SpacemanSpiff

I use Kagi too, it’s surprisingly snappy! Like seriously impressive for a small org. They talk about speed optimisation being critical for them as well. I find the result to be excellent as well. A true Google replacement/feels like Google in its prime.

I believe they have their own index and bot as well?

Yes….on a technical level. But the picture is bigger than that. Personally, I have a hunch that the choice of Rust is making Lemmy’s development slower. This seemed to be evidenced by the fact that Kbin has more functionality than Lemmy while having only been around for 2 months. Vs Lemmy’s 4 years. The Kbin dev has also been much more able to fix things on the fly during the surge in users. Whereas Lemmy will supposedly move off websocket use any day now.

Adoptability isn’t something to be discounted. The fact that there any more people out there familiar with PHP may give Kbin an edge over time. And let’s be honest, in real-world test PHP can very often be faster then - less-than-mature-Rust codebase.

Kind of like how Lemmy made up communities? Lol you don’t call microblogs “toots” either. Seems every fediverse software has their own terminology.

They do have their own crawler and I believe supplement from bing only when necessary. Something like 98% of results are from their own index across all users. They actually have a breakdown for every search.

I think they actually have the largest independent index outside of Google/Bing right now.

Others that exclusively use their own index are:

Mojeek (results aren’t great)
Kagi (result are pretty good imo)

I understand your idea, but I think it would defeat the purpose of the fediverse. It would create single points of failure that are un-correctable.

I also think many people forget that Reddit never functioned any differently. Everyone seems to have forgotten (and I’m not saying you have!) that there are and were always multiple subreddits for any given topic. With slightly differing names. The only reason people are forgetting this is because eventually one or a handful became pre-eminent and the others died or became transformed into something more niche.

I think it’s a problem that will ultimately correct itself, but I think a tags based system, like hashtags in Mastodon, would be a better solution for tying communities/magazines together through metadata.

I think it’s a fair point. They won’t be able to remain federated to many instances if their point of contention is open-enrollment.

I understand needing the Lemmy moderation tools to improve and that it’s temporary, but the damage to their own communities and users may not be temporary.

Their users will turn inward and end up preferring their own communities—which is fine. However it also means that non-beehaw users will shy away from those communities in favour of others, lest their home site get de-federated at some point for the same reasons. These effects combined means slow-to-grow, low-visibility communities in the fediverse, and increases the chance that their communities may dwindle if others of the same subject become pre-eminent outside of Beehaw.

In short, while I understand their reasons, I think that it risks making Beehaw.org permanently insular and ultimately much more similar to a non-fediverse website.

I think Google peaked about 6-8 years ago now and then started slipping at an ever accelerating rate.

It’s almost useless for me when searching anything remotely technical or otherwise niche.

I almost consistently need to go to the second page of results now, something I don’t remember doing since like 2009.

I find Bing acceptable. Brave search works well. But I’m actually using Kagi now since I’m hoping their paid model will actually mean I’m not the product.

Via the ActivityPub protocol that fediverse software uses :)

Pretty cool stuff!

Kbin can also directly interact with Mastodon users and toots because of this. Kbin magazines can natively contain both “threads” from Kbin and Lemmy, and microblog “posts” from Kbin and Mastodon. (And other software depending how they map these features.)

I think the “dynamic updates” behaviour is tied to Lemmy’s use of websockets instead of http. Kbin uses http. The Lemmy devs have stated they’re going to move off of websockets in the future as they present scaling issues with the way the software is written.

The websocket protocol allows bi-directional push communication regardless of the previous request which means that new posts are constantly triggering server side updates which then appear like a page “refresh” on clients.

Arguably, while websockets have very cool realtime features compared to http, for a Reddit-like content aggregate their use can quickly overwhelm usability without significant retooling.

It depends on what you mean by other servers…Lemmy instances? I haven’t seen any differences between kbin instances.

Kbin is entirely different software from Lemmy; it’s a completely different backend.

So that would be why it’s “better” in that regard than Lemmy instances. My understanding is that it’s a bug in Lemmy that will be fixed soon.

The “popping” of the feed in Lemmy is tied to their use of websockets instead of http. I believe the devs stated they’ll be moving off of websockets in the future.

This must be a Lemmy thing right? I haven’t experienced this at all on kbin.

Thank you, I think you’re right! I misread it for sure.

Small correction. Kbin is not a Lemmy instance. It’s an entirely different backend on the ActivityPub protocol. It federates with “threaded” fediverse sites like Lemmy, and with microblogging sites like Mastodon. Users on kbin can create and interact with both kinds of posts and kbin magazines “Lemmy communities” can natively include both.

Edit: sorry @atypicaloddity I totally misread your comment. Ignore me!