philipp's blog

zeitungsbote

philipp's blog

zeitungsbote is of my bigger projects, which I remixed multiple times over multiple years.

One of the problems of journalism - despite payments, that's a whole different beast - is curation. A classic newspaper represents a curated collection of articles, some interest you more and might even have been the reason why you bought this edition, some are less interesting. But the point is: You could not pick which articles to buy, you either bought the full newspaper or a different one. The only pre-selection you did was the choice of the paper, which in itself was a choice of tone and worldview. Once settled on a paper, you'd read through the articles that interest you - of course - and, depending on the time you had at hand, also some that you'd never consciously have picked yourself.

Online journalism, in 2017 and still today, suffer from a lack of curation. Superficially less interesting articles have it much harder now to be read, which means that today more than ever, truth must not be important to be read, it needs to be entertaining. We're lucky if it's both, we're in trouble when it's important but boring.

I tried to fix that.

Community to the Rescue

In 2017 it still seemed like a good idea to make ranking a community effort, and my initial idea for zeitungsbote was a mix between Reddit, Stack Overflow and Hacker News: Users would post links to news, other users would interact with it.

For users, positive interactions would lead to benefits such as access to moderation tools, your voice (read: submissions) carrying more weight and more.

I put a lot of effort into various ranking algorithms, like interactions over time, views, view-to-read ratio1, similarity to other submissions (to avoid a front page full of articles around the same issue) and much more.

My goal was to create two main curation mechanisms: The front page, which should allow anyone checking the news once per day to find a collection of articles which bring them up to speed about what's going on. And what I called the Sunday Morning Coffee, which aimed to give the reader an overview over what happened in the past week.

It worked until it didn't

It worked, it really worked. My community was too tiny to run into the eventual problems of StackOverflow2, and I soon used RSS feeds from various media so that I didn't need to rely only on user-submitted content.

I was proud of what I've built, and people generally liked it.

Then Covid happened. Since Covid came with a drastic change of pace for society, every news outlet produced article after article with very little new content. And I was hit with a severe case of news fatigue, which is not a good thing to have when the project you're working on is about news. Additionally, Covid also meant that I had no more free time at hand, so I needed to pause zeitungsbote as well, and I eventually deleted it.

I'm no longer convinced that zeitungsbote as envisioned in 2017 solve the problem as I see it, but I still think it exists and it's worth solving. Who knows, maybe I'll build another iteration in the future.

  1. I considered it a strong signal for an underrated, but interesting article if people who saw it, read it, even if they didn't upvote it.↩

  2. For example, the idea that competent posters (as evidenced by the upvotes) are also competent moderators is a bit of a stretch.↩

#pet-project-sematary