Now that we are making accessible instance information through the API I think it's time to rethink instance discovery. Coming from an information sciences standpoint we should differentiate the obtainable data from the information. You start making it possible for someone to write an application to crawl instances based on "who knows who" excluding those with a flag stating they wish to be excluded. You can then form a dynamic database. THEN working from that it's time to curate

With this method you can decententralize instance discovery by not requiring admins to manually add themselves to one list that keeps all the data. Anyone can gather the data for their own discovery application. Without third party services.

The data is not the information. Each application or service can then use their own methods to filter and curate the list. Of course joinmastodon.org should have some kind of picker but drawing on an open database is more reliable

Like, I could pull the database then make my own curated non-dynamic list I publish based on my own criteria.

The presentation to the user is how we give this data meaning and make it into information. We could even establish like "we all know these categories are what joinmastodon uses" but allow that to not go through a middleman who is a single individual and not really accountable to anyone, and inevitably not going to agree with everyone in the work on everything

Mostly I want to make it so admins aren't adding themselves to tons of different lists when instead they can set something on their instance that let's the curators come to them.

@shel 💯 on that. Lmk which things are missing from the instance API. Admin username and list of languages could safely be on there. Categories or tags or whatever - that's the complicated bit...

@gargron maybe the code of conduct could made accessible so an application could RegEx it

@shel I was thinking... You know software licenses, like MIT, GPL, Apache... You know how they are compatible, and you say, "GPL-compatible license"... What if we worked out different models of CoC, gave them names, and categorized by compatibility to them?

@shel @gargron i had this very thought when i first joined masto last April https://mastodon.cloud/@lifning/284521
that could be a really helpful abstraction to simplify explaining to people what's usually bogged down in minutia

@lifning @gargron thats a VERY interesting idea.... i think you'd want to form some kind of third-party committee to form such a thing

@shel @lifning Maybe it could be a github repository where ppl could submit the staff as PRs.

Starting models would probably be the m.s. CoC which is copied by a lot of places, and the Awoo.space one?

@gargron @lifning so like, what i'm thinking is:

We want a suchandsuch-compatible CoC Status you could... self-mark as? A Community Conduct Compatibility Policy (CCCP? lmao jk jk)

How are you claiming to be when you identify as this. A safe community? A harassment-free community?

after you pick that goal; you need to have a group basically be a thinktank to discuss and figure out what that means. Stuff like "Policy against hatespeech" perhaps

@shel @lifning @gargron this is a lot of politics and i'm not sure for how much benefit. do discord servers have a similar thing?