#indiewebcamp 2014-07-26

2014-07-26 UTC
#
gRegor` The extra u-uid check in step 2 seems superfluous
#
gRegor` So I think I'm missing something
#
JonathanNeal hmans: i’ll try it again at home, got a complaint about invoker missing https://gist.github.com/jonathantneal/10b2ad57295471693abd
#
KevinMarks_ hm. we shoudl check who wrote that - I think #4 means use the hcard on the page, without going back to the pointed to page for a fuller one
#
hmans JonathanNeal, I promise there will eventually be better documentation. Getting people involved in the project hasn't been a priority so far. :P
#
KevinMarks_ so you'd get an hcard without a photo in that case
#
KevinMarks_ (using my pages as examples)
#
hmans JonathanNeal, it's possible once you sude your `gem` will look in other gem directories than when you're ~jneal
#
gRegor` KevinMarks: Maybe this is ambiguous in #4 "rel-author link's href". I took that to mean the rel-author href from the original page we started at. But maybe it means the current page?
#
KevinMarks_ 4 says "on the post's page"
#
KevinMarks_ so I think it's saying "if you can't find hCards in the other places, use the inline one as a fallback"
#
gRegor` Oh. Then I think it's ordered incorrectly. It shouldn't be after "1. go get the destination page and parse it for microformats2" ?
#
gRegor` Gotcha
#
gRegor` I was reading 2-4 as "after you get the mf2 of the rel-author link"
#
KevinMarks_ clearly this needs calrifying
#
gRegor` :)
#
gRegor` Your interpretation there makes sense
#
KevinMarks_ 2+3 should be indented under 1 I think
#
gRegor` I think I'll do that and ask tantek's review
#
KevinMarks_ and maybe clarify which page is which
#
KevinMarks_ you have to be pretty used to tantek's style to parse it
#
hmans JonathanNeal, http://hmans.io/wxz404 :P~
#
gRegor` :)
cuibonobo_ joined the channel
#
KevinMarks_ hmmm https://angel.co/indiewebs
johncash joined the channel
#
gregorlove.com edited /User:Gregorlove.com/sandbox (+1414) "Draft of improvements to authorship/#Determining" (view diff)
#
gRegor` KevinMarks_: What do you think of this? Nesting those two further made it look more confusing, with all the indented numbers. http://indiewebcamp.com/User:Gregorlove.com/sandbox
#
gregorlove.com edited /User:Gregorlove.com/sandbox (-4) "/* Determining */" (view diff)
#
KevinMarks_ that is clearer
#
gRegor` !tell tantek Wanted to get your feedback on this proposed clarification for /authorship. http://indiewebcamp.com/User:Gregorlove.com/sandbox
#
Loqi Ok, I'll tell him that when I see him next
#
gRegor` !tell tantek context is conversation with KevinMarks here http://indiewebcamp.com/irc/2014-07-25#t1406331176
#
Loqi Ok, I'll tell him that when I see him next
#
gregorlove.com edited /User:Gregorlove.com/sandbox (-10) "/* Determining */" (view diff)
#
gregorlove.com edited /User:Gregorlove.com/sandbox (-1) "/* rel-author parsing */" (view diff)
#
hmans Is this the go-to gem for parsing mf2 when you're a Rubyist? https://github.com/g5/microformats2
#
gRegor` KevinMarks_: the rel-author parsing for http://www.kevinmarks.com/fragmentions.html fails because your rel-authr doesn't hve a trailing slash, but it does on the kevinmarks.com h-card :)
#
gRegor` Pretty minor, though. Wonder if I can safely strip the trailing slash from URLs for the comparison
#
KevinMarks_ also I am fast and loose about kevinmarks.com and www.kevinmarks.com
#
gRegor` :)
#
KevinMarks_ I should probably do href="" on the homepage
johncash joined the channel
#
Loqi fo sho
#
KevinMarks_ !tell tantek would a talk about fragmentions be suitable for that conf?
#
Loqi Ok, I'll tell him that when I see him next
tantek joined the channel
#
Loqi tantek: gRegor` left you a message 19 minutes ago: Wanted to get your feedback on this proposed clarification for /authorship. http://indiewebcamp.com/User:Gregorlove.com/sandbox
#
Loqi tantek: gRegor` left you a message 18 minutes ago: context is conversation with KevinMarks here http://indiewebcamp.com/irc/2014-07-25#t1406331176
#
Loqi tantek: KevinMarks_ left you a message 1 minute ago: would a talk about fragmentions be suitable for that conf?
chrissaad and johncash joined the channel
#
JonathanNeal no excuses hmans we like your software and we’re gonna try it
#
JonathanNeal =)
snarfed and ben_thatmust joined the channel
#
ben_thatmust hey guys
#
ben_thatmust and/or gals
gRegor`, krendil and tantek joined the channel
#
tantek let's see in reverse order
#
tantek KevinMarks: yes, fragmentions would be great to give a talk on, show them working at the HTML5DevConf
emmak joined the channel
#
tantek but I'd say let's work on making /fragmentions look like a proper spec - ok to also use microformats.org/wiki/fragmentions since we have a "spec-like" template/look already setup there.
#
tantek as editor, your choice
#
tantek gRegor`: ironically, it's hard to see the diffs on your sandbox from the original article so I can't tell what you've changed or why
#
tantek so I can't give informed feedback :(
#
tantek I guess I can try to disambiguate /authorship and maybe add some comments / explanatory stuff
Rev_Illo joined the channel
#
gRegor` I can edit the actual page if you prefer.
#
gRegor` Basically I wanted to make sure it made sense.
#
tantek clearly that's what I get for not "showing my work" as it were in terms of explaining the derivations of the steps in the algorithm
#
tantek assumes responsibility
#
gRegor` I separated out the rel-author sub-algorithm, because when it was indented properly it was too many numbers that were confusing.
#
gRegor` Like "4 1 1"
#
gRegor` I also added explicit page references "A" and "B" to disambiguate "the page"
#
gRegor` I didn't think that I should have put the original version in sandbox, then diffed from that. Good idea.
#
tantek gRegor`: in re-reading the pieces of step 4, I did write them as you would code them
#
tantek rather as you would group them in code
#
tantek I've tried *really* hard to avoid any use of A or B variable references in the pseudo-code
#
tantek to make it more prose-like understandable
#
gRegor` 4.1.2 and 4.1.3 should be indented under 4.1, correct?
#
tantek no
#
tantek they require 4.1.1 - it's a literally ordered list that must be executed in order
#
gRegor` 4.1.4 deals with the *post* page, not the rel-author page
#
gRegor` That was my confusion before. I thought it was about the rel-author page after reading 4.1.1.
#
tantek 4.1.4 *requires* that the page has a rel-author link, which 4.1 tests
#
tantek thus 4.1.4 is under 4.1 because 4.1 already did the test!
#
gRegor` I think referring to page A and page B makes it much clearer
#
tantek I don't because then the reader has to decode what A or B mean
#
tantek it's flatter to simply write it out every time
#
tantek requires fewer things to keep in your head from different places
#
gRegor` You have to distinguish two pages either way.
#
tantek however, might help to use the same phrase for each page each time
#
gRegor` :shrug: we have a demonstration of it being confusing as-is. And I'm pretty technical :)
#
tantek and italicize them so they stand out like vars
#
tantek being confusing as-is is not the same as more/less confusing than an alternative
#
tantek being confusing as-is is just an indicator that they should be improved
#
gRegor` Which is what I was trying to do? Heh
#
tantek adding a layer of abstraction almost always makes things *harder* to understand
#
gRegor` There's an algorithm and a sub-algorithm. I separated the sub-algorithm and used a more explicit term than "the page" to be clear that it's a separate page (the rel-author page) in the sub-algorithm. I don't see how that's an abstraction.
#
gRegor` I guess if it shouldn't be "A' "B", it should be something else that makes it different
#
gRegor` "psot page" and "rel-author page"?
#
tantek so that's what I'm iterating on right now - renaming them shorter and more descriptively
#
tantek it's not really a sub-algorithm since it involves data from the the algorithm
#
tantek it's just an algorithm with multiple levels
#
gRegor` Once you parse the rel-author page, the comparisons deal strictly with that page.
#
gRegor` If I understand it correctly
#
tantek except for 4.1.4 which deals with both
#
gRegor` That's back on the original page, not the rel-author page.
#
tantek ah I see the distinction you're making. ok
#
gRegor` Cool. :) Yeah, that's what helped me understand the whole algorithm better "what do I do with this page" vs "what do I do with the other page"
#
tantek and now that I'm rewriting that I'm reflattening more of the algorithm
#
tantek and trying to abstract the "author-page" as you suggested
#
tantek gregor I think I have a 1-depth less max shallowness iteration
#
tantek let's see if it is any more understandable.
#
tantek I added handling of URL author properties also which helps with author property usage like rel-author but contextualized inside an h-entry
#
tantek.com edited /authorship (+782) "flatten algorithm, group author property h-card discovery, group rel-author / author-page retrieval and processing, handle URL as author property, use author-page vs. h-entry's page" (view diff)
dns53 joined the channel
#
tantek gRegor`: take a look - pretty heavy re-write to flatten and better cover more cases in the proper order
#
tantek also named pages like "author-page" vs "h-entry's page"
#
gRegor` reading
#
gRegor` "if there is no author-page and the h-entry's page is a permalink page" Is there a microformat way to tell if the page is a permalink page?
#
gRegor` Looks good. The <code> makes the key parts stand out, too
snarfed and tantek joined the channel
#
tantek thanks gRegor`!
#
tantek re: way to tell if the page is a permalink page? I suppose that's out-of-band information, as usually by the time you've found the h-entry and are parsing it, you know whether you got it from a permalink, or from a stream / homepage, or perhaps from an archive page
#
tantek gRegor`: btw, what do you think of " if author property is an http(s) URL, let the author-page have that URL " ?
#
tantek inspired by barnabywalter's requests for me to fix my h-card on my home page, I updated the h-entry's on my home page to have <a class="u-author" href="http://tantek.com/"><,/a> - which should work to simply reference the primary h-card on the page as my authorship h-card, with name, photo etc.
#
tantek anyway, just braindumping that here in case it seems useful as background
#
tantek off for a bit
chrissaad joined the channel
#
gRegor` !tell tantek I thought if the p-author was just a string, that meant it was just the author name, not that it should be parsed further to tell if it's a URL. But you know microformats way better than me. :)
#
Loqi Ok, I'll tell him that when I see him next
#
gRegor` !tell tantek e.g. <span class="p-author">John Doe</span>
#
Loqi Ok, I'll tell him that when I see him next
#
gRegor` !tell tantek what was the author microformat like before you changed it to u-author?
#
Loqi Ok, I'll tell him that when I see him next
#
aaronpk hello indiewebcamp!
#
Loqi aaronpk: bnvk left you a message 1 day, 3 hours ago: what state is https://github.com/benwerd/indiereader in? The Readme is a bit thin at the moment
#
Loqi aaronpk: barnabywalters left you a message 11 hours, 17 minutes ago: your replies feed page doesn’t have in-reply-to markup on! you should add it: http://pin13.net/mf2/?url=http%3A%2F%2Faaronparecki.com%2Freplies
#
Loqi aaronpk: tantek left you a message 7 hours, 1 minute ago: if you're interested in giving a talk about IndieWeb sometime 2014-10-20..24, in particular how you've built or setup your own indieweb site, check this out and make a proposal! http://html5devconf.com/cfs-oct2014.html - feel free to mention that Tantek recommended you make a proposal (I was on their speaker board for a while) - and that offer is open to
#
aaronpk been a long day in real life
dns53 joined the channel
#
aaronpk !tell barnabywalters what do you mean add in-reply-to markup? I would have to add reply context first, right? (at least showing the URL i'm replying to?)
#
Loqi Ok, I'll tell him that when I see him next
#
aaronpk !tell bnvk "a little thin" yeah lol. the reader is a little rough around the edges still, and doesn't have any automatic way to update posts yet (like wordpress cron). also i'm not actually using it regularly which is a more important metric.
#
Loqi Ok, I'll tell him that when I see him next
glennjones joined the channel
#
Loqi glennjones: tantek left you a message 7 hours, 10 minutes ago: if you're interested in giving a talk about IndieWeb sometime 2014-10-20..24, in particular how you've built or setup your own indieweb site, check this out and make a proposal! http://html5devconf.com/cfs-oct2014.html - feel free to mention that Tantek recommended you make a proposal (I was on their speaker board for a while) - and that offer is open to
#
@andigalpern RT @t: going to Homebrew Website Club Wed 30th 18:30 @MozSF, 17:30 writing hr http://indiewebcamp.com/events/2014-07-30-homebrew-website-club & http://www.facebook.com/events/682246018521339 (ttk.me t4XD2) (twitter.com/_/status/492897095355875328)
#
Loqi [mention] https://brid-gy.appspot.com/repost/twitter/t/491832528580247552/492897095355875328 linked to http://indiewebcamp.com/events/2014-07-30-homebrew-website-club (webmention)
#
gregorlove.com edited /authorship (+6) "/* Algorithm Design Notes */ hCard => h-card" (view diff)
#
gregorlove.com edited /authorship (+1) "/* Name avatar display in comments */ hCard => h-card" (view diff)
#
gRegor` !tell tantek It would be good to capture the explanation of the u-author case here: http://indiewebcamp.com/authorship#Algorithm_Design_Notes
#
Loqi Ok, I'll tell him that when I see him next
#
gRegor` Should the microformat wiki link on /hCard be put on /h-card somewhere (under See Also / a History section), and set /hCard as a redirect?
#
aaronpk gRegor`: thar's a question for #microformats
#
gRegor` I mean the IWC wiki page /hCard
#
gRegor` It's just a link to the microformats wiki. I was thinking we could consolidate it into our /h-card page
#
aaronpk ohh
#
aaronpk yeah
#
aaronpk i didn't even know there was an hCard page
#
aaronparecki.com edited /hCard (-20) "Redirected page to [[h-card]]" (view diff)
#
gRegor` I just found it poking around, fixing some "hCard" text
#
aaronpk yup, it was created july 4 2012
#
aaronpk never touched since then
#
aaronpk that was well before the h-card page existed
KartikPrabhu and chrissaad joined the channel
#
KartikPrabhu wassup Loqi?
#
KartikPrabhu Jeena: do you not sen webmentions yet... or did I just not get this one: https://jeena.net/notes/177 ?
glennjones and tantek joined the channel
#
Loqi tantek: gRegor` left you a message 1 hour, 58 minutes ago: I thought if the p-author was just a string, that meant it was just the author name, not that it should be parsed further to tell if it's a URL. But you know microformats way better than me. :)
#
tantek Good evening
#
Loqi tantek: gRegor` left you a message 1 hour, 55 minutes ago: what was the author microformat like before you changed it to u-author?
#
Loqi tantek: gRegor` left you a message 1 hour, 57 minutes ago: e.g. <span class="p-author">John Doe</span>
#
Loqi tantek: gRegor` left you a message 1 hour, 55 minutes ago: what was the author microformat like before you changed it to u-author?
#
Loqi tantek: gRegor` left you a message 1 hour, 4 minutes ago: It would be good to capture the explanation of the u-author case here: http://indiewebcamp.com/authorship#Algorithm_Design_Notes
#
Loqi tantek: gRegor` left you a message 1 hour, 4 minutes ago: It would be good to capture the explanation of the u-author case here: http://indiewebcamp.com/authorship#Algorithm_Design_Notes
#
tantek Duplicates much Loqi?
#
Loqi woot!
#
tantek No woot
#
tantek Assuming Gregor reads the logs. "author" is just a property. Implies nothing about name. Any semantics more than a string (or sub object like an h-card) is up to the vocabulary or consuming application.
#
tantek The combination of barnabywalters asking for me to fix / update my home page h-card, and snarfed askin about use-cases where you have minimal if any h-card associated with a post or home page got me thinking how I would markup authorship of posts in my home page stream.
#
tantek So I thought about how can I express authorship if my posts without altering the presentation. And I came up with using an empty <a> tag with u-author and an href of / for my home page.
#
tantek It's an experiment and proposal all in one.
#
tantek And live on tantek.com so you can try it out
#
tantek Of course down level / previous impls that treat "author" as just a string and display it will work "fine" and just show my URL instead of my full name. Like we used to do with our nametags at SXSW. (See recent PDF talk on IndieWeb)
#
KartikPrabhu !tell tantek: I think the whole p-author vs u-author is a misconception people have about mf2 in general. Not realising that p-author an u-author only change the parsing and are actually the 'same' property. Maybe this could be improved upon in mf2 documentation.
#
Loqi Ok, I'll tell him that when I see him next
tantek joined the channel
#
Loqi tantek: KartikPrabhu left you a message 4 minutes ago: I think the whole p-author vs u-author is a misconception people have about mf2 in general. Not realising that p-author an u-author only change the parsing and are actually the 'same' property. Maybe this could be improved upon in mf2 documentation.
#
@t New rule: make tech proposals to improve web publishing on your site, or 1st learn how to. Prev http://tantek.com/2013/219/t7/new-rule-learn-html-before-innovate-osfw3c (ttk.me t4XG3) (twitter.com/_/status/492927267039752192)
#
tantek KartikPrabhu: yes they are the same property - "author" - hence I've started to refer to the unprefixed property name in such algorithms
#
tantek could you elaborate or point to the misconception you're seeing?
#
KartikPrabhu quite a few people have asked about whether they should use p-author or u-author to markup 'author info' (author only used as example). This seems to me to stem from not realising that u- or p- only talks about the parsing and not the property itself.
#
KartikPrabhu same with u-in-reply-to or p-in-reply-to
#
@absalomedia RT @t: New rule: make tech proposals to improve web publishing on your site, or 1st learn how to. Prev http://tantek.com/2013/219/t7/new-rule-learn-html-before-innovate-osfw3c (ttk.me t4XG3) (twitter.com/_/status/492928456158486528)
#
tantek I guess I haven't see anyone ask about should they use u-author - could you provide a URL for reference / context?
#
KartikPrabhu wasn't gregor` asking about it which you replied to?
#
KartikPrabhu I have had this question more about in-reply-to in IRC... will try to look for log reference
#
tantek KartikPrabhu: gregor was specifically asking about my changes to /authorship
#
KartikPrabhu tantek: couple of places such question have been asked/answered are: http://indiewebcamp.com/irc/2014-03-08/line/1394321980 http://indiewebcamp.com/irc/2014-03-12/line/1394676651 http://indiewebcamp.com/irc/2014-03-12/line/1394676679 http://indiewebcamp.com/irc/2014-04-08/line/1397018268
#
@kevinmarks RT @t: New rule: make tech proposals to improve web publishing on your site, or 1st learn how to. Prev http://tantek.com/2013/219/t7/new-rule-learn-html-before-innovate-osfw3c (ttk.me t4XG3) (twitter.com/_/status/492930204906438656)
#
KartikPrabhu again this seems to come up for in-reply-to most when people are trying to markup/consume replies an reply-contexts
#
@OnTheWebz RT @t: New rule: make tech proposals to improve web publishing on your site, or 1st learn how to. Prev http://tantek.com/2013/219/t7/new-rule-learn-html-before-innovate-osfw3c (ttk.me t4XG3) (twitter.com/_/status/492931278333677568)
#
tantek KartikPrabhu: would be good to see the specific examples
#
KartikPrabhu will keep an eye out in the future
#
KartikPrabhu is not very good at fining references
#
tantek easier to capture them when they occur to some place findable
gRegor` joined the channel
#
tantek welcome back gRegor`
#
KartikPrabhu tantek: yes... I should learn to do that
friedcell joined the channel
#
gRegor` Read the logs. The use of the author property makes sense.
#
tantek cool
#
gRegor` I'm in the process of rebuilding some of my mf2 parsing because it was one of the first indieweb things I did and I have a better understanding now, so I'm improving some of it. I'll check for a URL in the author property.
#
gRegor` My rel-author parsing used to just be: return the first h-card from the author-page, heh
#
gRegor` Some properties are consistently URLs so I have to remind myself that the p-, u-, etc. is just for parsing.
#
gRegor` Like in-reply-to that KartikPrabhu mentioned
#
gRegor` (is consistently a URL)
#
gRegor` so far :)
#
@hmans It's funny how there's pretty much no #indieweb participation in Germany, the country probably loudest about the surveillance shenanigans. (twitter.com/_/status/492965625074294784)
#
KartikPrabhu hmans: there was an indiewebcamp in Berlin this time
#
hmans Yeah?
#
hmans I've seen on /IRC-People that some of you are from .de, it's just that it doesn't feel there is any real traction. I keep hearing/reading "STOP USING SILOS! THE NSA IS WATCHING US ALL! Oh, what, I need to set up something myself? Screw that, you nerds"
#
hmans With all the protesting in this country you'd think they'd be a bigger part of movements like this
#
KartikPrabhu hmans: here is the list of people who signed up for IWC 2014 in Berlin - http://indiewebcamp.com/2014/Guest_List#Farther_East
#
hmans I'll make sure to be there next time.
danbri, ShaneHudson, Rev_Illo, glennjones and barnabywalters joined the channel
#
Loqi barnabywalters: aaronpk left you a message 6 hours, 33 minutes ago: what do you mean add in-reply-to markup? I would have to add reply context first, right? (at least showing the URL i'm replying to?)
#
barnabywalters !tell aaronpk RE adding in-reply-to markup, yes — it’s quite an important part of the post, after all, the only thing letting parsers know that the post is a reply post
#
Loqi Ok, I'll tell them that when I see them next
fourtonfish and barnabywalters joined the channel
#
@roundtrip RT @lucas_gonze: Webmention URIs: https://medium.com/@lucas_gonze/bec4e0dda3c6 (twitter.com/_/status/493004429650825216)
scor, Rev_Illo, Sebastien-L and dariusdunlap joined the channel
#
@Yesiten Newest on Wikipedia: IndieWeb Via: http://yesiten.com/go/ (twitter.com/_/status/493032771883761664)
bnvk joined the channel
#
Loqi bnvk: aaronpk left you a message 9 hours, 28 minutes ago: "a little thin" yeah lol. the reader is a little rough around the edges still, and doesn't have any automatic way to update posts yet (like wordpress cron). also i'm not actually using it regularly which is a more important metric.
#
bnvk !tell aaronpk: gotcha, thanks for the status re: the reader ;)
#
Loqi Ok, I'll tell them that when I see them next
#
PMurphs_ kay sorry i cant do this anymore i am way too hung over
#
PMurphs_ wbhat
#
PMurphs_ how dix this windoe open
#
PMurphs_ did*
chrissaad joined the channel
#
aaronpk good morning!
#
Loqi aaronpk: barnabywalters left you a message 3 hours, 18 minutes ago: RE adding in-reply-to markup, yes — it’s quite an important part of the post, after all, the only thing letting parsers know that the post is a reply post
#
Loqi aaronpk: bnvk left you a message 26 minutes ago: gotcha, thanks for the status re: the reader ;)
wolftune joined the channel
#
aaronpk so step 1 of this project is complete! I now have my bike rides and runs on my website! http://aaronparecki.com/metrics/2014/07/08/194743/
#
bnvk aaronpk: do you track your pushups with the Beeminder app?
#
aaronpk bnvk: no it's my own app https://github.com/aaronpk/PushupCounter-iOS
#
bnvk hehe
#
aaronpk it's actualy a micropub client
#
bnvk nice
#
bnvk does your nose touch the screen?
#
aaronpk yeah
#
PMurphs_ oh
#
PMurphs_ hi aaronpk
#
PMurphs_ so it turns out that anamanaguchi is great live
KevinMarks_, barnabywalters, chrissaad, jschweinsberg and tantek joined the channel
#
tantek good morning #indiewebcamp!
#
pdurbin tantek: it's always *just* past morning for me when you say that :) ... what's new?
#
tantek the /authorship algorithm got a pretty big rewrite (nearly the same functionality, hopefully more understandabl)
snarfed joined the channel
#
tantek snarfed, kylewm still any confusion about in-reply-to being a URL vs an object etc.? per http://indiewebcamp.com/irc/2014-04-08/line/1397018268
#
snarfed tantek: i don't think so. we get plenty of both, and our code handles both fine now
#
snarfed thanks!
#
snarfed https://github.com/snarfed/activitystreams-unofficial/blob/master/microformats2.py#L42
#
tantek snarfed, how about rel-in-reply-to vs. u-in-reply-to - any challenges there?
#
snarfed i try as hard as i can to never ever challenge anything markup related :P
#
tantek lol
#
kylewm huh, was going to ask gRegor` the same question re: http://indiewebcamp.com/irc/2014-07-26/line/1406359445
#
snarfed i maybe kind of hopefully understand the difference btw rel-* and u-*
#
tantek page scope vs. object scope
chrissaad joined the channel
#
snarfed yes
#
KevinMarks We tried making rel have object scope before but it was confusing
#
tantek and HTML5 settled on rel having page scope by default, so it was easier to go with that.
gRegor` and brianloveswords joined the channel
#
gRegor` Morning, indiewebcamp
#
gRegor` What was the question regarding in-reply-to, kylewm?
#
tantek gRegor`: KartikPrabhu showed some examples of people being apparently confused by rel vs u- vs. p-
#
gRegor` I should not comment on mf2 when it's late at night. Re-reading I realize my comment wasn't correct.
#
gRegor` As is usually the case when I talk mf2, haha
#
Loqi fo sho
#
gRegor` That hurts, Loqi
#
Loqi :)
#
tantek unintentionally delayed Loqi put down
#
tantek s/put down/putdown
#
Loqi tantek meant to say: unintentionally delayed Loqi putdown
#
aaronpk whoa new sidebar for the wiki, just saw that
#
tantek hey welcome back aaronpk ;)
#
tantek yeah that was a result of discussion with gRegor` about what all the things the wiki redesign effort was trying to do
#
tantek and seeing what we could quickly solve incrementally
#
aaronpk miss so much even being offline for a day now
#
tantek hey that was days ago ;)
#
aaronpk ooops
#
tantek incremental > put everything in one big release
#
tantek and re: missing things, somehow I missed how nice barnabywalters has made the time-ordered display of interaction on his posts: http://waterpigs.co.uk/notes/4UcCe5/
cuibonobo joined the channel
#
tantek that's some Tumblr-level coolness going on right there!
#
tantek barnabywalters++ for nice time-ordered display of interaction on his posts
#
Loqi barnabywalters has 51 karma
#
barnabywalters tantek: thanks! the design was inspired mainly by tumblr
#
aaronpk hmans: did you get your indieauth questions answered yesterday? I saw some chatter about wanting your site to be its own auth server
#
tantek.com created /reply-to (+25) "r" (view diff)
#
tantek.com created /reply_to (+25) "r" (view diff)
#
cuibonobo i just tried to login in to barnabywalters site and i got a `NoMethodError at /auth` from indieauth.com
#
tantek.com edited /in-reply-to (+17) "clarify in-reply-to for use on post permalink page's in particular" (view diff)
#
tantek well hopefully that makes the "reply to" questions easier to discover answers for - cc: KartikPrabhu
#
barnabywalters cuibonobo: are you by any chance using firefox nightly?
#
cuibonobo i'm on firefox 31
#
barnabywalters I just reproduced that problem but spotted that the me parameter was only http://wa rather than my full URL, which looks like a firefox autocomplete bug. I typed the whole thing and logged in fine
#
aaronpk cuibonobo: oh thanks
#
aaronpk can anybody spot the bug here? https://github.com/aaronpk/IndieAuth/blob/master/controllers/auth-web.rb#L213
#
cuibonobo i still have the stacktrace open. query_string has `me=https://jenmontes.com&redirect_url=http://waterpigs.co.uk`
#
barnabywalters doesn’t speak ruby
#
aaronpk shouldn't need to know ruby to spot the bug ;)
#
tantek aaronpk - loss of knowledge as to whether http or https?
#
aaronpk should be an "else if", not just "else"
#
barnabywalters aaronpk: how come that isn’t a syntax error then?
#
tantek aaronpk: that's not a syntax error?
#
tantek heh
#
tantek also doesn't speak ruby
#
aaronpk it's not a syntax error because ruby
#
barnabywalters or is ruby just really permissable about that stuff
yakker joined the channel
#
tantek is there a rubylint to catch such errors?
#
aaronpk ruby syntax doesn't have curly brackets or anything, so the stuff between the "else" and "end" is just part of the block
squeakytoy joined the channel
#
gRegor` aaronpk: That goto fail shouldn't be there? ;)
#
aaronpk also barnabywalters that should be redirect_uri in your query string, not redirect_url, which is why it hit that line in the first place
#
tantek !tell acegiak could you take a look at https://indiewebcamp.com/in-reply-to and let me know if it helps clarify rel-in-reply-to vs. u-in-reply-to vs. p-in-reply-to usage and consumption?
#
Loqi Ok, I'll tell them that when I see them next
#
barnabywalters aaronpk: thanks for the heads-up, I shall fix it
#
tantek !tell skinny,crystal_ go ahead and add yourself to indiewebcamp.com/irc-people so we know which nickname to !tell you in the future :)
#
Loqi Ok, I'll tell them that when I see them next
#
tantek !tell skinny,crystal_ could you take a look at https://indiewebcamp.com/in-reply-to and let me know if it helps clarify rel-in-reply-to vs. u-in-reply-to vs. p-in-reply-to usage and consumption?
#
Loqi Ok, I'll tell them that when I see them next
#
tantek we don't seem to have an area on the wiki where we show indieweb examples and screenshots of displaying interactions (comments, reposts, faves/likes) on a post
#
aaronpk wow really?
#
tantek well if we do I can't find it
#
tantek not on /comment-presentation nor /interactions
#
aaronpk nor /reply
#
aaronpk nor /responses
#
tantek yeah I still need to split up /reply vs /comments
#
gRegor` I have one small example here. Feel free to link to or embed on other pages: http://indiewebcamp.com/User:Gregorlove.com/webmention#Displaying_webmentions
#
tantek aaronpk I'm not sure how useful the responses article is for describing, it seems more historical discussion at this point (which is good to capture, but just not as explanation)
#
tantek also "responses" just sounds like another word for "replies"
#
tantek hence I've been using the term "interactions" (with a post) instead
#
tantek thinking that implies a broader scope than just replies, broad enough to include reposts, likes, bookmarkings etc.
#
tantek thoughts? (namingishard)
#
tantek (also "responses" smells of programmer-centric naming, e.g. HTTP responses)
#
tantek s/programmer/plumbing
#
Loqi tantek meant to say: (also "responses" smells of plumbing-centric naming, e.g. HTTP responses)
#
aaronpk yeah I didn't know about the responses article
#
bear well, respond is normal english when you follow up with someone
#
bear I just responded to your query about this
#
tantek bear, right, another word for a reply to someone
#
tantek hence less suitable for including things like favorites, reposts
#
bear ah - I get your point - yes
#
bear reply/respond favourite repost
#
bear hmm, interaction seems too clinical
#
bear collaborate?
#
bear feels more inclusive than interact
#
tantek collaborate may imply more closeness than there is, and other unintended meanings too
#
bear raises the "naming is hard" flag
#
cuibonobo barnabywalters: i was checking out the post that tantek linked to (http://waterpigs.co.uk/notes/4UcCe5/) and noticed that interactions by snarfed don't have a context — they just list his name
#
cuibonobo i suppose they are favorites because they have a star next to them, but i don't think he favorited twice… so something is being lost
#
gRegor` "collaborate" to cover comment/like/repost/reply? That sounds weird.
#
bear agrees and retracts collaborate
#
gRegor` I like "responses" overall.
#
bear I was reading the synonym list and thought it looked good
#
bear goes to find food
#
gRegor` cuibonobo: I wonder if snarfed re-sent the webmention and it wasn't parsed properly as an update
snarfed joined the channel
#
cuibonobo gRegor`: i'm not sure. that might be the case. both interactions point to the same URL. yet, when you visit the URL, it's blank
#
gRegor` That's snarfed's implementation of likes. Doesn't display anything visible, but the mf2 is there.
#
gRegor` Or it was. I think I've seen visible likes on his site since then.
#
snarfed https://snarfed.org/responses#likes
#
gRegor` Oh hai, snarfed
#
snarfed ducks
#
cuibonobo snarfed: another weird thing I noticed: https://snarfed.org/2014-02-16_re-barnaby-walters-%E2%80%A2-idea-a-microformat-for-download-signatur
#
cuibonobo (though in my browser it just appears as a bullet point)
#
cuibonobo instead of %E2%80%A2
#
snarfed cuibonobo: yup, url encoding
#
snarfed my indie replies/likes/reposts' markup gradually changed over time
#
snarfed if you view source, you'll see there's a valid u-like[-of] in there, just no visible text
#
snarfed (as gRegor` mentioned)
#
snarfed current likes are better
#
cuibonobo i was just wondering whether having a unicode character in your URL was making barnaby's parser confused
#
snarfed cuibonobo: i doubt it. it's probably because there was no text in the e-content
#
Jeena KartikPrabhu I should send and get webmentions, did you send one to me too?
#
cuibonobo snarfed: ah. gotcha
#
tantek is going out for a coffee and will read the naming debate as it continues. Currently leaning towards "interactions" for the reasons above. "responses" being too synonymy with "replies".
dariusdunlap and barnabywalters joined the channel
#
gRegor` Why is the synonymy of responses and replies problematic?
#
gRegor` /replies redirects to /comment, so I think we could call them comments instead of replies.
#
gRegor` comments/likes/reposts under the umbrella: responses?
#
tommorris has a read of snarfed's site
#
tommorris likes the design and IA
#
snarfed tommorris: i wouldn't recommend it
#
snarfed gRegor`: it's problematic if it's an umbrella term, but people also think it means replies specifically
#
snarfed tommorris: thanks!
#
gRegor` Ah, ok
#
gRegor` I dunno then. Back to code. :)
#
tommorris snarfed: and the wedding photographs on the Pottery Barn site are good too. I say that as someone who likes both photography and weddings (having both officiated one and written lots of grumpy letters to my MP demanding the right to be able to get married). ;)
#
snarfed tommorris: thanks! i appreciate it. our photographer was awesome
#
snarfed hmans: looks like you're trying to get json versions of bridgy's mf2? try ?format=json . details at https://www.brid.gy/about#source-urls
#
snarfed hmans: you might also be interested in http://facebook-activitystreams.appspot.com/ , http://twitter-activitystreams.appspot.com/ , and http://instagram-activitystreams.appspot.com/
#
cweiske.de edited /distributed-indieauth (+130) "/* Level 1: Enable web sign-in */" (view diff)
#
aaronpk not quite, cweiske
#
aaronparecki.com edited /indieauth.com (+9) "move TOC to bottom" (view diff)
#
kylewm barnabywalters: for likes/retweets from bridgy are you just using the time received as the time of the like?
#
barnabywalters kylewm: OTTOMH I can’t remember, but it is likely. I don’t handle bridgy webmentions any differently to other webmentions
#
snarfed fwiw, not all bridgy likes will have created twitter (and maybei think
#
snarfed er, sorry, misfire
#
snarfed the silos themselves don't always report like times (eg twitter, instagram), so bridgy omits them in those cases
#
@GlennF RT @t: New rule: make tech proposals to improve web publishing on your site, or 1st learn how to. Prev http://tantek.com/2013/219/t7/new-rule-learn-html-before-innovate-osfw3c (ttk.me t4XG3) (twitter.com/_/status/493100806954971136)
#
kylewm barnabywalters: that's really cool, i think i prefer that to the facepile i use... also are some reactions/interactions out of chronological order on purpose?
#
barnabywalters kylewm: er, probably not :)
#
barnabywalters it is possible that I’m displaying a different datetime to the one I sort by, or that timezones are at work
#
cweiske.de edited /distributed-indieauth (+484) "move level 1 to 0, add new level 1." (view diff)
#
cweiske.de edited /distributed-indieauth (+4) "/* Level 1: Choose an authorization server */" (view diff)
#
kylewm barnabywalters: just fyi i'm looking at the last two on http://waterpigs.co.uk/notes/4UcCe5/
#
barnabywalters hmm, that last one is a duplicate of a mention seen above, so I suspect that has something to do with it. I’ll take a closer look tomorrow
#
bret watching https://ind.ie/summit/videos/#RichardStallmanInterview while doing chores. its typically a bad idea to try to school RMS in person, as you are just going to be arguing :p
#
@notenoughneon Got an #indieweb feed reader running: http://notenoughneon.com/feed.php Its still pretty buggy. (http://notenoughneon.com/2014/7/26/1) (twitter.com/_/status/493106818579562496)
#
bret I think RMS's primary mistake is so much of his energy and sacrafice goes to eliminatig non-free software right now rather than building with the best tools available to actually reach parity
#
bret its a losing battle trying to talk people into using trisquel
#
cuibonobo bret: agreed
#
bret i *think* it may stem from his idea that all software can be maintained?
#
bret or most
#
bret you can code yourself into a corner
cuibonobo joined the channel
#
gRegor` Heh, I love all RMS requests for photos / videos of him.
#
bret he is a really amazing guy and has a great argument for free software. I just dont think he should be driving that ship any longer because he is out of touch, and thus not very effective
#
bret he is unable to address the consumer device space other than the microwave example or "dont use it buy an old one its not that big of a deal"
#
aaronparecki.com edited /distributed-indieauth (+43) "/* Level 1: Enable web sign-in */ clarify level 1" (view diff)
#
aaronpk oops that was supposed to say level 0
#
aaronpk thinking i'm gonna call this OwnYourRun
#
aaronpk runkeeper/strava -> micropub
glennjones and bitraten joined the channel
#
bret :D
snarfed joined the channel
#
gRegor` Woo, pretty sure my updated library handles the entire /authorship algorithm
#
gRegor` testing testing
Kopfstein joined the channel
#
@ShaneHudson RT @notenoughneon: Got an #indieweb feed reader running: http://notenoughneon.com/feed.php Its still pretty buggy. (http://notenoughneon.com/2014/7/26/1) (twitter.com/_/status/493120699650048000)
#
GWG I need inspiration
yakker and chrissaad joined the channel
#
gRegor` Inspired by kittens, GWG? https://www.youtube.com/watch?v=FtX8nswnUKU
#
GWG gRegor`: I feel like there is still more to do before I'm done with what I'm working on
#
GWG I need an opinion again
#
gRegor` On?
Leeky joined the channel
#
GWG Well, I'm working on the final revision of my design for my site
#
GWG Compare http://tiny.n9n.us/
#
GWG to http://david.shanske.com/
#
fourtonfish GWG: A silly question, but which one is the new design?
#
GWG The tiny.n9n.us, titled Theming Test Site
#
GWG fourtonfish: Basically, I rewrote the entire thing.
#
gRegor` Just general opinions, or something specific?
#
GWG Well, I'm iterating.
#
GWG What is missing with the new design before I deploy it?
#
gRegor` At the second-to-smallest media query breakpoint, the square image in the sidebar is too wide for the container.
#
fourtonfish GWG: Ah, I see. I don't know, I like the david.shanske.com version better, actually. I'm not a fan of round borders, but that's just a personal preference, of course. I really like the navigation bar under the banner. On the test site it's just too...disjointed?
snarfed joined the channel
#
fourtonfish Also -- I would drop the clock.
#
fourtonfish I mean the clock widget.
#
gRegor` I think div.response should have the margin-top removed, so the rounded corners align nicer with the parent container.
#
GWG fourtonfish: I needed placeholders
#
GWG gRegor`: Div.response is not part of the design. It is part of the response plugin, actually. Independent CSS styling.
#
GWG I needed stuff in the sidebar
#
GWG On the main site, I put the clock in temporarily, but never found what I wanted to do with the space.
#
GWG fourtonfish: Disjointed how?
pfefferle joined the channel
#
fourtonfish Let me take a screenshot -- maybe it's just me? I'm on Chrome/Ubuntu (both latest).
#
GWG I've done both Chrome and Firefox on Fedora
#
fourtonfish http://i59.tinypic.com/2zf7ip0.png
#
GWG fourtonfish: I see it. But I don't know what you mean by it
#
fourtonfish I just like how on david.shanske the navigation bar and the banner go together. There is no space between -- that's all I meant.
#
GWG Oh
#
GWG That
#
fourtonfish The fact that the navigation bar is wider than the banner just makes it look like a mistake, IMHO.
#
fourtonfish If you want an honest opinion, I would go with http://david.shanske.com/ -- just make the corners sharp. Remove the clock widget, and add missing icons in the navigation bar so it's consistent.
ShaneHudson joined the channel
#
fourtonfish (I'm not a designer, so yeah, just a personal opinion.)
#
fourtonfish Oh yeah, one more thing:
#
GWG Neither am I
#
fourtonfish When you hover a title of the post, for example Unified and Contextual Communications -- the icon shouldn't be underlined.
#
fourtonfish The way it breaks just looks a bit messy.
#
fourtonfish Actually now I see that it goes for every link with an icon.
#
fourtonfish I hope it's clear what I mean -- when you hover Unified and Contextual Communications or Comment -- the icon's underline is off compared to the text.
#
GWG fourtonfish: I'm trying to rebuild the design I built for the live site using the new platform I used. There's a reason I need that, as it is more flexible.
#
GWG But I'm making some minor tweaks.
#
GWG So, ultimately I want the site to look like a refined version of the old one
#
Loqi gives GWG the site to look like a refined version of the old one
#
fourtonfish Well, then I would just make the navigation bar same width as the banner.
#
GWG Well, the banner is a scaling issue I'm working on.
#
fourtonfish Even if it stays rounded/spaced out.
#
GWG For example, if you refresh, you get the foundation. I just turned off the other mod
#
fourtonfish Yeah, I see.
#
gRegor` Submitted a PR to sandeep's authorship-test-cases that covers tantek's new use of u-author. https://github.com/sandeepshetty/authorship-test-cases/pull/1
#
gRegor` Also, authorship-test-cases is a thing. :) Spent so much time working through the authorship algorithm, I hadn't explored the rest of that page much, heh.
#
fourtonfish GWG: Well I think once you fix that, it will look much better. There is also the problem with the icons -- compare when you hover JUNE 13, 2014 1:00 PM and DAVID SHANSKE -- the icon next to the date doesn't get underlined.
#
fourtonfish (Actually, I'm just looking at the theme disabled, so I'm not sure if it's the same with it on.)
#
GWG Well, I need to determine if I'm fixing anything at the lower or higher level.
cuibonobo__ joined the channel
#
fourtonfish GWG: Sorry, can you elaborate on what you mean by lower/higher level?
#
GWG When you saw me just change the styling...one is layered on top of the other
#
fourtonfish Right. Well if you just mean whether to edit the CSS for the first layer or the second layer, I guess you are using Bootstrap for the first layer, so I would do all my edits to the second one.
#
fourtonfish So that way you can update Bootstrap.
#
GWG Well, the header is styled on the first layer
#
fourtonfish I mean -- when a new version comes out.
#
GWG There is a small custom CSS file.
#
GWG Then the modifications to the style go up higher.
#
fourtonfish Well, like I said -- if you ever want to update Bootstrap, it's better not to touch the CSS file and just edit your custom CSS. Yeah, this is not the idea solution as you might end up with some unnecessary code (and other small issues), but unless you have tons of traffic, it's not even worth mentioning. I do this for smaller projects -- I actually load Bootstrap from a CDN and then load my custom CSS (but I don't usually do
#
fourtonfish a lot of edits, otherwise I just roll my own CSS).
#
fourtonfish *ideal solution
friedcell, scor, dariusdunlap, chrissaad, danbri, tantek and tantek_ joined the channel
#
tantek_ good afternoon #indiewebcamp
#
GWG Good afternoon, tantek
trodrigues joined the channel
#
tantek any ActivityStreams folks around with any input on the what to name the combination of comments/reposts/likes ?
#
tantek speaking of such combinations, really liking the presentation of all three here: http://notenoughneon.com/2014/7/26/1
#
tantek emmak++ for some nice comments/retweets/favorites presentation on a post permalink
#
Loqi emmak has 4 karma
#
KartikPrabhu I use 'response' to refer to all such activites
#
tantek KartikPrabhu: it doesn't bother you that 'response' just sounds like a synonym for 'comment' or 'reply' ?
#
KartikPrabhu 'response' to me denotes any type of reaction to a post, could be a reply, comment, mention, like or repost
#
KartikPrabhu maybe 'reaction' could be better
#
tantek yes I think I've heard reaction as well
#
tantek interesting pattern. interaction, reaction
#
KartikPrabhu feedback could be another
gRegor`, snarfed and KartikPrabhu joined the channel