#microformats 2018-04-15

2018-04-15 UTC
[kevinmarks], nitot, KartikPrabhu, barpthewire, gRegorLove_, [jeremycherfas], [tom_wolf] and [jgmac1106] joined the channel
#
@kevinmarks ↩️ If I fetch your site in response to a webmention, and parse it for an h-card is that processing PII, building a profile? On the other hand, if I republish on my site with microformats comment markup, does that count for data portability? (twitter.com/_/status/985538040214052865)
[miklb], [jgmac1106], KartikPrabhu, [kevinmarks], [jeremycherfas] and [pfefferle] joined the channel
#
@prepanorg Web::Microformats2 by jmacdotorg@github - http://prepan.org/module/nYhnPU5b9bY #prepan (twitter.com/_/status/985591512493834240)
#
@perlnews PrePAN: Web::Microformats2 - Libraries for parsing Microformats2 metadata from HTML or JSON (jmacdotorg@github) - http://prepan.org/module/nYhnPU5b9bY (twitter.com/_/status/985591768929320960)
#
@JmacDotOrg Just published a @PrePANorg RFC for my proposed Web::Microformats2 Perl module. http://prepan.org/module/nYhnPU5b9bY (twitter.com/_/status/985596945342369792)
#
Zegnat That looks nice jmac!
#
jmac Thanks!
#
sknebel do you have a web interface to break.. ehm, TEST it somewhere?
#
Zegnat Not sure there is lots to break, does it do mf1 back-compat?
[miklb] joined the channel
#
jmac sknebel: It skips with a test suite that pushes a copy of github.com/microformats/tests through it. And it passes something over half of those, with the rest marked TODO.
#
jmac s/skips/ships/
#
sknebel sadly, that test suite isn't always right either :/ but good start
#
jmac That's my understanding (particularly with dates, right?) but just getting a start is my goal.
#
jmac Zegnat: It doesn't pay attention to anything other than tags with MF2 attributes right now. (It pretends to read rels for the sake of JSON serialization, but this is a lie.)
#
Zegnat gotcha
#
sknebel yes, dates were one thing. happy to try help if you are unsure about something
#
Zegnat Dates are “normalised” in the tests, to help parsers. But I would argue that leads to the oposite, because the spec is pretty specific about the date string values expected in the JSON output
#
jmac I have a bunch of "Hey, wait a minute" questions about some of the tests that don't seem to quite agree with the spec, but I'm leaving those for later
#
jmac Even as they are they are super-useful in demonstrating the overall flexibility a parser needs, with tons of examples of what MF2 can do
#
Zegnat Feel free to open issues on any tests by even the slightest doubt of their output. Lots of us have wanted to fix them but keep getting sidetracked because it isn’t the most fun job ever. If you are looking at them anyway, best imediately note down which ones are questionable
#
sknebel or make issues on your own parser and ping us or wherever you're most comfortable :)
#
jmac Oh, sure... I'll ask some questions out loud when I'm next in a mode to work on testing again, thanks. I have indeed been making immediate notes.
[snarfed], [jgmac1106] and [kevinmarks] joined the channel
#
[kevinmarks] The python parser "fails" a good chunk of them too
[miklb] and [eddie] joined the channel