6 ms·

It would be a lot easier to get around vote manipulation detection if the code was public.

FeatureRush 10y ago
Crypto industry is a step ahead in this game, they would release white paper proving that even the creators of the system can't manipulate upvotes.
- dragonwriter 10y ago
  This isn't crypto, and there's no concrete definition of manipulation that would support proof (unlike the kind of threats crypto deals with); it's a fuzzy problem domain.
krapp 10y ago
If that's true, it implies that vote manipulation detection is already easy to get around. It's just relying on security through obscurity and probably direct human intervention to cover for unpublished flaws.
- dang 10y ago
  Aren't you holding us to an impossible standard here? The problems we're talking about are not cryptographic; they're much closer to, say, law enforcement. Has anybody figured out how to model such things with mathematical guarantees?
  - krapp 10y ago
    It's just depressing that a site that calls itself Hacker News, and that centers around discussions about software, web applications and security, by so many intelligent and capable people, is less open and less hacker friendly than Wordpress. And I understand why - the problem is that votes mean something to startups that post here, they have a potential real money value. But it's unfortunate that any insight into how the HN staff deals with the problem of being gamed is denied to the community at large, out of fear of being gamed even more. Plenty of people here would be willing to learn Arc, audit and contribute to the codebase if you let them, and that scrutiny could be beneficial, not just harmful.
    - i336_ 10y ago
      The idealism here is really noteworthy, and (as just another commentator/user here) I must say I like it very much. As you yourself note, however, I can't see any possible way to make what you're describing practically work out: there is absolutely NO way to tell the good guys from the "growth hackers". :( The other problem is that the bunch of people willing to do this work - which (I say this non-critically, please note!!) is similar to IRL work for a community organization or similar, in that it's not attractive or "widely scoped shiny impact that looks awesome on paper" - is much, much smaller than the bunch of people with evil grins just waiting for more tidbits to use against everyone else.
yellowapple 10y ago
Security by obscurity is not obscurity. It would be a lot easier to detect and fix flaws in the vote manipulation detection if the code was public.

Crypto industry is a step ahead in this game, they would release white paper proving that even the creators of the system can't manipulate upvotes.

dragonwriter 10y ago
This isn't crypto, and there's no concrete definition of manipulation that would support proof (unlike the kind of threats crypto deals with); it's a fuzzy problem domain.

krapp 10y ago

If that's true, it implies that vote manipulation detection is already easy to get around. It's just relying on security through obscurity and probably direct human intervention to cover for unpublished flaws.

dang 10y ago
Aren't you holding us to an impossible standard here? The problems we're talking about are not cryptographic; they're much closer to, say, law enforcement. Has anybody figured out how to model such things with mathematical guarantees?
- krapp 10y ago
  It's just depressing that a site that calls itself Hacker News, and that centers around discussions about software, web applications and security, by so many intelligent and capable people, is less open and less hacker friendly than Wordpress. And I understand why - the problem is that votes mean something to startups that post here, they have a potential real money value. But it's unfortunate that any insight into how the HN staff deals with the problem of being gamed is denied to the community at large, out of fear of being gamed even more. Plenty of people here would be willing to learn Arc, audit and contribute to the codebase if you let them, and that scrutiny could be beneficial, not just harmful.
  - i336_ 10y ago
    The idealism here is really noteworthy, and (as just another commentator/user here) I must say I like it very much. As you yourself note, however, I can't see any possible way to make what you're describing practically work out: there is absolutely NO way to tell the good guys from the "growth hackers". :( The other problem is that the bunch of people willing to do this work - which (I say this non-critically, please note!!) is similar to IRL work for a community organization or similar, in that it's not attractive or "widely scoped shiny impact that looks awesome on paper" - is much, much smaller than the bunch of people with evil grins just waiting for more tidbits to use against everyone else.

yellowapple 10y ago

Security by obscurity is not obscurity. It would be a lot easier to detect and fix flaws in the vote manipulation detection if the code was public.