Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. San Francisco Onion Futures Company(onionfutures.com ↗)
    35comments
  2. Android 17 is the first since 3.x to add new APIs without releasing to the AOSP(grapheneos.social ↗)
    357comments
  3. SDCC – Small Device C Compiler(sourceforge.net ↗)
    12comments
  4. Typesafe-computer-use drives a Mac toward a goal for 1/50th of a cent per step(github.com/awlevin ↗)
    9comments
  5. Human brain is two separate organs, Stanford Medicine-led research finds(stanford.edu ↗)
    2comments
  6. Science Is Open Software(jepedersen.dk ↗)
    22comments
  7. Cloudflare Quick Tunnels(cloudflare.com ↗)
    272comments
  8. How OpenAI Used Its Own LLMs to Design Its Jalapeño Chip(ieee.org ↗)
    74comments
  9. Saving another 100TB of RAM(cloudflare.com ↗)
    59comments
  10. Why building a Rust LSP is hard(rust-glancer.github.io ↗)
    18comments
  11. NASA-IBM Lunar Foundation open-Source Geospatial AI Model(usra.edu ↗)
    discuss
  12. How to Write with an LLM(sockpuppet.org ↗)
    309comments
  13. You can run Git on object storage if you re-make packfiles(tigrisdata.com ↗)
    1comments
  14. Goroutine Leak Profiles(go.dev ↗)
    2comments
  15. Xcode 27.1 Beta Release Notes(developer.apple.com ↗)
    84comments
  16. The first new cat species discovered in 100 years(nationalgeographic.com ↗)
    89comments
  17. Show HN: Cactus Needle 3: 8-29MB automation models can match DeepSeek V4 Flash(cactuscompute.com ↗)
    83comments
  18. OpenJev(openjev.com ↗)
    255comments
  19. Photon-Emission-Guided Laser Fault Injection Enables RP2350 Secure Debug(ledger.com ↗)
    64comments
  20. The Farnese letter(simonklee.dk ↗)
    6comments
  21. Cache-to-Cache: Direct Semantic Communication Between LLMs (2025)(arxiv.org ↗)
    12comments
  22. Minimal Phone 2(minimalcompany.com ↗)
    206comments
  23. Cyclomatic Complexity in C#(ndepend.com ↗)
    17comments
  24. Claude Code now reads AGENTS.md if there is no Claude.md(claude.com ↗)
    214comments
  25. LispBM is a concurrent Lisp for microcontrollers with message passing(lispbm.com ↗)
    3comments
  26. Inside ZCode: Silently uploading your Git history to the cloud(ferstar.org ↗)
    96comments
  27. Warez: The Infrastructure and Aesthetics of Piracy (2021)(archive.org ↗)
    43comments
  28. Alibaba open-sources AI model that can detect cancer and nearly 150 conditions(scmp.com ↗)
    11comments
  29. How SpaceX streamlined the Raptor engine(construction-physics.com ↗)
    68comments
  30. The Implications of Linguistic Illegibility for LLM Security(arxiv.org ↗)
    26comments

A Group Is Its Own Worst Enemy

25 pointsby 15y agocodinghorror.com
8 comments
15y agoHN ↗

I'm curious: is there any site where:

1) logged in users can vote for what they like, both up and down. A profile is built identifying other users who vote like you, and so your vote influences the post prominence for those others more than for other users who have different voting patterns

2) anonymous users get post prominence based on a smattering of different group votes

Its a bit like movie recommendations. In this way, trolls will quickly be grouped with other trolls and those who consistently vote up articles they like will see post prominence from like-minded devotees, and a single board can have a wide range of subjects that end up not needing categorization because they are self-organising.

(Would this work for, say, HN?)

15y agoHN ↗

I think (1) roughly describes HubSki, but that's going from descriptions I've heard rather than personal experience.

15y agoHN ↗

It's not too far off. On Hubski, the content you see is based on who you follow (people and/or tags). Importantly, you don't just get the posts of the people you follow, but the posts they share with you. (you might be interested in following someone that doesn't post at all, if they find and share quality posts). Likewise, posts you share go to your followers. -You can share and unshare.

My thinking on Hubski has been guided by the idea that groups should form and dissolve as people wish, and that Hubski become a multitude of overlapping and ever-changing groups. As user following-relationships are color-coded (http://hubski.com/about), you tend to build a familiarity with a number of people that are connected with you this way.

IMO it's not just content that should be the result of choice, but also how familiar you want to get with members of the community. I'm currently doing more work to this end.

In a nutshell, I feel effort is best spent enabling users to build and break boundaries and groups as they are wont to. -It's not as important that you behave as I would like you to, if I can easily avoid you.

15y agoHN ↗

I wonder about the implications of this WRT free speech in general. While ostensibly, free speech is a pillar democratic societies, not all forms of expression are actually allowed (think: porn or public nudity).

This implies that free speech is less about the content, but about the freedom to assemble and exchange information. That is, freedom manifests itself in self-selected groups of peers, who can then moderate the content of their communities.

"Free speech" is not a guarantee for a platform for any particular message, but rather a promise (by political powers) not to suppress free private exchanges of information; while public spaces (streets, squares, airwaves) may still be moderated.

15y agoHN ↗

With respect to free speech and, in general, freedoms within a group, it's important to realize we're dealing with two boundaries. The first is the rules the group agrees on that are enforced. The second are your own personal boundaries that you choose.

I think the tendency is to make the enforced rules light; we like to tip the balance in the favor of more freedom as much as possible. That said, this works best if individuals make an effort to be genuinely concerned about the interests of others in the group and therefore draw much more conservative boundaries around their own behavior.

In other words, groups need lots of social restrictions on behavior to work well, but it works best when each individual has a lot of room to decide what those boundaries are. Only the egregiously bad behavior is formally addressed as a group.

Sometimes people don't realize there are two boundaries going on and rationalize rude behavior under the umbrella of freedom or free speech.

15y agoHN ↗

Kuro5hin - brings back memories. Rusty released the source code as Scoop and I hacked on its Perl for a while.

Scoop was the original code base for The Daily Kos if I remember rightly. I took a swatch at the sources but no signs, its probably been boiled down to bones now.