#microformats 2019-08-26

2019-08-26 UTC
qotta_ joined the channel
#
[vendan] yeah, looks like python is keeping `7:00am`, and not changing it to `07:00am`
#
[vendan] they are the same value, effectively...
#
KartikPrabhu there are also others. I'm making a list :P
#
KartikPrabhu yeah the datetime normalization was a pain to write up in mf2py
#
[vendan] time is just a pain, no matter what
qotta_ and qotta joined the channel
#
[vendan] just added my fairly out of date go parser as well
JeanDiscord[m] joined the channel
#
KartikPrabhu [vendan]: there is another Go parser https://github.com/willnorris/microformats
#
KartikPrabhu not sure how actively developed it is
#
[vendan] yeah, it's based on my parser
#
KartikPrabhu oh cool
#
[vendan] I might see about pulling some of it back in, but I'm a bit miffed about it being "clone and push new repo" and not just a "fork repo on github"
#
[vendan] but it's nifty to see it being up there as a "unofficial Google" whatever
#
KartikPrabhu oh willnorris is around so maybe it can be sorted out
#
KartikPrabhu lot of the parsers are now in the microformats org on github if you feel like adding the Go one to the list
#
KartikPrabhu https://github.com/microformats
#
[vendan] I'd be fine with that
[Michael_Beckwit and [tantek] joined the channel
#
[tantek] mf2py should be released with the img alt support *not* behind a flag. that's a worthy update
[vendan] joined the channel
#
[vendan] so, it's somehow working... better then it should be?
#
[vendan] but my mf2.andyleap.dev is now able to calculate a "base" set of data between all the parsers
#
[vendan] and it's able to go "past" the minimum set
#
[vendan] it's goal is basically the "base" should either be a superset or subset of all the outputs of the parsers
#
[vendan] (it can be a superset of one, and a subset of another)
#
[vendan] ok, figured out the flaw, now it should be working correctly
#
[vendan] I'll probably pull my parser out for now, cause it's throwing the results off a bit (cause it doesn't support various things)
#
[vendan] fun thing that means is it's even easier to check to see if the parsers are in agreement with each other about the test being wrong!
#
[vendan] just select the test and then hit parse. If they both flip back to blue, they agree with each other
#
KartikPrabhu regarding the h-event/ampm test the spec does not say that an "am" value should be normalized to include a leading "0" so "7am" would become "7:00" which seems wrong. ref: http://microformats.org/wiki/value-class-pattern#Date_and_time_parsing
ZapierDiscord[m] and [vendan] joined the channel
#
[vendan] the standard I've always followed for that kind of time display stuff is that HH means leading zero and H means no leading zero
#
KartikPrabhu yeah. the spec only uses the "HH" notation to demand that "12" be "00"
#
[vendan] this lines up with the minute, as MM -> 00, not 0, (though 7:0am would be a horrifying format to print for the time...)
#
KartikPrabhu [vendan]: I agree. I only think the spec should be more specific
#
[vendan] 👍
#
KartikPrabhu actually, the spec does not even say to match a "H:MMam" so in some sense a "7:00am" value should be ignored!!
#
[vendan] `HH is the 24 hour "hours" in the time, from 00 to 24, with optional leading 0 for values less than 10.`
#
[vendan] that's just in the parsing section
#
[vendan] I think that def. means allow `7:00am`, but the wording is a little confusing
#
KartikPrabhu yes. but the value-class-pattern spec lists a whole bunch of patterms to match but does not have any "H:MMam" in it
#
KartikPrabhu but I agree "7:00am" is valid and should be parsed as "07:00" and similarly a "7:00pm" should be parsed as "19:00"
#
KartikPrabhu so that would need a change in the spec, and in mf2py
gRegorLove joined the channel
#
KartikPrabhu so on the whole, apart from whitespace and datetime format stuff it seems mf2py is doing more than the tests want
#
[vendan] which is probably more on the "tests need to be updated" side, I think?
#
KartikPrabhu yeah, but the extra stuff mf2py is doing is not in the spec :P I have to find the issue/discussion for why that was put in :P
#
[vendan] there's also some that I think is still in flux, like the language/alternates stuff
#
KartikPrabhu haven't done this in 2 years
#
KartikPrabhu yeah the "alternates" stuff is what I was looking at
#
[vendan] heh, I've been out of it for a bit over 4 now... 😞
#
KartikPrabhu well then good to have you back. this tests vs parser site has been very useful :)
#
[vendan] heh, was a lot of fun to put together
#
[vendan] I like the backend side of things... frontend, not so much
#
KartikPrabhu is opposite
#
[vendan] (thank heavens bootstrap and similar stuff makes non-horrifying frontends somewhat closer to possible for folks like me)
#
KartikPrabhu do-not-start-me-on-bootstrap ;)
#
KartikPrabhu mostly though, whatever works
#
[vendan] ehhh, I'm not a huge fan, and I just know I know enough css/js to cause mass havoc 😄
#
[vendan] I tend to write a lot of TUI stuff for my personal use, web stuff like this is just more easily accessible on the internet and such
#
KartikPrabhu what's TUI? sorry non-dev here :P
#
KartikPrabhu Terminal-UI ?
#
[vendan] yeah, console based UIs
#
[vendan] it's one of the few really cross platform UI paradigms, cause basically every system has a terminal
#
KartikPrabhu true
#
KartikPrabhu steeeeeeeeeeeeep learning curve though
#
[vendan] (there's electron and such, but I refuse to use anything that pulls that much extra cruft in)
#
[vendan] ehhh, CLI's have a steep learning curve
#
[vendan] TUI's don't have to
#
[vendan] they tend to have a lot of keyboard shortcuts, but they can also have mouse based interfaces and menus and such like you might expect out of a GUI
gRegorLove joined the channel
#
KartikPrabhu fair
#
[vendan] for instance: https://raw.githubusercontent.com/andyleap/editor/images/editor.png is the editor I use
#
[vendan] you can click `File` and it's got a menu that drops down with open/save/exit and such
[snarfed] joined the channel
#
KartikPrabhu oh great! now I have to redo my virtual environment to work on mf2py
[Rose] joined the channel
#
Zegnat Just a quick scrollback as I was mentioned: textContent has a strict and easy definition, KartikPrabhu, I think you were thinking of innerText
#
Zegnat innerText looks a lot more like what people may expect (with linebreaks kept inside <pre>, paragraphs looking like paragraphs, that sort of thing), but also depends on CSS specifications rather than just DOM (and HTML)
#
Zegnat [vendan] if you are interested in checking out whitespace stuff, https://github.com/Zegnat/php-innertext tries to document some assumptions that may be needed for innerText in an mf2 parser context, and https://github.com/sknebel/python-innertext is a matching (as in: same tests pass) Python implementation
[grantcodes], [Rose], gRegorLove, qotta, [KevinMarks], [jgmac1106], galaxie, [tantek], [schmarty], [jackjamieson], [snarfed], bupkes, [vendan] and [Michael_Beckwit joined the channel
#
@LeahNeukirchen ↩ī¸ http://microformats.org/wiki/gender-formats (and note that this doesn't address the pronoun issue at all) (twitter.com/_/status/1166099149005176835)
#
@LeahNeukirchen ↩ī¸ ok, this is getting better: http://microformats.org/wiki/h-card-brainstorming#Pronouns (twitter.com/_/status/1166099692293414912)
[fluffy], gRegorLove, [snarfed], [KevinMarks], [schmarty] and [Michael_Beckwit joined the channel
#
@BekoPharm Hammered my blog into shape with #SemPress theme by @pfefferle for proper #microformats support used especially on the #IndieWeb. https://wordpress.org/themes/sempress/ (twitter.com/_/status/1166132522733965312)