this post was submitted on 19 Nov 2026
15 points (89.5% liked)

Fediverse

45019 readers
161 users here now

A community to talk about the Fediverse and all it's related services using ActivityPub (Mastodon, Lemmy, Mbin, etc).

If you wanted to get help with moderating your own community then head over to !moderators@lemmy.world!

Rules

Learn more at these websites: Join The Fediverse Wiki, Fediverse.info, Wikipedia Page, The Federation Info (Stats), FediDB (Stats), Sub Rehab (Reddit Migration)

founded 3 years ago
MODERATORS
 

Here is a brief concept: It is the federated/fediverse Web search engine (this may have been based on a older post i made)

But how is it federated, that part i hope i can explain it well,

How It May Work:

1: Instead of a crawler, it uses the older method of A Hand curated database of links and websites (By the instance mods), And no it is not like a lemmy link aggregation community, it is designed to replace DuckDuckGo, Google, Startpage, Kagi and more,

1.1: it is that the server admins add a entry to a database with that entry having a Link and a Name and other properties, which turns into search result, you see, and the database of Entries federate so you have more entries

2: The instances federate, so you have more possible search results

3: you can upvote or downvote specific search results (so if a page was SEO spam you get less of it)

4: it has integration with the rest of the fediverse

A Why:

Web Index's, And Crawlers, could be locked down so privacy first search engines would be locked down and have issues

which would cause panic, unlike social media, you cannot easily have an alternative

A Notice:

THIS CONCEPT HAS NO AI INVOLVED IN ITS MAKING WHATSOEVER

i hope you understood the concept

Tags For Mastodonians:

#technology #fediverse #foss #opensource #freesoftware #ideas #concepts

Edits To The Concept:

1: A Karma System for instances that get a lot of downvotes in their links

2: a Vote log seen by admins with tools to help detect vote manipulation

3: A Rate limiter for downvoting and upvoting

top 28 comments
sorted by: hot top controversial new old
[–] unitedwithme@lemmy.today 6 points 1 day ago (1 children)

Honest question: who tf has time to hand-curate results over a crawler or index?? That's insane! I would host, except that's a lot of work.

Maybe a way to query search results quantities to then itemize your results? Idk. But, that's just asking for a bad result, my $.02.

[–] Adniem@thelemmy.club 1 points 1 day ago

well i think that's counting the entire web which nobody does visit the entire web

but maybe i do some ways to bulk add (such as probably Regex)

and if i think you misunderstood it isn't you curate the results for "Dogs" but you just do make entries for example an image of a dog and you do add some proprieties, and the search engine tries to find it when you search "Dogs"

[–] gedaliyah@lemmy.world 5 points 1 day ago (2 children)

Are you suggesting that users manually add websites and keywords to a database?

[–] Adniem@thelemmy.club 4 points 1 day ago

yes but with some automation (such as Regex and probably more tho i haven't written that out)

[–] guillem@aussie.zone 1 points 1 day ago

Isn't that how the Mozilla Directory worked?

[–] breadsmasher@lemmy.world 11 points 1 day ago (2 children)

its all open source! go ahead and build your example, no one is stopping you 🥰 id love to see your implementation

[–] otter@lemmy.ca 1 points 1 day ago

I don't think OP is talking about adding a feature to Lemmy, this is a separate idea for a fediverse platform

[–] Adniem@thelemmy.club 1 points 1 day ago (2 children)

but maybe the thing that needs a how is how we may stop evil instances willing to ruin everything

[–] breadsmasher@lemmy.world 5 points 1 day ago (1 children)

what?

you put out the changes you think should be made.

go ahead, make the changes! Its all open source. free to do as you please.

Put a code change in to support your idea!

[–] Adniem@thelemmy.club 2 points 1 day ago* (last edited 1 day ago) (1 children)

wait i have the concept

instances can have karma if their results get constant downvotes they get put onto a low karma list for the admin to see

and also i have no programming skills i just made this post to do see how the community thinks of the idea before i go to do it

[–] breadsmasher@lemmy.world 4 points 1 day ago (1 children)
[–] Adniem@thelemmy.club 4 points 1 day ago

but what should i learn to make such a project

like what after HTML, CSS, And Javascript

[–] sinza@piefed.social 3 points 1 day ago (1 children)

Most ActivityPub software allows you to defederate, no? You can always implement a way to defederate known bad actors.

[–] Adniem@thelemmy.club 3 points 1 day ago (1 children)

but the question is

how do you find bad actors in the first place

[–] sinza@piefed.social 3 points 1 day ago (1 children)

While we are talking about software that does not exist, there are Mastodon blocklists that can be gathered. At the very least, the FreeDNS blocklist seems like a reasonable bare minimum.

If bad-faith actors are a huge concern -- and for this idea, this is a reasonable concern -- you can look into the idea of an island network.

[–] Adniem@thelemmy.club 2 points 1 day ago (1 children)

maybe a karma system for instances where if instances get downvoted a lot they are put onto that

and so does instances that have a large percent of users downvoting

and a rate limiter for downvoting and upvoting

[–] tales_origin@feddit.org 3 points 1 day ago

You just have to make sure somehow that no network of bots is downvoting all the stuff they don't like

[–] Kierunkowy74@piefed.zip 3 points 1 day ago (1 children)
[–] Adniem@thelemmy.club 2 points 1 day ago (2 children)

but that's not YaCy

it is a static databse, there's no crawler

[–] leanleft@lemmy.ml 1 points 12 hours ago (1 children)
[–] Adniem@thelemmy.club 1 points 11 hours ago (1 children)

you probably did not read the post fully

[–] leanleft@lemmy.ml 2 points 11 hours ago (1 children)

yeah. i think you are probably correct. srry bout that

[–] Adniem@thelemmy.club 1 points 11 hours ago
[–] Kierunkowy74@piefed.zip 5 points 1 day ago (1 children)

Without crawling a website, a search engine cannot know content of the website.

You are effectively reinventing an web directory.

Maybe it could be done as wiki or even federated one like ibis.wiki.

Another prior art would be many pinboards

[–] Adniem@thelemmy.club 1 points 1 day ago

i mean like you don't crawl it to know it exists but only to like get basic info

[–] anarcockworkorange@anarchist.nexus 4 points 1 day ago (1 children)

May I suggest you edit your title to something like "Workshop ideas for a federated search engine"? The current one won't attract lots of people. You may want to cross post this to a programming community as well.

[–] Auster@thebrainbin.org 1 points 1 day ago

Maybe a mix of the threadiverse's common search features, and how NeoDB handles their entries across instances?