#microformats 2018-01-11
2018-01-11 UTC
# 
[kevinmarks] There is a cobblers children aspect to the css there
# [kevinmarks] That it has fixed width margins on my phone
# [kevinmarks] My phone has a 4k screen and I have a near point about 6" from my eyes with glasses off, so I can read it, but that might not be broadly true
# 
KartikPrabhu microformats might not be about CSS advice, but it would be good if the site were easier to read on phones so that more people read it
# KartikPrabhu also, having a "old" look seems to deter people from taking content seriously
# KartikPrabhu yeah, i didn't say "retro"
# KartikPrabhu also, in a largely mobile world fixed width is not great even if it is "Retro"
# KartikPrabhu but, I don't want to go anywhere near styling wikis
tantek, KartikPrabhu, Kyle-K, nitot and [mlopatka] joined the channel; Kyle-K left the channel
# 
sknebel !mf2py http://microformats.org/
# 
Zegnat http://microformats.org/wiki/hfeed doesn’t suggest anything about url properties on feeds to me. The HTML doesn’t seem to have any classes like it either. So I really have no idea where the URL properties are coming from :S
# sknebel (https://github.com/microformats/microformats2-parsing/issues/11 was the issue for mixed mf1/mf2 we had recently again)
[kevinmarks] joined the channel
# [kevinmarks] Python one, yes
# [kevinmarks] I have extra backcompat for WordPress, but looks like they weren't in that theme
# [kevinmarks] It does have 2 identical urls for each post
# [kevinmarks] I wonder if we should change parsing to strip duplicates. The rels parsing does, but it is harder in the broader case.
nitot joined the channel
# 
ben_thatmustbeme Sigh, and no one mentions the most recently (re)written parser... The Ruby one
# sknebel !mf2-ruby http://microformats.org
# Zegnat Interesting. http://microformats.org/wiki/h-feed#Backward_Compatibility says “site-title” is the WordPress thing. If that is wrong, hopefully [kevinmarks] can give it an update!
nitot, adactio, [mail], [kevinmarks], KartikPrabhu, [mlopatka], [pfefferle] and [miklb] joined the channel
# 
gRegorLove This is my latest php-mf2 improvements parsing microformats.org: http://gregorlove.com/php-mf2/test.php?id=39
# gRegorLove rel=bookmark gets upgraded to u-url
# gRegorLove I think the backcompat spec is fine for now, tantek, appears to be a bug in php-mf2 0.3.2 when parsing nested mf1, like hEntry inside hFeed.
# gRegorLove I think T only if authored. That's in a pending PR.
# Zegnat But both T and space are allowed in HTML datetime values, so I would say mf2 follows that and definitely accept both: https://html.spec.whatwg.org/dev/text-level-semantics.html#datetime-value
tantek joined the channel
# gRegorLove Different question, Zegnat. Parser was formerly normalizing to add the "T" even if it wasn't authored that way.
# gRegorLove But the spec says it shouldn't add it if it isn't authored that way
# gRegorLove Understood :) In this example it's not changing the string. The "T" is authored.
# gRegorLove At least in the first example, didn't spot check the other h-entry on the homepage
KartikPrabhu joined the channel
[colinwalker], chrisaldrich, tantek and [keithjgrant] joined the channel
# 
@dissolve333 @Blogger any chance you will ever add microformats-2 to your templates, would love to be able to parse any blogger ever with it (twitter.com/_/status/951528572174336000)
# ben_thatmustbeme pokes the bear
# 
aaronpk oh look, snarfed requested mf2 support back in march https://groups.google.com/forum/#!topic/web-data-commons/WOeSOODtj3A
[kevinmarks] joined the channel
# [kevinmarks] I added mf1 to blogger more than 10 years ago: http://epeus.blogspot.co.uk/2007/08/microformats-in-blogger-hatom-support.html
# [kevinmarks] maybe plindner could have a chat with the blogger team
[mail] joined the channel
# ben_thatmustbeme i think the mf2 processing on apache's any23 is broken
# ben_thatmustbeme it processes part way, but seems to FAIL with errors
[miklb] joined the channel
# ben_thatmustbeme it fails parsing tantek.com and my site certainly
[dariusdunlap] joined the channel
# ben_thatmustbeme or rather it does, but just throws out a random error
# ben_thatmustbeme [Fatal Error] :170:3: The element type "input" must be terminated by the matching end-tag "</input>"
# ben_thatmustbeme yeah
# ben_thatmustbeme running it locally
# ben_thatmustbeme any specific page you want me to try?
# ben_thatmustbeme seems like it just blows up when its malformed html
# ben_thatmustbeme very rigid parsing
# ben_thatmustbeme don't have time to hack on it, not a java person
KartikPrabhu joined the channel
# ben_thatmustbeme don't even know how to do that
[eddie] joined the channel
# ben_thatmustbeme it seems to be using jsoup 1.7.2 by default, trying to build it with maven
# ben_thatmustbeme wow, that was from Jan 27, 2013
[colinwalker] and [kevinmarks] joined the channel
# [kevinmarks] The python one is a bit like that, the parsing depends which lib you are using
# ben_thatmustbeme looks like its here https://tika.apache.org/, this depends on edu.ucar:grib:jar:4.5.5 which is up to v 8.0 (
# ben_thatmustbeme though .... Unidata's NetCDF Decoders package is no longer being actively maintained or supported.
# ben_thatmustbeme taht was for grib which contains jsoup
# ben_thatmustbeme yeah
# ben_thatmustbeme anyway, gotta go catch a train
# ben_thatmustbeme someone should figure out how to point the common crawl people at this
# ben_thatmustbeme i can't imagine they are too happy with how much this must error out
tantek, [mail], [eddie], sebsel, [kevinmarks] and chrisaldrich joined the channel