#sknebelIn https://www.svenknebel.de/posts/2017/3/9/ I claimed that a the W3C annotation stuff would already be accepted by a webmention receiver as a valid basic "mention", and I today talked to someone from hypothesis and discovered that they already publish annotations in that W3C data format
#Loqi[Sven Knebel] notes for HWC Berlin 2017-03-08
#sknebelso if I had a receiver implementing it I could test my claim
#sknebel(and the claim would be a lot more useful if the set of "endpoints accepting JSON" would be non-empty ;))
#GWGsknebel: I'd read your writeup on this eagerly.
#aaronpkit wouldn't take much for me to accept JSON sources as "valid", but it would be more complicated to parse useful information out of them
KartikPrabhu and [chrisaldrich] joined the channel
#sknebelyes, that would be the next problem, and the point where there isn't enough standardization in generic "JSON" sources to do it without special cases
#aaronpkthe JSON sources would need to standardize on a vocabulary, like activitystreams 2.0, or this annotation JSON
#aaronpkyes, the LD bit was shoehorned into it. you can pretty much ignore the LD part if you don't care about it
#aaronpkit was meant to look like "normal JSON" as well
#GWGIs there a standard way to tell a client what your vocabulary is? I thought there was.
#sknebelGWG: JSON-LD seems to have a bunch of ways of expressing that (just manually looking at various sources of it right now, and they do it differently)
#GWGI haven't spent much time with JSON-LD, just the JSON I needed.
#sknebelhm, from a quick read it seems like scanning for a field named "@id" and linking to that URL might be useful -> thats more or less what u-uid is in mf2 and the sites I've looked at deliver human-readable content there
#[kevinmarks]JSON-LD is very flexible. This can be a good thing (as it can represent something like activity streams 2) but it also means there are a lot of ways to represent a given thing.
#tantek"vertical middles of ... text" is just rephrasing the above. by "text" do you mean the metrics, glyph ink, including ascenders? accents? descenders?
#aaronpkvertical-align:center doesn't work because they are inline elements i think
#tantekor rather, might not be the "center" that you have in mind
#ZegnatI frequently misremember the actual value keywords, but if you use CSS a lot you should at least be able to remember what it affects, imho.
#tantekalright, with multiple publishing uses of p-checkin / u-checkin that seem to "work", that's probably worth documenting as current "best practice"
#tanteknow the question is whether any consuming code will notice and do anything with p-checkin / u-checkin
#jonnybarnesif you have an h-entry note, with an associated download link, maybe its a pdf of a flier for an event the note talks about, or whatever, how would you markup the <a> tag?