#dev 2017-11-06

2017-11-06 UTC
[eddie] joined the channel
#
[eddie] aaronpk++ for the social icons repo! I use font-awesome and have had to make swarm be foursquare’s icon. Now I can fix that!! ??
#
Loqi aaronpk has 84 karma in this channel (1464 overall)
#
aaronpk haha yay
#
Loqi aaronpk: lol
#
GWG aaronpk: You inspired me to add more icons to my lists.
tantek and [kevinmarks] joined the channel
#
[kevinmarks] A huge number of the svgs uploaded to svgur.com are app icons
#
[kevinmarks] Also, see https://github.com/edent/SuperTinySocialIcons/
tantek, gRegorLove, KartikPrabhu, renem and [keithjgrant] joined the channel
#
[keithjgrant] aaronpk SVG is far more flexible for styling. If it's inlined, CSS can target individual nodes and style them: fill color, stroke style & color, animation. Also better accessibility.
tantek joined the channel
#
[keithjgrant] More flexible embedding options, too
#
aaronpk It looks like icomoon does a good job of setting up the svg sprite thing and it's only a little bit harder to use than the icon font
#
aaronpk I'll have to add some css rules to style the color of the icons specifically in every place they're used, rather than them picking up the same text color
#
KartikPrabhu aaronpk: you can change the colour of the icons itself
#
KartikPrabhu inside the SVG
tantek joined the channel
#
aaronpk I️ want it to have different colors depending on where it's used, for example in a link vs not
#
oodani Huh, that's interesting. I just noticed that microformats2 parsers handle my <script class="p-json-ld" type="application/ld+json"> elements differently to each other? mf2py does exactly what I'd expect, which is take the full JSON text as a property of the containing h- element. However, pin13 produces an *empty* string as the property's value, and go.microformats.io just completely omits the property. Which of those behaviours is considered correct?
#
oodani Judging from the spec http://microformats.org/wiki/microformats2-parsing, I think mf2py is doing the right thing: to parse a p- property, you only drop *nested* <script> elements, not the p- element itself.
#
Loqi [Tantek Çelik] microformats2 parsing specification
#
oodani (And yeah, JSON-LD is redundant, but Google wants it - so I figured it'd be better to provide it in a way microformats2 can cope with. ? )
#
KartikPrabhu oodani: no mf2 consumer will look for that prpoerty anyway
#
oodani Heh. You're not wrong.
#
KartikPrabhu also providing JSON-LD version inside a mf2 -property is needlessly redundant
#
oodani Yep, just mentioned that.
#
KartikPrabhu aaronpk: aah then yes CSS is your friend. also "color: currentColor" makes the icon inherit the font color f its parent
#
KartikPrabhu oodani: so yeah, I am confused what you are trying to do
#
oodani Trying to figure out which microformats2 parsers are buggy and how to fix them. This is just an example that happens to expose the difference in behaviour.
#
KartikPrabhu oodani: testing on unused non-standard properties seems not very useful
#
oodani Isn't kinda the point of microformats2 that it's arbitrarily extensible, since the class prefixes indicate how to retrieve the data regardless of whether the parser actually knows what the property is or not?
#
KartikPrabhu yes. but testing on properties that no one actually consumes seems moot
#
KartikPrabhu also in your example the p-* will make the JSON-LD into a plain text string so it won't be consumed as a JSON data anyway
#
oodani Yeah, I know, that's exactly what happens with mf2py.
#
oodani Other parsers do different things. They should really all be doing the same thing.
#
KartikPrabhu oodani: having a made-up counter-example is not going to motivate anyone to fix the differences
tantek joined the channel
#
oodani Okay. I just, y'know, happened to parse my site, which has this property on it, and noticed different parsers produced a different result. Thought someone might wanna know.
#
aaronpk That's kind of a hilarious example
#
aaronpk The fact that all the parsers do something different is interesting, but that's more a conversation for #microformats if we're just talking about spec level stuff with no actual consuming use case
#
oodani Ah, hadn't realised there was a specific channel for spec stuff. :)
#
gRegorLove What's your site, oodani? I work on the php-mf2 parser so my curiosity is piqued now. :)
#
oodani https://00dani.me - it's got the hilarious p-json-ld properties on both my h-card and all h-entries. :3
sebsel joined the channel
#
gRegorLove Yeah, found why php-mf2 is blanking it out. innerText() is a function called recursively to get a plaintext string. It skips <script>
#
gRegorLove Which might make more sense only on child elements
#
oodani Yeah, it probably shouldn't be skipping the whole root element if it happens to be a <script>.
Ruxton, tantek, [kevinmarks], amz3`, ben_thatmustbeme, [xavierroy], cweiske and [pfefferle] joined the channel
#
[pfefferle] good morning
#
cweiske hiho
#
Loqi rise and shine!
myfreeweb, zoglesby, TheGillies, schmarty, plindner, AlanPearce[m], mindB, eli_oat[m] and barpthewire joined the channel
#
@00d4n1 i just implemented support for sending webmentions! https://nya.as/eK (twitter.com/_/status/927486998557814785)
[kevinmarks], [pfefferle], treora, ScalaWilliam, tantek, jeremycherfas, snarfed, John___ and clintpatty joined the channel
#
jeremycherfas.net edited /Facebook (+553) "/* Big data collection allows intrusive inference */ Added link to Reply All episode about spying. --~~~~" (view diff)
dougbeal|iOS, gRegorLove, amz3`, snarfed, tantek and [miklb] joined the channel
#
tantek is tweaking his RSVPs display because the CSSWG (# css on irc.w3.org) discussed auto-sizing iframes!
#
aaronpk ooh
j12t_ joined the channel
#
snarfed myfreeweb++ for using https://fed.brid.gy/ !
#
Loqi myfreeweb has 3 karma in this channel (20 overall)
#
Loqi Ryan Barrett
#
snarfed eg https://unrelenting.technology/replies/2017-11-06-18-29-22 replying to https://octodon.social/@00dani/98957074067239839
#
Loqi [Greg] nice! looks like you don't have receiving yet though…
tantek joined the channel
#
myfreeweb snarfed++ for making https://fed.brid.gy/ :)
#
Loqi snarfed has 17 karma in this channel (312 overall)
#
Loqi Ryan Barrett
clintpatty, snarfed, plindner, John___ and [chrisaldrich] joined the channel
#
unrelenting.technology edited /Firefox_Share (-468) "R.I.P. Firefox Share. Removed in version 57." (view diff)
#
unrelenting.technology edited /Firefox_Social_API (+81) "it's dead" (view diff)
KartikPrabhu joined the channel
#
unrelenting.technology edited /Firefox (+61) "removed features" (view diff)
#
aaronparecki.com edited /h-review (+111) "/* See Also */" (view diff)
[eddie] and tantek joined the channel
#
GWG Good day
#
tantek !tell snarfed when you have a moment - mind trying out fed.brid.gy with https://activitypub.rocks/news/test-suite-running-and-impl-reports-page-up.html ? And if there's *anything* blocking you from filing an implementation report, definitely want to hear about it!
#
Loqi Ok, I'll tell them that when I see them next
myfreeweb and j12t joined the channel
#
tantek aaronpk yeah I did, as Note 2 here https://indieweb.org/payment#Payment_page_with_amount
#
tantek perhaps I'll make that its own heading
#
tantek ends up brainstorming some more
snarfed joined the channel
#
tantek.com edited /payment (+541) "/* Payment page with amount */ Payment page with reason, maybe without amount" (view diff)
plindner, eli_oat[m], mindB, AlanPearce[m], schmarty, TheGillies and zoglesby joined the channel
#
tantek aaronpk: https://indieweb.org/payment#Payment_page_with_reason
#
snarfed tantek: sure! it's on my todo list. low priority right now because bridgy fed is just server to server, which https://test.activitypub.rocks/ doesn't actually test at all. it's just an interactive form that generates an implementation report. so, kinda low value for me right now
#
Loqi snarfed: tantek left you a message 23 minutes ago: when you have a moment - mind trying out fed.brid.gy with https://activitypub.rocks/news/test-suite-running-and-impl-reports-page-up.html ? And if there's *anything* blocking you from filing an implementation report, definitely want to hear about it!
#
snarfed still, will do
#
GWG Hello, snarfed
#
tantek hmm - I think that's the intent, that you're supposed to follow the steps it asks you to and then report back on the results you're getting / seeing
#
tantek a bit of honor system rather than server-side tested
#
tantek still useful
#
snarfed tantek: eh kind of. it doesn't even have test data, urls or objects or pages or whatever. just abstractly describes features.
#
tantek hmm - abstractly sounds a bit bad
#
snarfed so, sure, it's a nice implementation report generator, but i wouldn't call it a test suite, at least the server to server part
#
snarfed hey GWG
#
tantek snarfed, can you file that as an issue? I take your word for it, just want to capture it so the folks in the WG know it's an issue
#
aaronpk yeah it's more like the webmention/micropub implementation report checklists but spread out in a list where you have to click each one to progress further
#
snarfed tantek: um, maybe? it seems like it's working as designed. not sure what i'd ask for...?
#
snarfed i'll file something asking for real testing
#
aaronpk i already mentioned it to cwebber and he said yes that's expected
#
tantek snarfed, yes, your concerns about what you expect to see / want as far ast testing vs. what you think you're seeing
#
tantek still worth documenting
#
tantek especially if it's enough to disincentivize or deprioritize filing a report
#
snarfed sure. i'll mention https://webmention.rocks/ and https://micropub.rocks/ as admirable examples
#
tantek (which it sounds like it is, which is ok, just need to capture that)
#
snarfed (to be fair, what i'm asking for is a lot of work, so i understand why it's not there)
#
tantek yup. I trust your github issue filing style :)
#
snarfed honestly when i do file an implentation report, it'll be easier to copy/paste the template file and type answers myself :P
#
aaronpk heh
#
snarfed tantek: in return, want to try bridgy fed yourself? you're pretty great at finding bugs too :P setup is just a few redirects in your .htaccess!
#
tantek oh boy
#
tantek maybe this afternoon during the Social CG f2f
#
tantek hey snarfed want to come crash the Social Web Community Group f2f here at W3C TPAC?
#
tantek at the Hyatt SFO
#
tantek today 15:30-17:30
#
tantek you'd get to meet cwebber, primary editor of ActivityPu
#
snarfed lol, hmm
#
tantek ActivityPub*
#
snarfed i met him at iws 2017!
#
tantek no he wasn't there
#
tantek you met AJ
#
snarfed ah!
#
snarfed maybe! looking
#
tantek realizes the meeting is an hour
#
snarfed 15:30-16:30? i can do that
#
snarfed i'm literally a mile away
#
aaronpk haha whoa
#
Loqi rofl
#
aaronparecki.com edited /payment (+91) "/* Payment page with reason */" (view diff)
myfreeweb, eli_oat[m], plindner, AlanPearce[m], mindB, zoglesby, TheGillies, schmarty, KartikPrabhu, clintpatty and tantek joined the channel