Replies (16)
I've got to think about this one a bit more but my first thought is that it might not make sense for "all purpose social apps" like amethyst to include general catallax events. an item you buy on nostr is sort of like a classified posting and makes sense in a social app context. catallax tasks would likely be more specific to a given employer, project, or whatever else people might use it for...
then again, perhaps clients might somehow use catallax events to drive or fund feature development and would only include some events relevant to their app... 🤔
I look forward to seeing how client developers want to use it, and to creative ways to inter-op with shopstr/marketplace kinds
ok I read in practice you are recreating nip-72 where you replace the community with an arbiter, the one who examines a job. (33400)
in 33401 the patron can put a job to do, for simplicity let's think of this as a long form content 30023 where the job is defined and has defined tags and the call to the arbiter if he accepts. when someone accepts the job it is added. After if this one does the job and the patron pays him we put the arbiter issues the receipt of kind 3402.
Job completed.
Free agents problem, they are never listed or examined and cannot be evaluated, that is, you will only evaluate the examiners and those who submit the job.
the free agant's npub gets referenced in the 33401 when the task moves to "assigned" status. and when then task is concluded and the 3402 is created, the result of the job allows you to evaluate the free agent.
so querying a whole set of events would allow you to evaluate the performance of all the workers involved. and future integration of WoT or GrapeRank would provide onlookers with their own tuneable algorithm for the various evaluations feasible to glean from the available data.
I don't really see the connection to NIP-72. anyone can post any of the Kinds, so "moderation" seems like the wrong way to think about it.
the core idea is more about providing tools for unaffiliated parties to get things done together by leaning on a 3rd, mutually-trusted, intermediary.
you don't have a univocal view of the merits, because they are mentions, you can make a kind 100xx with the merits, it can also be a badge or something similar, but badges are a thing in themselves, so maybe not highly recommended
the workflow is similar or has some analogies
there is one or more moderators, and notes are approved with kind 4550
Using Habla for long form you can tag a community kind 34550
Any Nostr event can be posted to a community.
Clients MUST add one or more community a tags, each with a recommended relay.
but it is a type of NIP not implemented or partially implemented so it is good for experiments
are you suggesting I leverage that NIP for this purpose? I'm open to the idea, but I'm trying to understand the center of what you're getting at. the comments are a bit scattershot even if each one makes some sense on its own.
I don't see what you mean. "univocal view of the merits"? what do you mean by that and why would it be desirable?
if there are multiple relays multiple implementations, you might not find all events instead of searching a list, but in the end you would still have to synchronize the list so not very useful
It's just a test that can be done.
perhaps I'll put a bounty up for re-implementing the reference client with these kinds 😉
good. some events might not want to be found.
my hypothesis is that's exactly why it hasn't gained traction.
probably also competing with AI, why is it important to "map" free agents because they are less numerous
Once again I'm not sure I'm getting your point.
Because of the above, I can't tell if this is relevant or not, but: there's no reason free agents couldn't be AI agents. if they do good work and have a positive reputation, who cares what's behind the nym?
