#microformats 2018-05-30
2018-05-30 UTC
voxpelli_, [cleverdevil], [jgmac1106], ChanServ, tantek_, pniedzielski[m], schmarty, wakest, [Natris1979] and KartikPrabhu joined the channel
# 
@syedrasel1234 #Full #Width #templates #Microformats
#Threaded #comments #Translation @Readygov
Photoshop & Illustrator any Work)
Contact Us :https://goo.gl/VLuTPY (twitter.com/_/status/1001680891691839489)
# 
gRegorLove For https://github.com/indieweb/php-mf2/issues/176, part of the issue appears to be the h-cite has no u-url.
# gRegorLove I'm reading http://microformats.org/wiki/microformats2-parsing##else+if+it's+a+u-* and it's not clear what to do when it doesn't have a u-url... fall back to parsing as p-* I think
# gRegorLove Aside from the stray example.com/foobar, which I know shouldn't be there. Still digging into that.
# 
KartikPrabhu "else use the parsed property value per p-*,u-*,dt-* parsing respectively " I think that means parse as u-* no?
# KartikPrabhu it is a bit ambiguous rule
# gRegorLove Actually hang on, that's not the part I meant to read, heh.
# KartikPrabhu in any case, the property should be a "value: something" instead of "something"
# gRegorLove Sure, aside from that though.
# gRegorLove KartikPrabhu, why does mf2py have "value": "" for the nested h-cite?
# gRegorLove When parsing that HTML
# KartikPrabhu I think it is parsing it as a u-* and goes all the way to the textContent rule
# KartikPrabhu I am now very confused what the "else use the parsed property value per p-*,u-*,dt-* parsing respectively " is supposed to mean
# gRegorLove I think it means jump down the page to "parsing a u- property"
# gRegorLove And agreed, looks like textContent.
# KartikPrabhu yeah, so by that interpretation mf2py seems to be doing the correct thing
# gRegorLove Which I think there's a separate mf2 parsing issue clarifying that algorithm? Heh
# KartikPrabhu is there?
# gRegorLove https://github.com/microformats/microformats2-parsing/issues/15 though maybe doesn't apply here
# KartikPrabhu yeah that is different
# gRegorLove So DOM textContent doesn't replace an img with alt, that's just an mf2 parsing convention for p-*?
# KartikPrabhu that is more like what the textContent should be
# KartikPrabhu I thought that was the spec so I coded it that way :P
# gRegorLove Not questioning it, I just wasn't sure, haha
# KartikPrabhu yeah. I think mf2py is doing things as currently in spec but feel free to point out wrongness
# KartikPrabhu but in any case the URL shouldn't make it in there right?
# gRegorLove Correct
# gRegorLove I thought fixing the 'value' first would be easier.
# gRegorLove Looks like it's part of Zegnat's text parsing updates. php-mf2 is extracting the img alt
# gRegorLove You can try dropping the img HTML in https://wiki.zegnat.net/media/textparsing.html and it gets the alt. I'll defer to him on that.
# KartikPrabhu hmm I think the text stuff is conflicitng with the ordering of the parsing
# KartikPrabhu for getting textContent I think both my algo an Zegnat's should give roughly the same thing
# gRegorLove If you put the <article> html in Zegnat's parsing it returns "this is alt"
# KartikPrabhu yeah because it is replacing img with alt or src, which the u-* parsing does not tell you to do
# gRegorLove That's the php-mf2 parser's current understanding of 'textContent'
# KartikPrabhu yeah mf2py only does img replacement when specified
# gRegorLove Which I realize is wrong
# KartikPrabhu it would be weird to get alt as a u-* value. src I can imagine being fine
strugee joined the channel
# gRegorLove I don't think textContent is going to be mf2-aware, though.
# KartikPrabhu gRegorLove: funny that http://pin13.net/mf2-dev/?id=20180530062251688 gives name = ""
# KartikPrabhu since it should be replacing it with the "src"
# gRegorLove one bug at a time :)
# KartikPrabhu lol!
# KartikPrabhu it isn't about textContent being mf2-aware, I think. It is more about what text equivalent makes sense in a mf2 context
# KartikPrabhu I think mf2py can handle the replace with "src" but not the "alt" case
# KartikPrabhu with minor modifications
# KartikPrabhu just like it fixes whitespacing depengin on the HTML element
# gRegorLove idk. I get confused every time I dig in-depth into the differences between textContent and the mf2 rules to parse p-*
# KartikPrabhu me too! :P
# gRegorLove So I'm not thinking about it much at the moment. Zegnat groks that a lot better
# KartikPrabhu so at the least that property should go as "value"
# KartikPrabhu that is a parsing bug and not a textContent bug
# KartikPrabhu basicaly any "u-*" there should give the same result
# gRegorLove I'm getting tired so I'm not following what you're saying. All I'm saying is "this is alt" comes from Zegnat's textContent algorithm
# KartikPrabhu gRegorLove: look at this example in contrast http://pin13.net/mf2-dev/?id=20180530063455820 I have changed the first u-photo to a u-like-of and that "spurious" thing is gone
# KartikPrabhu the "this is alt" is not the problem but the "http://example.com/foobar" in the first photo property is
# gRegorLove I'm working on the latter. And mf2py and php-mf2 disagree about the former.
# KartikPrabhu aah yes. they do. and there I think mf2py is follwing the spec to the letter
# gRegorLove That example is also wrong. photo from the h-cite showing up in the h-entry.
strugee joined the channel
# gRegorLove Retiring for the night soon, but I think I'm making progress on the implied photo issue!
# KartikPrabhu no that is implied photo :P
# gRegorLove what. Which of the implied parsing rules does it fall under?
# gRegorLove I read `else if .h-x>:only-child:not[.h-*]` as disqualifying this HTML from implying the photo
# KartikPrabhu step 5 of implied photo for the h-entry ;)
# KartikPrabhu "else if .h-x>:only-child:not[.h-*]>img[src]:only-of-type:not[.h-*], then use that img’s src for photo "
# gRegorLove If the implied photo is correct there then it should also be correct on https://github.com/indieweb/php-mf2/issues/176
# KartikPrabhu no, the photo is not implied if there is a u-photo already
# gRegorLove ok, true.
# gRegorLove But second HTML snippet, " .h-x>:only-child:not[.h-*]" stops the implied parsing. It's a nested h-*
# KartikPrabhu this example is more complex than I wanted :P
# KartikPrabhu that is only for "name"
# KartikPrabhu there are rules for stopping implied name but none for stopping implied photo or url
# gRegorLove I'm saying step 5 for implied photo does not match, because: .h-x>:only-child:not[.h-*]
# gRegorLove Step 6 has the same beginning of the rule, so it doesn't match either.
# gRegorLove Thus no gotten value
# KartikPrabhu oh! hmmm
# KartikPrabhu good point!
# KartikPrabhu ok so there is a bug in mf2py too
# KartikPrabhu which somehow has never been an issue!
# gRegorLove Think I've got this about fixed in php-mf2. Pretty sure it's an xpath issue in parseImpliedPhoto().
# gRegorLove I'll do some testing when I'm more awake tomorrow :)
# KartikPrabhu cool
[markmhendrickso joined the channel
[jgmac1106], [kevinmarks] and barpthewire joined the channel
[kevinmarks], nitot, adactio and [jgmac1106] joined the channel
barpthewire, nitot, [kevinmarks], [jeremycherfas] and [sebsel] joined the channel
# gRegorLove Zegnat, could you check into the h-cite value parsed in https://github.com/indieweb/php-mf2/issues/176? mf2py gets an empty string instead of "this is alt"
chrisaldrich, adactio and tantek joined the channel
# 
@kevinmarks #opensourcenortheast (@ Campus North in Newcastle Upon Tyne w/ @microformats) https://www.swarmapp.com/c/fzGrsizJLCV (twitter.com/_/status/1001877676867686400)
tantek and [tantek] joined the channel
# 
[jgmac1106] can you have a ton of h-cite on one page if they are in one h-entry?
# [jgmac1106] On a syllabus for example I may have 20-30 citations
# gRegorLove Certainly!
# [jgmac1106] cool
gRegorLove_ joined the channel
# [jgmac1106] though it was so daunting I haven't added it my syllabus template yet: https://github.com/jgmac1106/Syllabus/blob/master/index.html just on my course sit:e on line 57: https://github.com/jgmac1106/edu305/blob/master/modulethree.html
# [jgmac1106] need a reverse parser where I can copy an APA citation from Google Scholar and have it spit out the correct mf2, so somebody fly around the earth like Superman and make thigns work backwards. Proper APA citation and h-cite markup is a bear
# [jgmac1106] ...and before I get nerd corrected...yes Iknow that was about time travel and a bizzaro world would be better analogy
# [jgmac1106] and also since it was in the movie it was never canon anyway, but still felt like Superman when I figured out mf2 and APA
# [jgmac1106] Google Scholar doesn't have APIs.grrr what happens if you click on this: https://scholar.google.com/scholar?cites=12954661913662618326&as_sdt=8005&sciodt=0,7&hl=en#d=gs_cit&p=&u=%2Fscholar%3Fq%3Dinfo%3AKwj0nyNW1r4J%3Ascholar.google.com%2F%26output%3Dcite%26scirp%3D0%26hl%3Den
# gRegorLove_ An h-cite builder like the hCard builder would be cool.
# [jgmac1106] be really cool if you coudl get h-cite listed on the bottom of the citation modal in Google Scholar...but I am sure those companies pay big bucks to get there
# [jgmac1106] for new pubs the doi # would make this pretty straightforward as I think about it
# [jgmac1106] but that just links yopu back to the publishing silo actually
# [jgmac1106] click on the bibtex file everything is there that you would need
[Natris1979] joined the channel
# 
@jgmac1106 ↩️ Part of the reason I am working on a syllabus template using microformats2. The kind of readers I am seeing folks start to prototype will be awesome learning tools. https://github.com/jgmac1106/mf2syllabus see it on Glitch as well http://syllabus.glitch.me/ (https://jgregorymcverry.com/2975-2/) (twitter.com/_/status/1001894115552919552)
# [jgmac1106] @zegnat I may just try to map it out using a workflow of going to Google scholar>copying the BibTex and throwing that into a modal with a selection for citation style. AssumingI would just need to make the proper mf2 template for each style.
barpthewire joined the channel
# @jgmac1106 ↩️ Yeah my syllabus was never complete just the template.
The course is raging at https//:http://edu305.jgregorymcverry.com. I was using the experience to learn microformats and CSS Grid so it ain’t pretty.
While the Holy Grail (would love a more inclusive as in… https://jgregorymcverry.com/2979-2/ (twitter.com/_/status/1001899022125330432)
# @jgmac1106 @jaytiesse Just came across your and the @metalabharvard project in a thread about Open Syllabi. I have been pushing folks to keep syllabi as plain html as possible and use microformats2 versus more complex markup suggested by http://schema.org… https://jgregorymcverry.com/2984-2/ (twitter.com/_/status/1001903093393645570)
# [jgmac1106] kind of. I forgot to write https in a link in the note, the second is a truncated link to the rest of the note
# [jgmac1106] now to go edit a tweet and the post
# [jgmac1106] actually I just wrote schema.org either bridgy or Twitter added the http://
# 
Zegnat tantek: I think the consensus between sknebel, ancarda, and me, is that the theoretical problem of thundering herd is always going to exsist for something like readers. If everyone hosts their own reader, once you send a WebSub ping into the world that there is a new post, they are all going to wake up and come at you.
# Zegnat But there are several technical mitagations we could think of. Usual caching, of course. E.g. making your website resist DaringFireball/Slashdot. Another thing that would be interesting is to make readers not respond to a ping immidiately (thus staggering the load from the requester side), but you can never be 100% sure that every reader is doing that.
# [jgmac1106] So I use flickr CC images alot. If I am using them in an h-entry can folks let me know if I added u-photo to img tag correctly here: https://github.com/jgmac1106/flickr-cc-helper/commit/88d6c361812a5ed543c4dfdd33f113dcc932f2d4 I was going to try and make it more of an h-cite format...but baby steps
# [jgmac1106] okay, that is what I misunderstood
# [jgmac1106] p-author
# [jgmac1106] opps jinx
# [jgmac1106] okay will keep working on the bookmarklet to see if I can have it spit back an h-cite
# [jgmac1106] These are images embedded in my posts, technically a repost but really just an image that I dig that is also openly licensed. Just pull it in from Flickr API using Alan Levine's tool. I think I will be able to maker an mf2 version
# [jgmac1106] okay will share it there when I mess with the plug-in
sebsel and [kevinmarks] joined the channel
# 
[kevinmarks] So is it a u-featured image?
# [kevinmarks] In jgmac's context - one he wants to represent the post
[grantcodes] joined the channel
[wiobyrne], [Vanessa], [chrisaldrich], KartikPrabhu, tantek and [jgmac1106] joined the channel
# [jgmac1106] I think in my context yeah I would use u-feature as it it kind of the "hero" image of the page...especially if SM treats that like a twitter card I would add to my header
# [jgmac1106] http://edu305.jgregorymcverry.com/modulefour.html Ignore the image size thats a me learnign css grid thing and not even getting bad inline styling to override'
# [jgmac1106] hmm. I think I need to add an h-card on every page above a featured image
# [jgmac1106] really an h-entry is broken and needs an h-card
# [jgmac1106] its not, I forgot it. I am broken
# [jgmac1106] Yeah I messed up, I should be adding headers for accessibility purposes after <main>
# [jgmac1106] yeah total oversite, and I hammer my students on that
# [jgmac1106] been following since
# [jgmac1106] doing your instructional design while also teaching is such a rush. Why I want to spend this summer getting my tempaltes perfect
# [jgmac1106] what is name or irc channel I always mess it up...but here is where I ended up: https://gist.github.com/jgmac1106/6bef1656ab4761405d62350825f53f95
KartikPrabhu, [kevinmarks], tantek, [chrisaldrich] and [jgmac1106] joined the channel