (Think of a stereotypical Soviet government report.) Markets work very well for lots of things, but they can't establish honesty and trustworthiness. Instead, they need honesty and trustworthiness to function.

Still, it's better than them getting paid to tell you what someone else wants you to hear.

I don't recall ever having a problem with any actual reporting.

My objection is with the editorials. I've all but stopped reading "the news" (NYT, WSJ, Economist, my local paper, etc) as a result.

This is the best sort of thing that happens at HN.

"I don't like [thing]" "I work at [thing], come help me fix it"

You miss that the subject changed from media in general to NY Times. It is like mixing apples and ... well, rotten pears.

(Not claiming that NY Times is perfect. Of course. There are other good sources, too. But NYTimes isn't like the media I grew up with. At 20 I realized that all their coverage of subjects which I knew beyond the surface was garbage, at best. My specific example is DN, the largest Swedish morning paper, but could be most of the media.)

I'd be happy to be corrected of course.

I'm glad to see that the link to the study is now in. Also, the current version of the article includes phrases like "the study’s lead author, Dr. Robert Lustig"; did it originally? I can't find the link to the version history of the post.

I'm skeptical of this "software" explanation. Software can of course make citation management much easier, but I see lots of articles that don't even bother to mention the lead author of a cited publication; and how did the software get that way in the first place? The software reflects the priorities of the corrupt organization that produced it.

Having a chance to peek at a tiny area in manufacturing process, I can sort of imagine software replacing a more manual process by something faster, but much more messy and generally worse. It starts with three layers of corporate management above a subcontractor hired to write the software. Of course all requirements go through the entire chain, in what resembles and adult version of the game of telephone (aka. "deaf phone" or "Chinese whispers", the latter being particularly appropriate since what I saw, I saw in China...).

Somebody could probably sit down with the journalists for few weeks and come up with a software package that would fit their needs perfectly - if the development consisted of direct communication between the developers and journalists/editorial staff, without any management middlemen in between. Alas, that's not how software is procured in large organizations.

But I'm also skeptical. It could explain NYT's problems, but it doesn't explain the even worse problems every other news outlet has.

The reason articles about studies leave out a link to the paper is that the software doesn't support it? Why not just put the URL in, or just cite it normally?

Yeah. I can imagine they have lots of problems with preventing tidbits of information from getting lost in an environment of multiple people chaning the same block of free-form text, but if the article is literally about some particular paper, then starting the file with:

NOTE: Study "Effects of X on Y given XYZ" by Researcher N. Here, http://address-to-paper.org

(not the placeholder, the actual data) should help. I don't see such a line getting lost, and even if they happen to publish it by accident instead of incorporating into the text (as sometimes happens, things slip through), it would still reach the same goal anyway.

Good luck in your work, but unfortunately it wasn't the tech stack that prevented NYT from publishing the warrantless wiretapping story in due time.

The NYT managed to put together an absolutely massive printing and diatribution operation. If editorial cared about putting a citation in a story, they would do it. Getting a journal citation right is no harder than spelling the scientist's name right. Tech talent is not the blocker.

> sharp deadline pressure to publish fast before the competition

Sounds like the problem lies here and not on the technical side of writing, editing and publishing text.

I wonder what part of this decadence is imputable to ad revenu. If I were very naive, I would say all of it.

Reply on news.ycombinator.com