5 ms·

Any reason why HN source code is not published? The best I can think of is not to let people see the penalizing behavior, but having an open standard might actu

Any reason why HN source code is not published? The best I can think of is not to let people see the penalizing behavior, but having an open standard might actually help it improve rather than keeping it as a hidden secret that slowly gets discovered by independent malicious parties.

capableweb 3y ago
Here you can find the latest public code: http://arclanguage.org/ http://arclanguage.org/ But I don't think there has been any public updates since that release (edit: updates from YC. There is a community-led project that continues to improve the codebase after that release)
shawnz 3y ago
It shares the same lineage as this code: https://github.com/arclanguage/anarki/tree/master/apps/news https://github.com/arclanguage/anarki/tree/master/apps/news
- krapp 3y ago
  Anarki and its version of the forum is so far removed from the original Arc base that it isn't even worth comparing at this point.
dun44 3y ago
>The best I can think of is not to let people see the penalizing behavior I wonder if it would be possible to split HN (or any other HN-like site) into two halves: one would handle the actual storage of posts and comments, and another would handle the scoring, ie the order the comment tree branches appear in. Then we could have multiple of the latter, competing with each other on the scoring alone, with the actual content being common.
- solardev 3y ago
  Kinda like how reddit has "hot", "top", and "best"?
  - dun44 3y ago
    More like with Usenet where every user would have their own scorefile, which determined the order they would see posts. Except without having every user to manage it for themselves, instead choosing which "frontend" to use.
- jedberg 3y ago
  The reddit open source code was basically this. The whole thing was open, including the scoring, but all the spam functions were just stubs that returned True or False as appropriate, without all the logic. Then we would override those functions with a private library that had all the spam logic. I find your idea interesting about having a service that stores the links and comments (and presumably does spam controls too) but anyone can put their own scoring/sorting algo on it.
  - defrost 3y ago
    There's an interesting less granular pooled approach worth exploring .. seeing the aggregate scoring of {people you follow} | {friends} | {Dev community} | {carpentry fans}. It's the "crowd I find of interest" also found "this stuff to be interesting".
    - dun44 3y ago
      Except in this case you still don’t have any accountability, so the site (eg HN) can still shape the narrative by slightly altering the scoring.
- krapp 3y ago
  Hacker News has an API[0], awkward as it is (but not rate-limited) and Algolia does as well[1]. It isn't difficult to build a basic reader client. [0] https://github.com/HackerNews/API https://github.com/HackerNews/API [1] https://hn.algolia.com/api https://hn.algolia.com/api
- capableweb 3y ago
  Funnily enough, you're basically describing Bluesky/AT Proto: https://atproto.com/guides/overview https://atproto.com/guides/overview
tptacek 3y ago
There's probably not all that much to see past the decade-old public release from the arc project. The parts of it that aren't public past that are non-public by necessity, like the voting ring detector stuff.
aftbit 3y ago
I would guess it's part of the ethos of not spending more time and effort on HN than strictly needed. Perhaps also a bit of the standard silicon valley culture of building on open source but not actually contributing anything meaningful back.
rollcat 3y ago
> The best I can think of is not to let people see the penalizing behavior, but having an open standard might actually help it improve rather than keeping it as a hidden secret that slowly gets discovered by independent malicious parties. Good point. You can see it as a part of a "defense in depth" strategy. Ideally you want the system's security architecture to stand it's ground, even if everything (except for the private keys) gets leaked. In practice you'll conceal some of these details to gain an extra edge. Lisp famously allows your software to be extremely dynamic, so I don't see why such functionality shouldn't be implemented as a closed plugin, while the core of the site would remain open to public contributions. And boy does this website need a couple fixes... (dark mode? mobile CSS? large comment thread performance? dupe link detection?)

capableweb 3y ago

Here you can find the latest public code: http://arclanguage.org/ http://arclanguage.org/ But I don't think there has been any public updates since that release (edit: updates from YC. There is a community-led project that continues to improve the codebase after that release)

shawnz 3y ago

It shares the same lineage as this code: https://github.com/arclanguage/anarki/tree/master/apps/news https://github.com/arclanguage/anarki/tree/master/apps/news

krapp 3y ago
Anarki and its version of the forum is so far removed from the original Arc base that it isn't even worth comparing at this point.

dun44 3y ago

>The best I can think of is not to let people see the penalizing behavior I wonder if it would be possible to split HN (or any other HN-like site) into two halves: one would handle the actual storage of posts and comments, and another would handle the scoring, ie the order the comment tree branches appear in. Then we could have multiple of the latter, competing with each other on the scoring alone, with the actual content being common.

solardev 3y ago
Kinda like how reddit has "hot", "top", and "best"?
- dun44 3y ago
  More like with Usenet where every user would have their own scorefile, which determined the order they would see posts. Except without having every user to manage it for themselves, instead choosing which "frontend" to use.
jedberg 3y ago
The reddit open source code was basically this. The whole thing was open, including the scoring, but all the spam functions were just stubs that returned True or False as appropriate, without all the logic. Then we would override those functions with a private library that had all the spam logic. I find your idea interesting about having a service that stores the links and comments (and presumably does spam controls too) but anyone can put their own scoring/sorting algo on it.
- defrost 3y ago
  There's an interesting less granular pooled approach worth exploring .. seeing the aggregate scoring of {people you follow} | {friends} | {Dev community} | {carpentry fans}. It's the "crowd I find of interest" also found "this stuff to be interesting".
  - dun44 3y ago
    Except in this case you still don’t have any accountability, so the site (eg HN) can still shape the narrative by slightly altering the scoring.
krapp 3y ago
Hacker News has an API[0], awkward as it is (but not rate-limited) and Algolia does as well[1]. It isn't difficult to build a basic reader client. [0] https://github.com/HackerNews/API https://github.com/HackerNews/API [1] https://hn.algolia.com/api https://hn.algolia.com/api
capableweb 3y ago
Funnily enough, you're basically describing Bluesky/AT Proto: https://atproto.com/guides/overview https://atproto.com/guides/overview

tptacek 3y ago

There's probably not all that much to see past the decade-old public release from the arc project. The parts of it that aren't public past that are non-public by necessity, like the voting ring detector stuff.

aftbit 3y ago

I would guess it's part of the ethos of not spending more time and effort on HN than strictly needed. Perhaps also a bit of the standard silicon valley culture of building on open source but not actually contributing anything meaningful back.

rollcat 3y ago

> The best I can think of is not to let people see the penalizing behavior, but having an open standard might actually help it improve rather than keeping it as a hidden secret that slowly gets discovered by independent malicious parties. Good point. You can see it as a part of a "defense in depth" strategy. Ideally you want the system's security architecture to stand it's ground, even if everything (except for the private keys) gets leaked. In practice you'll conceal some of these details to gain an extra edge. Lisp famously allows your software to be extremely dynamic, so I don't see why such functionality shouldn't be implemented as a closed plugin, while the core of the site would remain open to public contributions. And boy does this website need a couple fixes... (dark mode? mobile CSS? large comment thread performance? dupe link detection?)