#microformats 2018-03-26

2018-03-26 UTC
#
gRegorLove KartikPrabhu, Is that in your github fork? I was going to try the tests you used.
#
KartikPrabhu gRegorLove: https://github.com/kartikprabhu/mf2py/blob/25560c7ba27ab75a88414657a7cbc1e14b3d0559/test/examples/datetimes.html#L132-L141
#
KartikPrabhu note that it is a leap year so the date should be "2016-03-02"
#
gRegorLove Ohh, I was looking in the wrong branch. Thanks!
#
gRegorLove Yeah, I'll make sure there's tests with and without leap year
#
KartikPrabhu it is also in "experimental" but I decided to fix version 1.1.1
#
gRegorLove I put it for milestone 0.4.3 for php-mf2 - release after next.
#
KartikPrabhu I am keeping "master" branch fixed with the main repo PRs
[miklb], matigo, [cleverdevil], [kevinmarks] and [pfefferle] joined the channel
#
Zegnat Nice to see clarification on the datetime value. If the reason is to make it simpler for the json consumer, I wonder if it's worth just saying to always normalise to an HTML datetime value. Presumably that's what people expect as non-vcp dt-* parsing is usually from time elements.
#
Zegnat https://html.spec.whatwg.org/#datetime-value
#
KartikPrabhu no I think mf2 properties should be as authored
#
KartikPrabhu vcp stuff is different I think
#
Zegnat To always normalise vcp output to an html datetime value then. Basically instead of having the list of example results, we can link to a spec that specifies then explicitly is my thinking
#
KartikPrabhu the current spec does that no?
#
Zegnat I am probably overthinking it. Just feel like we should be able to make use of the fact that HTML includes datetime parsing rules now 😅
#
[kevinmarks] Lol https://twitter.com/lynncyrin/status/977369342147936258?s=19
#
Loqi [@lynncyrin] *points out the failing tests* Is THIS your team? *rolls out all the linter errors* Is THIS your team????? *merges my PR, and all the status indicators turn green* pft, naw. I'm your lead engineer now
nitot_ joined the channel
#
Zegnat goes back to plain text parsing and leaves dates to others
barpthewire, [gerwitz], KevinMarks_, nitot, KevinMarks and nitot_ joined the channel
#
paketwisata edited /get-started () "(-3536)" (view diff)
#
paketwisata edited /get-started (-2) (view diff)
#
zegnat edited /get-started (+3538) "Reverted edits by [[User:Paketwisata]] ([[User_talk:Paketwisata|Talk]]) to last version by [[User:Tantek]]" (view diff)
#
Zegnat adminhelp: block user Paketwisata on the wiki?
nitot_ joined the channel
#
@ForestGdnWales The blog post pages are marked up with "microformats", so that the pages preview nicely in the different social networks, I used these references: https://moz.com/blog/meta-data-templates-123 and http://microformats.org/wiki/h-entry (twitter.com/_/status/978206908015370241)
#
@ForestGdnWales 3/4 The blog post pages are marked up with "microformats", so that the pages preview nicely in the different social networks, I used these references: https://moz.com/blog/meta-data-templates-123 & http://microformats.org/wiki/h-entry (twitter.com/_/status/978207391593517056)
#
Zegnat Ha, those quotes around microformats! ^^^
[jeremycherfas] joined the channel
#
Zegnat Alright, per change control, I have done a parser implementation before pushing for further spec changes: https://github.com/indieweb/php-mf2/pull/168
#
Loqi [Zegnat] #168 New algorithm for plain text values
#
Zegnat adminhelp: how do I get promoted on the microformats wiki so I can create a page? I would like to put the plain text values algorithm on the wiki for further iterations and brainstorming.
KevinMarks, kaushalmodi, Garbee, webchat362, [eddie], [kevinmarks], KevinMarks_, nitot, nitot_, barpthewire, [kim_landwehr], gRegorLove_, [cleverdevil], [Mike], [pfefferle] and [cb] joined the channel
#
aaronpk reran the test and it passed
#
Loqi aaronpk: gRegorLove left you a message 20 hours, 52 minutes ago: Can you add Zegnat to the indieweb github team so we can assign reviews?
#
Loqi aaronpk: gRegorLove left you a message 18 hours, 32 minutes ago: updated https://github.com/indieweb/php-mf2/pull/164, can you run CI again? PHP5.6 didn't get the updated test somehow https://travis-ci.org/indieweb/php-mf2/builds/358164691
[cleverdevil] joined the channel
#
aaronpk and added Zegnat
#
Zegnat wonders what powers he gained
tantek joined the channel
#
gRegorLove_ Woo! https://github.com/indieweb/php-mf2/milestone/1
#
Zegnat Time to get a release out? :)
#
tantek any more easy issues to resolve first? or everything else deferred til after the release?
KevinMarks joined the channel
#
Zegnat All the issues and PRs marked for 0.4.2 are checked off. The 2 standing PRs are deferred.
#
tantek great!
KevinMarks_ and nitot joined the channel
#
gRegorLove_ VCP ordinal dates should be easy (knock on wood) but I think it's lower priority to the vendor prefix fix, so I slated it for 0.4.3
#
gRegorLove_ https://github.com/indieweb/php-mf2/milestone/2
#
gRegorLove +1 for releasing php-mf2 0.4.2
[miklb] joined the channel
#
tantek I will be happy to change my markup to test VCP ordinal dates when they're ready in at least php and python parsers
#
tantek to help drive updates to clients of those parsers
[cb] joined the channel
#
Zegnat Interesting question: how many parsers must implement it before you change your markup accordingly, tantek? ;)
#
tantek at least whatever aaronpk and Loqi are dependent on (PHP), then one more (hence likely Python)
KartikPrabhu and KevinMarks joined the channel
#
Zegnat Stable version of Go already has it, so that could be the +1 after PHP.
#
tantek Zegnat: I should say, one more that has at least one client (publicly verifiable) that uses it
#
Zegnat Fair!
KartikPrabhu, kaushalmod, [colinwalker], nitot, barpthewire and kaushalmodi joined the channel
#
tantek edited /advocacy (+337) "online tutorials, start with a stack overflow answer that could be improved" (view diff)
#
Zegnat I actually had an SO answer come out of one of my php-mf2 patches. But it hasn’t been accepted as of yet :(
#
KartikPrabhu is confused what's the actual algorithm for the whitespace stuff?
#
KartikPrabhu a simple replace <br> by \n is not it
#
Zegnat You mean https://wiki.zegnat.net/media/textparsing.html ?
#
Zegnat I think https://github.com/microformats/microformats2-parsing/issues/15 captures the background pretty well.
#
Loqi [Zegnat] #15 What should mf2 textContent parsing result in? User expectation vs. DOM specification.
#
KartikPrabhu judas priest! that is a lot of work for getting text
#
Zegnat Should be easier to build than the actual HTML innerText: https://html.spec.whatwg.org/multipage/dom.html#the-innertext-idl-attribute
#
KartikPrabhu not if you have to write your own code
#
Zegnat Do you have access to innerText in Python?
#
Zegnat (not textContent, mind you)
#
KartikPrabhu i have access to some sort of text
#
Zegnat But do you have access to text where BR and P elements introduce line breaks, and other HTML whitespace is collapsed?
#
KartikPrabhu this is a lot of iterations of children which the HTML parser text-thing already does
#
KartikPrabhu no
#
Zegnat That’s what the issue was about. Users filing issues on parsers because the DOM tree’s textContent property didn’t match their expectations (e.g. <p>A</p><p>B</p> becomes "AB" without any whitespace). And standard DOM implementations like the one in PHP do not have HTML’s innerText available for a prettier output.
#
Zegnat We could try to normalise on HTML’s innerText, but that assumes CSS knowledge, which is often also lacking when you are working only with HTML parsers and DOM trees. So had to figure out something else.
#
KartikPrabhu will wait for this to make into some sort of spec
#
Zegnat The algo I wrote is based on expectations aaronpk colated and some simple CSS whitespace collapsing rules. It should also be relatively easy to implement, I have already done both a JS and PHP implementation.
#
KartikPrabhu it is not hard to implement, just that we would be iterating over the DOM tree so many times
#
Zegnat You are already doing that to some extend, as you have to walk the tree to replace IMG elements
#
KartikPrabhu yes to some extent
#
KartikPrabhu here we are walking over everything
#
KartikPrabhu anyway
#
Zegnat I’d love to see alternative proposals to #15 KartikPrabhu! You are a parser implementer, so please do weigh in with your opinion :)
#
KartikPrabhu I ma not very convinced about having the same text in all parsers as being useful
chrisaldrich, [stefp], [sdepolo], j12t, j12t_, KevinMarks, kaushalmodi, [asuh] and [kevinmarks] joined the channel