Skip to main content


Based on our decision in the last #SocialCG meeting, I set up a new GitHub repository for the #ActivityPub HTML Discovery Task Force.

https://github.com/swicg/activitypub-html-discovery

I added what I think are some of the most important tasks and user stories. I hope others contribute, too!

https://github.com/swicg/activitypub-html-discovery/issues

in reply to Evan Prodromou

I thought we said activities embedded in HTML were a bad idea?

(But that linking to activities via Link header or <link> tag was fine)

in reply to Evan Prodromou

The Semantic Web idea from Berners-Lee w/ intelligent agents…
in reply to Evan Prodromou

Re: Based on our decision in the last #SocialCG meeting, I set up a new GitHub repository for the #ActivityPub HTML Discovery Task Force.


Hi @evan@cosocial.ca! Has there been any additional movement on this task force?

I want to ensure that some of my interests are represented, e.g. an html route should issue an HTTP 308 to the ActivityPub route and vice versa based on received Accept header on mismatch.

Shouldn't have to make this explicit but not having this breaks a common UX flow (pasting the html url into an ActivityPub client)

in reply to julian

>I want to ensure that some of my interests are represented, e.g. an html route should issue an HTTP 308 to the ActivityPub route and vice versa based on received Accept header on mismatch.

Most Fediverse projects already support this (maybe with a different status code like 302 but I don't think it matters). What do you expect from the "task force"?

in reply to silverpill

Re: Based on our decision in the last #SocialCG meeting, I set up a new GitHub repository for the #ActivityPub HTML Discovery Task Force.


most


@silverpill@mitra.social you said it yourself 😉 I think a recommendation via CG report or FEP would be a good outcome for this task force.

A fellow admin attempted to retrieve a discussion via Iceshrimp and simply didn't know how to (the originating instance of the reply in question was SocialHub, which does not use the same url structure for their AP endpoints, nor are the endpoints exposed for guests)

cc @evan@cosocial.ca

in reply to julian

@julian @silverpill Yeah, I think that's why we created the task force; to make a Report.
in reply to Evan Prodromou

@julian @silverpill Also, I think 308 redirects are an OK option, but some platforms aren't going to be able to implement them. We're going to have to document other practices already used on the social web, like <link rel="alternate">.
in reply to julian

Here's your FEP:

# FEP-ce05: Discovering ActivityPub object from an HTML document

## Summary

## Content negotiation

If HTML document has an ActivityPub representation and request has an Accept header containing application/activity+json or application/ld+json; profile="https://www.w3.org/ns/activitystreams" media type, the server SHOULD return a response with HTTP 308 Permanent Redirect status code and a Location header containing ActivityPub object ID. The server MAY use 302 Found status code.

If ActivityPub object has an HTML representation and request has an Accept header containing text/html media type, the server SHOULD return a response with HTTP 308 Permanent Redirect status code and a Location header containing an URL of HTML document. The server MAY use 302 Found status code.

## Copyright

CC0 1.0 Universal (CC0 1.0) Public Domain Dedication

To the extent possible under law, the authors of this Fediverse Enhancement Proposal have waived all copyright and related or neighboring rights to this work.

-----

@evan @evan

This entry was edited (3 weeks ago)
in reply to silverpill

cool, thanks for the link. I'll add it to the issues list.

UPDATE:
https://github.com/swicg/activitypub-html-discovery/issues/18

This entry was edited (3 weeks ago)
in reply to silverpill

@silverpill @julian So, the Community Group sets up task forces to work on particular issues w/r/t ActivityPub.

We have the Webfinger, HTTP Signature, E2EE, Data Portability, Testing, Forums, and now Discovery TFs.

The goal of most of these task forces is to produce a Report, which documents best practices already used on the Fediverse, or proposes new mechanisms for getting tasks done.

in reply to Evan Prodromou

@silverpill @julian For discovery, this will be documenting the different ways to take an HTML document URL or a binary object URL and get the corresponding ActivityPub object ID.

For example: content negotiation, <link rel="alternate">, Webfinger, Link: header, OGP, embedded JSON-LD.

Use cases include following a person; liking a page; sharing an image.

https://github.com/swicg/activitypub-html-discovery

in reply to Evan Prodromou

is this related to our fediverse:creator tag in some ways?
This entry was edited (1 month ago)