#microformats 2020-09-13
2020-09-13 UTC
[tw2113], gRegorLove, [Emma_Humphries] and [chrisaldrich] joined the channel
# 
[chrisaldrich] GWG++ for organizing, hosting, and documenting
14WAADUQT, [asuh], [schmarty], [tantek] and [chrisaldrich] joined the channel
gRegorLove and [frank] joined the channel
[tantek] and cjk101010_ joined the channel
# 
ben_thatmustbeme the u-photo with alt text may have some other much more prevelent issue we hadn't considered.... it was assumed that the parser would parse alt text as the text value
# ben_thatmustbeme specifically i am finding multiple people who make an author tag with p-name on a div that contains a p-photo with alt="<their name>"
# ben_thatmustbeme so the alt text is expected to parse out to the text of a p-* > u-photo[alt] in to the textContent for the p-name to have any value
# ben_thatmustbeme also, i haven't found much of anyone using alt-tags in their content and many of those have alt tags that shouldn't be there... things like alt="a photo", alt="a jpeg", alt="Picture1" etc where the fault is not really on the parser at that point
# ben_thatmustbeme not done looking through all the examples i can yet
# ben_thatmustbeme Zagnat++
# 
Zegnat https://gist.github.com/Zegnat/65ed9a9fb0546fb8c4aa0c0b790b8a40 - updated! this validates all JSON from the test suite, without allowing any properties not defined in the schema. So that also verifies there is no unexpected extra debugging output in the test suite :)
jgmac1106_ and [jeremycherfas] joined the channel
# Zegnat rel and rels done, all tests still passing. Feel free to play with the schema. You can put https://gist.githubusercontent.com/Zegnat/65ed9a9fb0546fb8c4aa0c0b790b8a40/raw/1170ad1c3855b9b3af70a26f7bf67bc8968879d4/microformat.schema.json in to https://www.jsonschemavalidator.net/ and put any mf2 JSON next to it to (hopefully) see it pass!
# ben_thatmustbeme bit of silly pendentry, but the entirely empty string is marked valid
# ben_thatmustbeme i think at a miminumum it needs
{"items":[],"rels":[],"rel-urls":[]}# ben_thatmustbeme also, if it try any additional properties, we said its allowed to have other items, not related to this
# ben_thatmustbeme
{"items":[],"rels":[],"rel-urls":[], "otherthing":[]} should be valid# ben_thatmustbeme is "rels" and "rel-urls" allowed to be "", i would think it needs to be an array
# ben_thatmustbeme
{"items":[],"rels":"","rel-urls":""} is currently valid# ben_thatmustbeme i just realized you mentioned it being strict right now
# ben_thatmustbeme i tested a pull and parse of tantek.com and it validated fine
# Zegnat The minimum document is set to require items, rels, and rel-urls: https://gist.github.com/Zegnat/65ed9a9fb0546fb8c4aa0c0b790b8a40#file-microformat-schema-json-L43
# Zegnat aaronpk: I also remembered why I first only published a schema for a single microformat object. I originally wrote the schema (and most of /microformats2-json) for Micropub validation.
# ben_thatmustbeme ahh, yeah, no input at all
# Zegnat But because the Micropub spec refered to "the Microformats 2 JSON format" I started writing https://microformats.org/wiki/microformats2-json
gRegorLove and [chrisaldrich] joined the channel
# [chrisaldrich] Zegnat++
[tw2113], [tantek], gRegorLove, gRegorLove_, [frank], KartikPrabhu, [email096] and jgmac1106_ joined the channel
# Zegnat I have moved the schema into its own repo (https://github.com/Zegnat/microformats2-json-schema) and officially licensing the whole thing CC0. The simple web GUI for it will follow tomorrow (debating putting that in a separate repo, thoughts?)
[chrisaldrich] joined the channel