#social 2017-08-18
2017-08-18 UTC
timbl and eugr joined the channel
# 
eugr puckipedia: https://github.com/tootsuite/mastodon/pull/4630
# eugr this is probably the simplest way for me to deal with the OStatus remnants https://github.com/tootsuite/mastodon/pull/4632
# eugr i.e. okay, we're gonna publish invalid "id" attributes, which from a pure-AP implementation's perspective is junk, but at least it should work from Mastodon to Mastodon this way, and such junk "id"s will naturally disappear over time as OStatus becomes less and less used
xmpp-social, timbl and dlongley joined the channel
# Gargron puckipedia: curious, i am looking at https://web-payments.org/vocabs/security#publicKey and it uses "@id" instead of "id", what's up with that
# Gargron i am looking at LDS again and the document for json normalization is: https://json-ld.github.io/normalization/spec/
# Gargron "This document is a detailed specification for an RDF dataset normalization algorithm. The document is primarily intended for the following audiences:"
# Gargron "Software developers that want to implement an RDF dataset normalization algorithm."
# Gargron "Masochists."
# puckipedia Gargron: the AS2 json-ld context specifies that "id" is an alias for "@id"
# Gargron ah okay
# puckipedia which means that "@id" gets simplified into "id"
# Gargron should we have a "type": "Key" too?
# puckipedia I'd vote for either adding "Key" to the AS2 spec as sec:Key, or just saying "type": "https://w3id.org/security#Key" for now
# puckipedia well
# puckipedia maybe we could add the web-apyments context
# puckipedia so "@context": ["[as2 context]", "https://w3id.org/security/v1"]
# Gargron hmph
# puckipedia I found the first actual real-world use of a @context array btw
# puckipedia that hubzilla dev
# Gargron what do they do?
# Gargron also: i found a json-ld normalization gem for ruby, but it works on RDF data structures, i.e. not a simple hash. so it would require including the RDF ruby gem, converting json to hash, hash to RDF, then normalizing, then the same in reverse worder
# Gargron *order
timbl joined the channel
# puckipedia Gargron:
{"@context": ["https://www.w3.org/ns/activitystreams", {"me": "[magic envelope xmlns url]"}], ...}# Gargron ..magic envelope? what for..?
# Gargron ohh, they dont know about http sigs
# puckipedia nooo
# puckipedia it's http signatures *plus* magic envelope
# Gargron what
# Gargron why
# puckipedia shrug. hubzilla.
# puckipedia oooh
# puckipedia this is for signing the response
# Gargron but the response is TLS + DNS
# Gargron i.e. why sign that
# Gargron it's trustworthy enough
# Gargron i wonder if for our purposes we could just wing it and skip json normalization
# Gargron just hope that Oj.load -> remove signature subtree -> Oj.dump will be deterministic
# Gargron (i am bringing this up in the context of redistributing responses to one's posts, as well as deletes of things you've shared - both innovative features we want, whereas everything else is just aimed at feature parity)
# Gargron puckipedia: could you comment please? https://github.com/tootsuite/mastodon/issues/4607 did i do that fix in error?
timbl joined the channel
# puckipedia Gargron: I think I misread the code
ThibG joined the channel
# puckipedia here's where the magic happens for ActivityPub and many other related protocols
# ThibG so, I hear that https://www.w3.org/TR/activitypub/ isn't exactly current… is there a more recent version around?
# puckipedia well, there's been a few changes
# puckipedia mostly minor
# puckipedia and I guess the standardization of things like how HTTP signatures are used, and accept/reject follow etc aren't quite done yet
# puckipedia anyways, https://w3c.github.io/activitypub/#changes-9-may-to-present is the Bleeding Edge
# ThibG thanks
# puckipedia Gargron: I moved to using userId + "#key" for the public key format and always embedding the public key anyways
timbl joined the channel
# 
Loqi [ajordan] RESOLVED: PROPOSED: approve https://www.w3.org/wiki/Socialwg/2017-08-01-minutes minutes
timbl joined the channel