4 ms·

I don’t even really understand why things have to be notable for Wikipedia. If people wanted to maintain all of the Runescape articles in Wikipedia (as mentione

I don’t even really understand why things have to be notable for Wikipedia. If people wanted to maintain all of the Runescape articles in Wikipedia (as mentioned by nouser76)… why not? Why not have Wikipedia literally be the central place for all information for all these different things?

soulofmischief 2y ago
The most lazy approach possible to solve the issue that low-traffic and low-interest data becomes increasingly low-confidence as either no one is monitoring for changes, or due to information wars between parties with conflicting interest. I've personally been approached to build tech to monitor and revert changes matching keywords using residential proxy systems as a service for low-stakes clients (and declined).
- lolinder 2y ago
  > The most lazy approach possible to solve the issue that low-traffic and low-interest data becomes increasingly low-confidence Can you suggest some alternate approaches that you think would actually work? How would you approach removing the notability requirement without immediately opening the floodgates to hundreds of thousands of additional sketchy articles about niche topics that don't have enough interest to be vetted by more than one person per year?
  - soulofmischief 2y ago
    I've been advocating this for years, but I think a lot of problems on the internet could be solved with a universal, decentralized, anonymous reputation system. A user or entity maintain a private cryptographic key. Each service is given a different sub-key which anonymizes the entity for each service, if they choose. A myriad of voting systems could be employed depending on the needs of each service, via other site users, site admins or both. These differing voting systems (stars, points, etc) could be normalized through some standardized process into a final reputation score. Other services could then either consider the aggregate score, or only focus on ratings from particular other services, which could be grouped into service types for convenience. Ratings are cryptographically backed and could be fuzzed to increase anonymity. Scores would also be weighted by the relative reputation of other entities. Participating services would also themselves need to maintain a high reputation for their scoring to have much weight in the aggregate. Services could use these scores to wall of certain site features to entities without sufficient reputation, or even wholesale deny read/write access. Obviously there are particulars to work out to prevent brigading, etc. but there is a balance to strike such that farmed accounts could still quickly be punished the moment they begin spamming/astroturfing. Perhaps a flag/vouch system similar to HN could be employed. Voting rings could also be detected, though with limits given that each vote comes from an anonymized sub-key. This system would even protect against a future where the majority of content on the web will likely be aided or entirely produced by LLMs or AI. I welcome such synthetic content, but only if it matches or surpasses the general quality of human posts on the internet (which is much worse than one might think). On top of this, users could themselves use these scores as a way to filter out content which is signed with a reputation score below some threshold. I imagine some kind of hierarchical web, where some services are harder to gain write access to than others, allowing users to gain reputation in low-stakes environments with some level of decentralized moderation, but without compromising the integrity of higher-stakes information. Each piece of information on the internet should be signed with some sort of introspectable reputation score, and this kind of system would make it possible without compromising privacy and security of users in any political environment.
- johannes1234321 2y ago
  I would argue that quite some "fandom" content gathers more views than many "notable" pages on Wikipedia. That said, I think there is a distinction between fact and made up.
stevage 2y ago
There are many places on Wikipedia that explain exactly why. The short answer is, because it costs time and effort, and a little bit of money, to maintain all those articles, and they cause more problems than they are worth.
- Gormo 2y ago
  What does it mean to "maintain" an article as something distinct from posting and updating it?
  - rightbyte 2y ago
    It would be quite bloated if Wikipedia had 1000 Runescape articles about different minutia I guess. But I'd like if Wikipedia had 'subwikis' with that kind of stuff.
  - stevage 2y ago
    Amongst other things, looking out for vandalism and misinformation. Also managing categories, infoboxes etc as they get updated. There's quite a lot. Just look at the history of any article.
i80and 2y ago
It's been frustrating watching articles for stuff I cared about in the 2000's get deleted because the mere passage or time has rendered the topics non-notable
- bpeebles 2y ago
  It's not always clear but notability is not temporary[0] under English Wikipedia's guidelines. Standards of what makes something notable have shifted over the years and sometimes the different is hard to tease out. 0. https://en.wikipedia.org/wiki/Wikipedia:Notability#Notability_is_not_temporary https://en.wikipedia.org/wiki/Wikipedia:Notability#Notabilit...
  - i80and 2y ago
    I understand this to be the case, but in practice the distinction is mostly immaterial. Topics are notable for a decade, then not notable the next.

soulofmischief 2y ago

The most lazy approach possible to solve the issue that low-traffic and low-interest data becomes increasingly low-confidence as either no one is monitoring for changes, or due to information wars between parties with conflicting interest. I've personally been approached to build tech to monitor and revert changes matching keywords using residential proxy systems as a service for low-stakes clients (and declined).

lolinder 2y ago
> The most lazy approach possible to solve the issue that low-traffic and low-interest data becomes increasingly low-confidence Can you suggest some alternate approaches that you think would actually work? How would you approach removing the notability requirement without immediately opening the floodgates to hundreds of thousands of additional sketchy articles about niche topics that don't have enough interest to be vetted by more than one person per year?
- soulofmischief 2y ago
  I've been advocating this for years, but I think a lot of problems on the internet could be solved with a universal, decentralized, anonymous reputation system. A user or entity maintain a private cryptographic key. Each service is given a different sub-key which anonymizes the entity for each service, if they choose. A myriad of voting systems could be employed depending on the needs of each service, via other site users, site admins or both. These differing voting systems (stars, points, etc) could be normalized through some standardized process into a final reputation score. Other services could then either consider the aggregate score, or only focus on ratings from particular other services, which could be grouped into service types for convenience. Ratings are cryptographically backed and could be fuzzed to increase anonymity. Scores would also be weighted by the relative reputation of other entities. Participating services would also themselves need to maintain a high reputation for their scoring to have much weight in the aggregate. Services could use these scores to wall of certain site features to entities without sufficient reputation, or even wholesale deny read/write access. Obviously there are particulars to work out to prevent brigading, etc. but there is a balance to strike such that farmed accounts could still quickly be punished the moment they begin spamming/astroturfing. Perhaps a flag/vouch system similar to HN could be employed. Voting rings could also be detected, though with limits given that each vote comes from an anonymized sub-key. This system would even protect against a future where the majority of content on the web will likely be aided or entirely produced by LLMs or AI. I welcome such synthetic content, but only if it matches or surpasses the general quality of human posts on the internet (which is much worse than one might think). On top of this, users could themselves use these scores as a way to filter out content which is signed with a reputation score below some threshold. I imagine some kind of hierarchical web, where some services are harder to gain write access to than others, allowing users to gain reputation in low-stakes environments with some level of decentralized moderation, but without compromising the integrity of higher-stakes information. Each piece of information on the internet should be signed with some sort of introspectable reputation score, and this kind of system would make it possible without compromising privacy and security of users in any political environment.
johannes1234321 2y ago
I would argue that quite some "fandom" content gathers more views than many "notable" pages on Wikipedia. That said, I think there is a distinction between fact and made up.

stevage 2y ago

There are many places on Wikipedia that explain exactly why. The short answer is, because it costs time and effort, and a little bit of money, to maintain all those articles, and they cause more problems than they are worth.

Gormo 2y ago
What does it mean to "maintain" an article as something distinct from posting and updating it?
- rightbyte 2y ago
  It would be quite bloated if Wikipedia had 1000 Runescape articles about different minutia I guess. But I'd like if Wikipedia had 'subwikis' with that kind of stuff.
- stevage 2y ago
  Amongst other things, looking out for vandalism and misinformation. Also managing categories, infoboxes etc as they get updated. There's quite a lot. Just look at the history of any article.

i80and 2y ago

It's been frustrating watching articles for stuff I cared about in the 2000's get deleted because the mere passage or time has rendered the topics non-notable

bpeebles 2y ago
It's not always clear but notability is not temporary[0] under English Wikipedia's guidelines. Standards of what makes something notable have shifted over the years and sometimes the different is hard to tease out. 0. https://en.wikipedia.org/wiki/Wikipedia:Notability#Notability_is_not_temporary https://en.wikipedia.org/wiki/Wikipedia:Notability#Notabilit...
- i80and 2y ago
  I understand this to be the case, but in practice the distinction is mostly immaterial. Topics are notable for a decade, then not notable the next.