There’s also an ongoing project called Yacy which is trying to make a decentralized ever-growing index, but it’s small enough right now that the results are still quite poor even when federating with other users until you spend enough time fine-tuning it, which most people don’t want or have time to do.
That’s neat - though I would worry that it could get polluted easily, e.g. China, Russia, and/or fascists everywhere would very much like to control the conversation, so as it got a wider userbase that would be the time for it to cease functioning, whereas before that it could be allowed as something to occupy our time. So much of our daily lives are impacted by such geopolitics that we don’t even/often think of.
I hope that we (people writing FOSS) can explore the concept of voting more - e.g. how wikipedia does its edits with “trust actors”, similarly the Fediverse (& searching) could have a much wider pool of trusted community mods (or potentially a lesser category of that, lacking full post-removal capability as current ones do) where mods could vote and once something went below a certain threshold (e.g. 5x more down- than up-votes from such community curators), then a flag could go up like “this post has been marked as containing potential misinformation - are you sure that you would like to read it?” By distributing the load like that, it could help make this place MUCH more active, by lowering the barrier to moderation as it conjoins normal reading activities with a very simple button bush for most people. (and then a higher category of mods can double-check the curator mods, etc.)
On the other hand, authoritarian developers are unlikely to want to extend the Lemmy code along those lines, and rather go the opposite direction so that anytime someone says something even the slightest bit against their party line, boop the person becomes insta-banned in every single community that they have ever interacted in, even if never having commented but solely voted in them (I am not exaggerating this up - for one see the notes for the upcoming v0.19.4 release allowing this capability, and multiple recent threads discussing how this has already happened to numerous people - e.g. here is one excellent accounting of that process).
So anyway, if someone is willing to build a good system, then the people do seem willing to take it forward - e.g. see how many have already done so to Google Maps with all the reviews & pics including menus & such to share with everyone. But ofc there will be resistance to now doing that again for somewhere else, and perhaps still yet again if that one fails, and so on in perpetuity.:-(
Seems like you’re trying to address systemic problems of learned human nature at this point, which is something that can’t really be done through software alone. There needs to be a fundamental culture shift for that. Yacy is mostly good because it’s self-hosted with the option to federate to a decentralized index, and so is able to be tailored to your own needs without any external entity. However, the tailoring process means that the initial search results are absolutely atrocious compared even to Google at first, and then get much better over time. SearXNG, on the other hand, is a sort of self-hosted aggregator of search results from dozens of other search engines, rather than being a search engine in itself. It also serves as a proxy, since it is doing the searching and not you.
SearXNG does pretty well.
There’s also an ongoing project called Yacy which is trying to make a decentralized ever-growing index, but it’s small enough right now that the results are still quite poor even when federating with other users until you spend enough time fine-tuning it, which most people don’t want or have time to do.
That’s neat - though I would worry that it could get polluted easily, e.g. China, Russia, and/or fascists everywhere would very much like to control the conversation, so as it got a wider userbase that would be the time for it to cease functioning, whereas before that it could be allowed as something to occupy our time. So much of our daily lives are impacted by such geopolitics that we don’t even/often think of.
I hope that we (people writing FOSS) can explore the concept of voting more - e.g. how wikipedia does its edits with “trust actors”, similarly the Fediverse (& searching) could have a much wider pool of trusted community mods (or potentially a lesser category of that, lacking full post-removal capability as current ones do) where mods could vote and once something went below a certain threshold (e.g. 5x more down- than up-votes from such community curators), then a flag could go up like “this post has been marked as containing potential misinformation - are you sure that you would like to read it?” By distributing the load like that, it could help make this place MUCH more active, by lowering the barrier to moderation as it conjoins normal reading activities with a very simple button bush for most people. (and then a higher category of mods can double-check the curator mods, etc.)
On the other hand, authoritarian developers are unlikely to want to extend the Lemmy code along those lines, and rather go the opposite direction so that anytime someone says something even the slightest bit against their party line, boop the person becomes insta-banned in every single community that they have ever interacted in, even if never having commented but solely voted in them (I am not exaggerating this up - for one see the notes for the upcoming v0.19.4 release allowing this capability, and multiple recent threads discussing how this has already happened to numerous people - e.g. here is one excellent accounting of that process).
So anyway, if someone is willing to build a good system, then the people do seem willing to take it forward - e.g. see how many have already done so to Google Maps with all the reviews & pics including menus & such to share with everyone. But ofc there will be resistance to now doing that again for somewhere else, and perhaps still yet again if that one fails, and so on in perpetuity.:-(
Seems like you’re trying to address systemic problems of learned human nature at this point, which is something that can’t really be done through software alone. There needs to be a fundamental culture shift for that. Yacy is mostly good because it’s self-hosted with the option to federate to a decentralized index, and so is able to be tailored to your own needs without any external entity. However, the tailoring process means that the initial search results are absolutely atrocious compared even to Google at first, and then get much better over time. SearXNG, on the other hand, is a sort of self-hosted aggregator of search results from dozens of other search engines, rather than being a search engine in itself. It also serves as a proxy, since it is doing the searching and not you.
Complex problems will not be solved overnight, it’s true, but we can get there one step at a time!:-)