1. 30

As best I can tell, there’s a moderation practice along the lines of:

If a story relates to a current topic, and there’s been a previous post related to the topic within the last week, fold the new story into that previous story.

This has several downsides.

  • Posts suddenly vanish from the front page, often-times in the middle of active discussion.
  • Merging disrupts the comment threading. There’s rarely a consistent discussion after a merge. Oftentimes comments are quite specific to the particular article they were posted in response to.
  • Some relevant and interesting news effectively get hidden from a large part of the readership.

As a current example, you wouldn’t know by reading lobste.rs that Richard Stallman resigned from the FSF, because all follow-up stories were merged into the original accusations: https://lobste.rs/s/yxj6cd/remove_richard_stallman

Some suggestions for improving the current state:

  • Improve merging technically to help avoid the negative effects, e.g.:
    • bump stories to the front page if a front page story gets merged
    • allow browsing comments by original story
    • don’t lose “saved stories” on merge
  • And while that’s not addressed, some immediate tweaks to the policy, like:
    • cut the “merge window” to something like 24 hours
    • limit the maximum number of stories within on merged story, to something like 2 or 3
    1. 23

      (Um, holy crap, Richard Stallman resigned from the FSF? I’ve been on holiday for a week and I’m mostly offline but i do read lobste.rs and it’s about the only source of tech news/updates i care about, I had no idea rms resigned, and I’ve just found all the articles merged here and certainly more than one of them was worth reading.)

      Why isn’t this in control of the community rather than moderators? ‘Already Posted’ seems like the mechanism whrereby posts that cover a topic people feel has already been covered can be flagged. Perhaps there could be a convention that a few Already Posted flags are required before merge.

      1. 3

        Why isn’t this in control of the community rather than moderators?

        Because no one’s done it.

        Though to update my comment in the linked thread, I’ve seen enough bugs shake out of merging that doing and undoing it no longer sound so trivial. We probably want to refactor the data model to force us to think through the corner cases first (eg. if user hides a story and a link is unmerged from it, should they be hiding that standalone story?). I like your suggestion of requiring a few Already Posted flags.

    2. 17

      Like I mentioned in the original thread, it’s really bad to use this feature for anything except for effectively merging identical pieces. The original medium post was its own post, while the resignation announcements could’ve been merged, while the reflection medium post was also its own entirely independent piece of content. Each of these three groups have exactly nothing to do with each other content-wise except sharing a topic. And like I quipped in the original thread, if lobste.rs were to be consistent about a rule like this, we’d best merge every single topic into their megathreads weekly. Or maybe this is what tags were intended for, who knows.

      1. 4

        I second. I think the accusations of unfair moderating really cluster around the fact that, despite being the same person, these are separate events that we are discussing.

        I feel like the discussion about if this content is on-topic for lobsters is secondary to the fact that there are 3 different stories being collapsed incorrectly into one. If any one of those is off-topic, let the community decide. Let’s not muddy the issue by claiming RMS is a single topic in this case.

    3. 15

      Lobsters’ story-merging policy has definitely made it harder for the community to follow and discuss the RMS story. When I first heard the news, I was definitely interested to read what Lobsters community recations, and this policy has been a bit frustrating.

      However, I’m sympathetic to the notion that Lobsters is supposed to be a venue for discussing technology, not for discussing people, no matter how technology-relevant they may be. This kind of rolling-news-cycle of news and reactions and rebuttals-to-reactions really only shows up for stories about people, and maybe that time somebody deleted an NPM package and knocked half the internet offline. The point is, if the merging policy is making it difficult to discuss tech-scene drama, perhaps it’s just doing its job. If we want to discuss tech-scene drama, we can go to the Orange Site, or Reddit, or Slashdot, or IRC, we don’t have to discuss it here.

      1. 7

        I did some digging into the question of venue for discussing technology or people. Search currently turns up 693 stories tagged person, 8 of which have over 100 points. The filter page counts the person tag as having 727 stories, placing it in the 60th percentile for tags. This suggests to me that we actually do have a non-trivial number of users that do want to discuss people. Is this not what the filtering mechanism is for?

      2. 3

        The other big merge-fest I remember was when Github was acquired by Microsoft. But in general, I think the story-merging function is a good thing.

      3. 1

        The community already voted majority to discuss people and political issues. That’s already in. This thread is about the story merging policy.

    4. 8

      That’s a correct statement of the practice. Thanks for reverse-engineering it; I’ve added it to the about page. (Not to forestall discussion and put my foot down, just catch up on a missing detail. Changing a sentence is of course trivial if we adopt a different practice.)

      Has merging felt like a problem on any story that’s not breaking news about a contentious topic? I compiled a tsv of all merged stories so we can see the scope of the topic, but my gut says no, or very few. (And shout-out to @vale for compiling stats, I really appreciate efforts to try to build a shared foundation for discussion; reminder of the /about note that I’ll run queries.)

      If the saving feature doesn’t work properly on merged stories, that’s a bug, and please file it. And if saving is broken, hiding probably is as well. The way merged stories are implemented (self-reference instead of a story record with a one-to-many relationship a submitted link/text) seems to be a regular source of low-level bugs. Just this week I identified two more and I recall several more I don’t have time to search for. That’d be a significant refactoring but solve the issues at the source.

      Browsing by story or knowing which link a comment refers to seems already addressed by those comments including the title of the submitted link. Could you elaborate on why these are problems?

      I’m a big fan of story merging and the current week timeline because:

      1. It keeps discussion together. When other sites split discussion across multiple links it feels like 90% of each discussion reiterates the same points and maybe 10% is generating interesting new ideas. When Lobsters merges a story we don’t waste commenters’ time on reiteration so a much higher percent of the discussion is really high-quality, novel material.
      2. It keeps the front page uncluttered; I’d rather see 25 different stories on the homepage.
      3. It reduces the value of writing and submitting hot takes for content marketers. This summer feels like we’ve seen a steady increase in topical-but-promotional stories on Lobsters.*
      4. It’s useful for hiding. If you’re really not interested in a thread, you click ‘hide’ and it’s gone. Maybe you’ll see it again in a week, but probably not.

      If the problems of merging come down to it being bad for news, I’d rather remove news than merging. There are many other sites for news and it comes with a lot of downsides like trolly hot takes, sloganeering, personal attacks, rewarding thin blog posts that reiterate breaking news, and just generally being more about novelty and quick response than reflection, context, experience, and collaboration. I don’t feel like there are sites that do deep dives into computing as well as Lobsters does and I want to protect and invest in that.

      (And I see the other meta thread on politics that sure feels related, but I need to leave for a wedding.)

      * On this: rough estimate here, but with our current level of traffic of ~15k developers per day, a link on the front page sends an amount of high-cost demographic traffic that’d cost $10-25k of Twitter/adwords advertising or sending recruiter messages on LinkedIn. We’re going to see more marketing attempts because there’s more potential value to companies. News and reactions to news are cheap to produce.

    5. 8

      Quick pointer to two related comments. I’ll try to find a break at work to respond fully this this, but more likely it’ll be tomorrow morning.

    6. 9

      I have no problem with the current merging strategy since the things that get merged are often hot topic news and generate little to no useful discussion. It’s one of the reasons I’ve not been active on Lobste.rs the last while.

    7. 4

      I’m in favor of the enhancements suggested here.

      I’ve had issues with anchor links to comments not seeming to work on some platforms[1] - the comment is highlighted but it’s not on the top of the page. I’ll try to reproduce and file a bug if I find the time.

      Bumping stories makes a lot of sense, it would address the issue where people complain the story is being “buried by the mods”, and it would also signal to those interested that “new stuff is added to the dumpster fire discussion”.

      It would be nice if the merged comments had their own “heading”:

      * Story 1
       * Story 2
       [comments to original story]
       ## Story 2 comments
       [comments to the original story]

      [1] Chrome on iPad, but seems to work fine on Chrome on iPhone)

    8. 3

      I’m in favor of all of these ideas or something like them. I was also confused by the lack of story about the resignation on the front page. Thank you, @A-Za-z0-9, for helping me understand why. I did some quick and dirty statistics below, which suggests to me that there are other people on the site wishing to discuss topics like this. (Posting separately so data and conclusions can be evaluated separately.)