#microformats 2018-08-17
2018-08-17 UTC
tantek__, [jgmac1106], jwheare12, averell26, mattfo0, [chrisaldrich], PaulCapestany, tripleslash2, KartikPrabhu and [asuh] joined the channel; vivus left the channel
# 
gRegorLove Merged the implied tests updates after willnorris and ben_thatmustbeme gave the thumbs up.
# gRegorLove Now to see how php-mf2 does with it.
# gRegorLove Hm, packagist is pointing to a commit from March for dev-master. https://packagist.org/packages/mf2/tests#dev-master
# gRegorLove php-mf2 fails 30 tests, but some are just things like URLs not being made absolute. microformats/tests assumes you pass in a base url of http://example.com/
# gRegorLove some whitespace stuff
MetBril, Nineteen, wodencafe0, [mrkrndvs], primalz10, [asuh], KartikPrabhu, [iambismark], tesu9, [metbril] and depleted25 joined the channel
# 
depleted25 https://www.youtube.com/watch?v=ADn2IJnTRyM
Exagone31320, graingert18, macker24, alphaseg4, macky, Hello7126, gamma1, raSter^16, Guest60650, Guest13524, adactio, [pfefferle], steven22, StephenS22, gidzit, Milos0, lorimer23, [grantcodes] and edong2322 joined the channel
# 
ben_thatmustbeme gRegorLove: https://ben.thatmustbe.me/mf2tests/ has been updated
# ben_thatmustbeme a lot of the tests that are failed on things are whitespace related
[jgmac1106], Typhon23, cylon512_, MyIgel, [kevinmarks], pj23, Random9, wols, [keithjgrant], [xavierroy], [grantcodes], GWG, willnorris, Guest72310, rkta, Madbrad20018 and eldritch12 joined the channel
# 
willnorris ben_thatmustbeme: I've got a couple of patches for mf2-tester whenever you've got a minute to review. First to add a license to the repo, and second to add the go parser to the table
# 
Loqi willnorris: tantek left you a message on 2016-06-20 at 1:04am UTC: FYI I created https://github.com/microformats/microformats2-parsing/issues to collect future issues.
# Loqi willnorris: tantek left you a message on 2016-06-20 at 1:26am UTC: could you add the resolution / and or spec text we came to at IndieWeb Summit to this issue? http://microformats.org/wiki/microformats2-parsing-issues#unicode_generation_in_JSON
# Loqi willnorris: tantek left you a message on 2016-07-06 at 2:30pm UTC: as someone who has participated in microformats2-parsing issue discussions / resolutions in the past year at least, could you offer your opinion, even if just a summary +1/0/-1 on my documenting of our existing rough processes here: https://github.com/microformats/microformats2-parsing/issues/1#issuecomment-230912142
# willnorris wow, uh thanks Loqi. Two year old message, eh? Better late than never I guess :)
# willnorris guess that shows how long I've been out :(
# willnorris hey :)
[tantek] and [jgmac1106] joined the channel
vivus, mattfo0, Guest73346, [eddie], BackUP29 and \mSg16 joined the channel
# ben_thatmustbeme thanks willnorris
# ben_thatmustbeme merged
# ben_thatmustbeme also, how do i get the go microformats parser installed? I don't see any documentation on it other than an alpha git repo
# willnorris See the readme in my go branch of mf2-tester
# willnorris The "go get" command
# willnorris (on mobile, will send direct link when I'm back at my desk)
mobijubo8 joined the channel
# ben_thatmustbeme i will take a look, have to run for the moment
xCore13, [jgmac1106], [eddie], jhesketh25 and GTAXL10 joined the channel
# ben_thatmustbeme thanks willnorris, https://ben.thatmustbe.me/mf2tests/ is updated
# ben_thatmustbeme realized i had not pushed the diffs code
# ben_thatmustbeme so thats up now
# ben_thatmustbeme there are a few tests that just no-one seems to agree on, lol
# 
KartikPrabhu seem to be mostly in backcompat
rosseaux26 joined the channel
# ben_thatmustbeme yeah, i'm looking at the v2/h-card/impliedname
# ben_thatmustbeme just looking at checksums, not a single match
# ben_thatmustbeme there are a bunch that are whitespace things
# KartikPrabhu yeah the whitespace stuff is also new-ish
# ben_thatmustbeme there are some 20 boxes that will turn green when 4.0.8 of ruby goes out
# KartikPrabhu I would rather code to the spec than these tests
# gRegorLove We should probably specify the base URL to use with the tests. Some are just absolute URL differences: https://ben.thatmustbe.me/mf2tests/python/microformats-v2/h-card/impliedname.json.diff.txt
# gRegorLove ben_thatmustbeme, Would we be able to set up your mf2tests thing as CI for the test suite?
# ben_thatmustbeme gRegorLove: all of them are supposed to get the base url of example.com, which is what i give all the parsers
# ben_thatmustbeme thats why i couldn't use the ruby command line script that the gem installs, as you can't pass it one
# ben_thatmustbeme so if they aren't getting the correct domain, its on the parser
# gRegorLove Sorry, I mean some of the tests' json files don't have an absolute URL, but they should.
# ben_thatmustbeme ? which ones?
# ben_thatmustbeme i thought we had them all
# KartikPrabhu the one gRegorLove just linked to
# ben_thatmustbeme oops, missed that
MetBril joined the channel
# KartikPrabhu the <base> should be used to make things absolute
# gRegorLove Oh, willnorris already PRed the fix for that one :)
# KartikPrabhu oops ignore my comment, mf2py is failing that for some reason even though it should use <base>
# KartikPrabhu wait are URLs supposed to be made absolute in content>html ?
# KartikPrabhu I thought no
# gRegorLove No
# willnorris There are also some cases of e- properties "html" value getting urls expanded, but I don't think they're supposed to (only the "value" should, right?)
# gRegorLove h-card/impliedname.json fix is merged
# willnorris Or if they should be expanded, I don't think that's clear in the spec
# KartikPrabhu willnorris: yeah exactly . I don't think they should be but the tests have those expanded out
# willnorris Yep. Was going to send a PR for that later, but happy to have someone else beat me to it
# willnorris Wasn't there discussion somewhere about updating the spec to actually say they *should* be expanded? I can't find it now, so may have just imagined that.
# KartikPrabhu not that I recall
# willnorris Okay, then I'll assume I imagined it 😀
# gRegorLove There's a proposal to move the "normalized absolute URL" line down, inline on http://microformats.org/wiki/mf2p#parsing_a_u-_property
# KartikPrabhu but I can see it as a use-case for consuming for instance displaying the content > html in comments
# gRegorLove But that's only on u-*
# willnorris That's probably what I was thinking about
# willnorris Yeah, it does seem like a reasonable thing to consider, since those relative links become useless in any other context
# KartikPrabhu right
# KartikPrabhu but currently the lean has been towards not "disturbing" the authored html
# KartikPrabhu but this seems like a valid exception. we could either leave it to the consumer or do it in the mf2 parsing
MJ9410 joined the channel
# gRegorLove My first thought is that should be part of a sanitization process by the consumer, but also once it's parsed you no longer have the <base>, if it exists
# gRegorLove decides to not think about it too much for now :)
# KartikPrabhu gRegorLove: hmm yeah the <base> case is tricky
# KartikPrabhu willnorris: gRegorLove: maybe open an issue on the parsing repo?
Miron28, ws2k326, dimi34ka13, gRegorLove, [kevinmarks] and [tantek] joined the channel
# ben_thatmustbeme oh, also gRegorLove, CI for the test suite?
# ben_thatmustbeme added a totals to the bottom of the page there too :D
# gRegorLove ben_thatmustbeme: Yeah, kevinmarks was asking about it: https://github.com/microformats/tests/issues/88
# ben_thatmustbeme hmm, i don't see why you couldn't set it up for informational reasons, seems sort of weird in my brain that you are running the test suite against parsers to make sure the tests make sense while writing the parsers to test against the test suite
# ben_thatmustbeme but thats just my brean
supercool10, tantek__, [iambismark], tantek and [kevinmarks] joined the channel
# 
[kevinmarks] well, what ben just built (and aaron kind of did before, and so on) is a kind of CI results page, so using one integrated into github makes some sense
# [kevinmarks] glen's code in the repo is another version of this
milkii, [grantcodes], [keithjgrant], mattfo0, sam28, u1dzer026, [tantek] and [dougbeal] joined the channel
# [kevinmarks] Hmm, not sure that will's comment on http://microformats.org/wiki/microformats2-parsing-issues#unicode_generation_in_JSON can work - we have an HTML fragment, not a document so we can't convey the encoding. Normalising to utf8 is then right answer
# KartikPrabhu [kevinmarks]: i think the mf2 spec should specify explicitly that utf8 is to be used
# KartikPrabhu also v
# [kevinmarks] Yes, and that any html source encoding should be converted
james41382, Aleszandro and [iambismark] joined the channel
bradenslen, wgma, cheim, naos16, tantek and Strog21 joined the channel