2016-06-15 UTC
shepazu_, jasnell, bblfish, bblfish_, jasnell_, pdurbin, peacekeeper and KevinMarks joined the channel
# 15:27 cwebber2 the thing I like least about ActivityStreams is the use of mixedCase for property names ;) peacekeeper and KevinMarks2 joined the channel
# 15:55 cwebber2 originally, the text about uploading stuff vaguely said "upload to the file endpoint" # 15:55 cwebber2 however we hadn't specified a file endpoint on the profile object, so I'm adding one # 15:56 cwebber2 should it be "file"? That doesn't seem clear that it's an endpoint. micropub does media-endpoint # 15:56 cwebber2 however none of our other urls, which are endpoints in a sense, say "endpoint" # 15:59 cwebber2 so we could do be mediaEndpoint to be consistent with MicroPub # 15:59 cwebber2 but again, it looks odd becuse we don't have "fooEndpoint" for anything else # 16:00 aaronpk it's kind of a collection, it just can't be queried like a collection right? # 16:00 cwebber2 aaronpk: hm I don't think so. It's pretty much just an upload endpoint. # 16:01 cwebber2 does that sound sensible to you based on the above discussion? bblfish joined the channel
# 16:03 tsyesika could be, another solution could be to have a "endpoint" property which is just a mapping for more "endpointy things" (needs better phrasing) # 16:03 cwebber2 tsyesika: hm, I'm not a bit fan of that... we don't *have* any more endpoint'y things :) # 16:04 cwebber2 I think we'd need an idea of more kinds of endpoints we'd need/want if we were to do that # 16:04 tsyesika it's true which is why i was waiting for other people to suggest better ideas heh # 16:05 tsyesika did we decide this was out of scope, i'm not sure, i'm sure we have other endpoints though? # 16:07 aaronpk also interesting that they both took an RPC style approach rather than using HTTP verbs # 16:08 cwebber2 tsyesika: I need to think about that, or maybe even work on what it looks like in implementation :) # 16:09 cwebber2 for now, I mark it as fileEndpoints, and make a note for myself to think about or open an issue about how oauth discovery is done # 16:09 cwebber2 I'm admittedly a bit brain-fuzzy at the second about how that works. For some reason I thought we had solved it already # 16:09 cwebber2 but maybe that's because the pump api uses .well-known and we don't. tantek, peacekeeper, KevinMarks and KevinMarks2 joined the channel
# 16:44 tsyesika cwebber2: well we mention "oauth2 with bearer token" but completely ignore how to get that, which for me is a problem, if a client doesn't know how to get a token because it's undefined, how will clients and servers talk to each other # 16:46 tantek hey Social Web WG, I updated https://www.w3.org/wiki/Socialwg#sched which had not been updated in a while with what has happened, and general expectations for moving forward. I think the changes should be uncontroversial reflections of the working group's work, but definitely speak up if you see anything unclear or contestable. # 16:49 cwebber2 yeah I'm capturing the oauth endpoint thing as an issue bblfish joined the channel
jasnell joined the channel
bblfish, jasnell, jasnell_, Arnaud1, tantek, KevinMarks, peacekeeper, bblfish_ and shepazu joined the channel
cwebber2 tsyesika: so, I'm updating the binary file upload stuff to reference https://github.com/w3c-social/activitypub/issues/23
aaronpk is it these? https://www.w3.org/TR/activitypub/#collections
tsyesika http://w3c-social.github.io/activitypub/#actor-objects example numbr 2
tantek hey Social Web WG, I updated https://www.w3.org/wiki/Socialwg#sched which had not been updated in a while with what has happened, and general expectations for moving forward. I think the changes should be uncontroversial reflections of the working group's work, but definitely speak up if you see anything unclear or contestable.