|
Top-level
 I've started formalizing what I'm using for filtering on #fedbox into a FEP[1]. > I’d like to discuss an easy way to query a collection, for example, using a well-defined query parameter or an HTTP header. You’d send a request to the collection ID URL, specifying the ID of the object you’re checking for. In my implementation you use iri={url} of the item you're searching. If it's there you receive a collection with just it as an element. 14 comments
 @mariusor as VK shows, some people just add everyone as a friend. VK has a limit of 10000 friends per account. Also, are you sure about *arbitrary* fields? I store everything in a relational database, so I can only really query on fields that have indexes on them unless I'm okay with it going through the entire table row by row (I'm not). Maybe you should specify a response for when the query client asked for isn't possible to easily satisfy. @mariusor individual server? No. Most (all?) fediverse projects won't scale to a typical high-load infrastructure — most you could do is have the DB on a separate server. But you absolutely can have thousands, and even millions, of followers from all over the fediverse. Eugen, for example, has 467586 followers. @mariusor what I'm saying is that there's really no upper bound to the length of a collection. Especially when it's a collection that's the publicly-appendable kind. Group members, walls, forum topics, photo albums in groups, all those things. And you don't need a powerful server to host a collection with a million elements either. |
Gregory's Server
@grishka @activitypub
An example of it working is this (search for 2 activities): https://federated.id/inbox?iri=https://federated.id/activities/7db0d6fc-a007-4eee-998a-a28babf7b577&iri=https://federated.id/activities/486c7b6d-7feb-406f-bb4b-d64224c5d3f5