Attached: 4 images
As many of you know, I posted recently about my experiences and outlook on Kagi, the paid search engine. It's gotten some positive press recently, ironically right after I made my blog post about why I no longer liked or trusted it. This blog post was called "Why I Lost Faith In Kagi" and was a pretty simple quick collection of my thoughts that I primarily wrote so it'd be easier to find again later to link to people when discussing Kagi versus making it a fedi thread I couldn't search for easily later. Across the four social media platforms I linked this blog post on, I'd say it got a total of about 40 likes and few reblogs.
https://d-shoot.net/kagi.html
I say this because this morning I woke up to an email from Kagi's CEO, Vlad, who had seen the post and was upset about it. I have an email address listed on my blog (which is why I didn't bother removing it from these logs), which is what he sent his emails to. I am posting this entire email chain in this thread and will briefly post my thoughts about it, but I feel like it's something that needs to be seen. Please take note of the subject of the email as well (EDIT: It got cropped out sorry, the subject is "Fatih [sic] can not be lost"). Also, since the alt text would get extremely long with some of the transcripts, I've provided a text dump of the emails here for screen reader users and will offer a more abridged description in the alt text: https://d-shoot.net/files/kagiemails.txt
the other poster was most likely trying to stir up shit given we’re on a federated network (see also their other post), but I have been obsessed with the idea of a federated search engine this weekend
it’s definitely a worthwhile idea, but anything that does federated indexing will have to deal with moderation and abuse prevention first. those are hard problems in ordinary federated networks, but search engines specifically are targets of hostile SEO and spam in ways that few other systems are. those aren’t insurmountable problems though — just ones that need to be solved before federation is practical.
I dunno. Google is currently overwhelmed by SEO optimized crap (at least I guess so, I stopped using it years ago because of that) and I think a federated system where members have different prioritizations would make SEO optimization very very hard to abuse.
I wasn’t trying to stir shit up. You still haven’t explained how a federated search sounds like a good idea. Just more jargon about the implementation. Running a search from some guy’s shitty laptop or cheap vm sounds horrendous.
Again, what is the point? Do you expect every instance to give different search results or specialize in one type of search? Federation made sense for social media, how does it make sense for search?
Because what are you gaining from a federated search? It makes sense for social media where different instances can specialize in their own topics. How does it make sense for search? What is the benefit of federating a search engine? Will each instance specialize in some specific type of search?
If I run a search are you going to query the federated network? If not then what is the point of federating? If so, how are you going to make it perform and how will you make search results consistent and reproducible?
So far I’ve seen no explanation for how end users benefit from a federated search. It’s just a buzzword that techbros like because of the fediverse.
Because then the results would no longer be modified / pushed by the CEO du jour. It would truly give me honest and reliable results. That’s why. See Reddit as an example in social media, but the point stands.
Why terrible?
you’re right
the other poster was most likely trying to stir up shit given we’re on a federated network (see also their other post), but I have been obsessed with the idea of a federated search engine this weekend
it’s definitely a worthwhile idea, but anything that does federated indexing will have to deal with moderation and abuse prevention first. those are hard problems in ordinary federated networks, but search engines specifically are targets of hostile SEO and spam in ways that few other systems are. those aren’t insurmountable problems though — just ones that need to be solved before federation is practical.
I dunno. Google is currently overwhelmed by SEO optimized crap (at least I guess so, I stopped using it years ago because of that) and I think a federated system where members have different prioritizations would make SEO optimization very very hard to abuse.
I hope so! this is something I’d like to implement, once I’ve got some free deployment resources and time.
I wasn’t trying to stir shit up. You still haven’t explained how a federated search sounds like a good idea. Just more jargon about the implementation. Running a search from some guy’s shitty laptop or cheap vm sounds horrendous.
Again, what is the point? Do you expect every instance to give different search results or specialize in one type of search? Federation made sense for social media, how does it make sense for search?
Because what are you gaining from a federated search? It makes sense for social media where different instances can specialize in their own topics. How does it make sense for search? What is the benefit of federating a search engine? Will each instance specialize in some specific type of search?
If I run a search are you going to query the federated network? If not then what is the point of federating? If so, how are you going to make it perform and how will you make search results consistent and reproducible?
So far I’ve seen no explanation for how end users benefit from a federated search. It’s just a buzzword that techbros like because of the fediverse.
Because then the results would no longer be modified / pushed by the CEO du jour. It would truly give me honest and reliable results. That’s why. See Reddit as an example in social media, but the point stands.