#microformats 2015-10-29

2015-10-29 UTC
nitot, tantek, rawtext and fuzzyhorns joined the channel
#
kylewm noticing that http://microformats.org/wiki/h-card doesn't say much about venues. just people and organizations
#
kylewm is that on purpose?
#
tantek venues are just a subset of organizations
#
tantek there are many subsets / types of organizations
#
tantek I can add that to an FAQ
#
tantek edited /h-card (+10) "/* Status */ top-level backcompat" (view diff)
nitot joined the channel
#
tantek kylewm: note that "venues" is mentioned in one of the examples in the wild
#
tantek (just double-checked the example too)
#
kylewm :+1:
#
tantek do you have suggestions for improving the spec regarding applicability to venues?
#
kylewm tantek: no, I was just looking for the right terminology to describe an h-card that is a place rather than a person, "venue", "location", etc.
#
tantek "venue" is a named location
#
tantek location could be an adr or geo
#
kylewm ok, that helps, ty
#
@pypi_updates mf2util 0.2.9: Python Microformats2 utilities, a companion to mf2py https://pypi.python.org/pypi/mf2util/0.2.9 (twitter.com/_/status/659585704469635072)
nitot, gRegorLove, fuzzyhorns, tantek, MeanderingCode and jimtyhurst joined the channel
#
tantek edited /Template:cc0-owfa-license (-3) "-linebreaks" (view diff)
#
tantek edited /microformats2-parsing (+83) "expand header per feedback re: status, feedback" (view diff)
fuzzyhorns, tantek, Erkan_Yilmaz, adactio, MeanderingCode, TallTed, pfefferle, gRegorLove and rawtext joined the channel
#
benthatmustbeme edited /jf2 (+472) "add important links" (view diff)
fuzzyhorns joined the channel
#
kylewm ben_thatmustbeme: why does your service generate syndication: { url: ... } instead of just syndication: ...?
#
ben_thatmustbeme aaronpk: was in favor of trying to make anything that is a url be url:
#
kylewm providing a mime-type for content: is slick, i like that
#
ben_thatmustbeme so that it requires no knowledge of the actual vocabulary
#
kylewm hmm... so should dates be "published": { "datetime": "..." }?
#
aaronpk kylewm: ha interesting point
#
aaronpk mayyybe
#
aaronpk the other reason for alwaysh having urls be {url:...} is that it lets you optionally attach meta data that describes the url, since URLs are always referring to some other things
#
aaronpk most common of course is the mf2 vocab of the contents of the URL, like name, content, etc
#
aaronpk but it also allows for things like mime type of images
#
kevinmarks so is this a way of getting the rel-urls info back in?
#
ben_thatmustbeme aaronpk: from what i can tell the ur: thing isn't needed for things like syndication.
#
ben_thatmustbeme at least to make json-ld people happy that is
#
ben_thatmustbeme this is all getting so confusing
#
aaronpk no i wanted that in there regardless of json-ld
#
kevinmarks you want to be able to reconstruct the url markup?
#
aaronpk yeah, and not have to regex on the value to know if it's a url
#
kevinmarks that is a bit like xoxo, except that wasn't nested
#
kevinmarks schema review problems between apple and google: https://twitter.com/jamesthomson/status/659814941852082176
#
Loqi [@jamesthomson] Well, cheers Google for completely wrong star ratings on my apps…
#
kylewm kevinmarks: I tend to think the rels/rel-urls would be better served by a separate parser. I never want to run the whole mf2 parser just to get like the webmention endpoint
#
aaronpk good point, i usually don't use both kinds of values at the same time
fuzzyhorns joined the channel
#
kevinmarks historically we have had more interaction between rels and formats
#
kevinmarks but the 'rels apply to the whole document' view undermined that
#
kylewm yeah, that makes sense
#
kevinmarks but if you want to get type,hreflang etc info for the {url: } struct, the only place we have that at the moment is int he rel-urls
#
aaronpk well the parser still has access to the rels, so it could fill those values in
#
kevinmarks a bit like the xoxo one: http://www.unmung.com/xoxotojson?url=http%3A%2F%2Faaronparecki.com%2Farticles&pretty=on
#
aaronpk correct me if i'm wrong, but the "text" value on that json is from the link tag, and hasnothing to do with the contents of the URL?
#
kevinmarks reads decade old code to check
#
kevinmarks I think it is the link text, yes
#
aaronpk i guess not so much a code question but the intent
#
kevinmarks yes, it's the link text http://microformats.org/wiki/xoxo#Special_Properties
#
ben_thatmustbeme wipes his brow
tantek joined the channel
#
ben_thatmustbeme https://github.com/w3c-social/Social-Syntax-Brainstorming/wiki/jf2#context
#
ben_thatmustbeme to do it the correct way you have to put json-ld contexts on each of the h-* pages, and i am not going through all that
#
aaronpk on the wiki pages?
#
ben_thatmustbeme yeah, or on some other page you point to instead of the wiki page
#
aaronpk ah funny
#
aaronpk we can deal with that later :)
#
ben_thatmustbeme honestly, i think its fine as is
#
aaronpk probably
#
ben_thatmustbeme alternatively an updated http://microformats.org/wiki/existing-classes for mf2 ones would work
#
kevinmarks tantek, was there a decision to only put mf1 existing classes in that page, or did we just stop updating it ?
#
tantek I don't know who was updating it in the first place.
#
tantek it seemed like something that would quickly get out of sync so I didn't put any effort into it.
#
kevinmarks ah OK
#
kevinmarks I suppose https://github.com/tommorris/mf2py/blob/master/mf2py/backcompat.py is a more current reflection
#
tantek kevinmarks: for backcompat - the h-* specs have the full lists
#
tantek that's the canonical list - not any one particular implementation which may have more or fewer
#
kevinmarks makes sense
fuzzyhorns and tantek joined the channel
#
@RMVGaines I've said it before and I'll say it again, forms of "structured data," including MicroData, MicroFormats, and... https://www.facebook.com/rmvgaines/posts/10156245605435437 (twitter.com/_/status/659873427965255680)
KartikPrabhu and TallTed joined the channel