Hacker News Reader

Comment by hitekker | Hacker News Reader

Three factors.

1. Reddit's internecine subreddit wars required constant admin intervention and mediation. Reddit only has so many hands on deck, so more of than not, they pulled engineers in to help moderate.[1]

2. Reddit is unprofitable, so they have trouble affording top talent. [2]

3. Finally, Reddit's vision for the future has been hazy for the last decade. If the navigators can't chart a course, all the rowing in the world won't move the ship in the right direction. I suspect the overdone April Fools experiments are a result of engineering not being given technically compelling, or perhaps meaningful features to labor over.

Sources:

[1] https://www.reddit.com/r/TheoryOfReddit/comments/58zaho/the_...

[2]http://www.thedrum.com/news/2017/04/02/reddit-sets-eyes-prof...

View on HN

1. This doesn't require developer talent, or any real expense at all, social media marketing interns will do the job at barely any cost.

2. They've had plenty of money from investments though, their dev team is reasonably sized.

3. This one seems pretty accurate.

1. I agree that it shouldn't, but they did so. Updated post with link. In another post, he emphasized the "war" part much more, but I think he deleted it. It was fairly angry.

2. I'm having trouble finding their burn rate; the linked article at least hints that they'll use this money to attract talent.

Abuse, like everything else, changes at scale.

There's only so much moderation a single person can handle in a day. The NY Times moderation team posted some stats a ways back, it worked out to about 800 items/person/day. Which, in an 8-hour day works out to an item every 35 seconds. Sustained. Throughout the entire day.

Techs can leverage patterns, even if it means a lot of ad hoc code.

Building training systems is even better.

Considering what Reddit's been through in the past couple of years by way of direct information attacks (even not counting any systems-level attacks), involving coloured tablets and giant orange peaches, I don't find the story that dev time has been accounted for particularly unreasonable.

1. Yeah but no engineer should have to get their hands dirty in community moderation.

2. Come on, it's Reddit. I feel like many many developers would be interested on brand alone. I was, for a time.

3. Possible. But there's a difference between adding random directionless features (which they haven't really done), and not doing anything at all (which they have).

1. Right, they shouldn't, but unfortunately, they had to.

2. From what I hear, Reddit wants the best, but the best are aware that Reddit's brand conceals well-documented dysfunction:

https://www.recode.net/2015/7/13/11614676/reddit-chief-engin....

https://np.reddit.com/r/IAmA/comments/2iea97/i_am_a_former_r...

The links go on.

3. I worked recently at a startup where the product was a shambling, swelling mass of features, implicitly designed to obscure its missing core value proposition to both investors and to the founders. There is definitely a difference in the process; perhaps not a meaningful difference in the outcome, which for those invested in the company is impotence, toxicity, despair, etc.

Re 1. if they're working on moderation code then they should get their hands dirty regularly.

Reply on news.ycombinator.com