Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. I built non-autoregressive decision models with RL a year ago(convaiinnovations.com ↗)
    175comments
  2. AI-generated posters don’t have to be horrible(john.hartnup.uk ↗)
    560comments
  3. Human brain is two separate organs, Stanford Medicine-led research finds(stanford.edu ↗)
    200comments
  4. Btrfs/ZFS/bcachefs under workloads classic benchmarks skip(bartosz.fenski.pl ↗)
    3comments
  5. A graphical desktop for the ZX Spectrum(github.com/mindbox77 ↗)
    85comments
  6. Tin: full-text search for Postgres(planetscale.com ↗)
    57comments
  7. “The Secret Life of Circuits” is here(coredump.cx ↗)
    59comments
  8. Android 17 is the first since 3.x to add new APIs without releasing to the AOSP(grapheneos.social ↗)
    598comments
  9. Show HN: CUA-S1 – A System One Model for Computer Use(github.com/trycua ↗)
    discuss
  10. Supabase (YC S20) Is Hiring for OrioleDB(supabase.link ↗)
    discuss
  11. Black Holes or Black Hole Stars? Astronomers Spar over 'Little Red Dots'(quantamagazine.org ↗)
    23comments
  12. New evidence for hidden chambers beyond Tutankhamun's tomb(nature.com ↗)
    13comments
  13. People who know the most often sound the least certain(vrash.substack.com ↗)
    9comments
  14. GPT-6 Astra Solves a WWI German Radio Cipher(prinzai.com ↗)
    146comments
  15. San Francisco Onion Futures Company(onionfutures.com ↗)
    133comments
  16. Suzanne Ciani's Buchla Cookbook(echo.orpheusinstituut.be ↗)
    discuss
  17. Almost Never Use AI to Write Anything Substantive(erichgrunewald.substack.com ↗)
    40comments
  18. Cloudflare Quick Tunnels(cloudflare.com ↗)
    302comments
  19. How to Write with an LLM(sockpuppet.org ↗)
    365comments
  20. If math is more than proof, we need to better celebrate the rest of it(terrytao.wordpress.com ↗)
    208comments
  21. What Zig felt like, coming from Rust(besok.github.io ↗)
    150comments
  22. You can run Git on object storage if you re-make packfiles(tigrisdata.com ↗)
    29comments
  23. Saving another 100TB of RAM(cloudflare.com ↗)
    93comments
  24. Communication by means of modulated Johnson noise(pnas.org ↗)
    23comments
  25. Asking Authors About Their Own Papers(medium.com/tmlrorg ↗)
    51comments
  26. How OpenAI Used Its Own LLMs to Design Its Jalapeño Chip(ieee.org ↗)
    123comments
  27. SDCC – Small Device C Compiler(sourceforge.net ↗)
    26comments
  28. Adventures in Microcontroller Circuit Debugging(bigmessowires.com ↗)
    1comments
  29. Science Is Open Software(jepedersen.dk ↗)
    57comments
  30. Ray Ozzie and the Optimism of Being Early(reproof.app ↗)
    20comments

Ask HN: Dark Patterns on HN

4 pointsby 6y ago
6 comments
Maybe I am biased, but increasingly feel HN is subject to number of dark patterns related to potentially meritless rep-mining, voting, flags, etc.

What are examples if any that users are seeing and how might they be better addressed?

Would HN consider being more open about their policies or allowing trusted 3rd party to audit the effectiveness of these policies?

6y agoHN ↗

You seem to have asserted there is meritless rep-mining and provided no details. Then you ask others to provide details about where they see meritless rep-mining. I know you already know about this repository documenting HN

https://github.com/minimaxir/hacker-news-undocumented/blob/m...

So I'm really not sure what more you are after with respect to how things work.

Why would HN be interested in having a 3rd party audit their policies? What possible positive outcome would be in it for them?

6y agoHN ↗

Right, I am a HN super user.

I repeatedly get accounts to 500, then move onto a new account; no reason other than by doing so I am never able to downvote regardless of the desire to do so, let’s me easily benchmark perception of my comments & excluded community’s bias to upvote known users, it’s a minor form of OpSec, etc.

Here are a few examples:

- HN automates a lot of mod related tasks, but to my knowledge rarely prior to establishing such automation defines a measurable outcome, then tests to confirm the desired outcome holds true;

- Recently discovered user that’s been shadowban 5 years will likely never know they are ghosted unless I tell them;

- Other examples which are likely considered “secrets” that might not be appropriate openly disclose; I personally object to this;

- as for the audits, it would allow HN to get an independent opinion in a way that doesn’t result in data they normally withhold from being public.

Happy to address or clarify anything on request.

6y agoHN ↗

Right, I am a HN super user. I repeatedly get accounts to 500, then move onto a new account

Please don't. For HN to be a community, users need some identity for other users to relate to. If you change your identity routinely, you're interrupting that functioning. Otherwise we may as well have no usernames and no community, and that would be a different kind of forum. https://hn.algolia.com/?sort=byDate&dateRange=all&type=comme.... This is in the site guidelines: https://news.ycombinator.com/newsguidelines.html.

If you don't want to be able to downvote, I'm sure we could figure out a way to make that work.

6y agoHN ↗

Okay, understand; unfortunately that means I will just stop posting on HN, since I neither have the desire to have “an identity” on HN, nor do I intend to disregard any guidelines. Thanks for all the work you do to keep HN running, please don’t take this personally, understand and agree with your position, just not good fit for me. Cheers!

6y agoHN ↗

I would be interested in a way to disable downvotes from affecting the comment ranking.

6y agoHN ↗

as for the audits, it would allow HN to get an independent opinion in a way that doesn’t result in data they normally withhold from being public.

While I see obtaining feedback on user behavior and satisfaction valuable, I'm still unsure of how a public audit would benefit HN. It might already be the case that HN conducts private audits and we never know.

I don't really feel that HN has a community or reputation problem in the way that I feel Stack Overflow does so without understanding what you see as the motivation for a public review, I'm not sure what HN or the community would get out of the results.

One thing I would say is that I personally think "shadow bans" are a little petty. I don't agree with the idea of making the site artificially slow and making participation "seem" to work when it does not. That said, I accept that this is the way HN has chosen to work.