rektide"tantek> also - "alignment with things like JSON-LD" has always appeared purely political, rather than technical for any actual real-world use-case." i find this statement terrifying
tantekoshepherd: or just "url" has worked fine in numerous objects, in numerous deployments, so bother with "self" which is largely legacy from Atom/XML which is dying.
Loqitantek meant to say: oshepherd: or just "url" has worked fine in numerous objects, in numerous deployments, so why bother with "self" which is largely legacy from Atom/XML which is dying.
tantekrektide, non-alignment is not only not destructive, but essential to exploring multiple varying ideas *in parallel* (more efficiently) to allow for better options to be encouraged, found, developed
tantek!tell tiborKatelbach yes, telecon dialing in etc. information is at https://www.w3.org/wiki/Socialwg#Telecons and be sure to be on IRC as well for minutes, additional converstaion, queuing.
rektidethe latter of which reiterates the "we're being different because it's important not to cooperate" viewpoint in frighteningly explicit terms for me
tantekArnaud: as the most experienced chair here, do you have a URL to a "How to W3C telecon 101" type document which explains the basics of calling in, Zakim, q+, q- etc.?
tantekah, "self" is used *inside* a "links" collection as an implied *rel* - that makes much more sense than just having a property/field named "self".
oshepherdHmm, I need to get around to implementing garbage collection of the ActivityStreams database in Impeller (Android app); its' kinda silly that its' reached 400MB on my phone...
tantekfor that reason in microformats we re-used "uid" from vCard/iCalendar for *universal(global)* unique identifier, which by convention we've only used URLs for.
tantekno need to be ambiguously overloading like that though - hence "uid" is a better option if you want to be explicit about being global/universally unique
wilkieright, even with html, 'id' can be considered global to the page, which is the discrete unit http uses, html isn't going to define a rigid global context, so that makes sense.
oshepherdRegardless, whatever JSON-LD uses doesn't affect us, because JSON-LD lets you define aliases for its' "built in names", because one of the goals is that you should be able to use JSON-LD to map most existing arbitrary JSON data to 'linked data'
oshepherdThe intended "alignment" is that AS would be a specified subset of JSON-LD, and that you would ignore the crazy @properties behind the curtain (indeed, the spec should prohibit them)
oshepherdJSON-LD has two "canonical forms"; one is "crazy RDF triples in JSON", one is "canonical JSON via the mapping", the mapping can be defined such that we always get the nice canonical form people actually want
wilkieoshepherd: my intuition is that it is over-engineered, but that's always what the voice in my head says. looking at it with a clearer head, they have some interesting flexibility wrt interop decisions there.
tantekfor those that find graphs useful. I find it simpler to treat streams (activity and other) as simply a list of objects with properties with values
jasnell-github [w3c-socialwg-activitystreams] jasnell pushed 1 new commit to master: http://git.io/Ul3paA
tantek !tell tiborKatelbach yes, telecon dialing in etc. information is at https://www.w3.org/wiki/Socialwg#Telecons and be sure to be on IRC as well for minutes, additional converstaion, queuing.
oshepherd rektide: I cite my /real/ AS1 profile as an example which uses rel=self to point to itself: https://gist.github.com/oshepherd/8c879449b3a55316cd73
trackbot Sorry, Arnaud, I don't understand 'trackbot makes some of this stuff easier'. Please refer to <http://www.w3.org/2005/06/tracker/irc> for help.