2013-06-26 UTC
spinnerin joined the channel
# 00:00 tantek acegiak - it does - you need to read them in prose in place in the phrasings that barnaby proposed on the page
# 00:01 acegiak isn't this a perfect use case for the rev attribute?
# 00:01 tantek acegiak - the rev attribute is unusable (bad at usability)
# 00:02 tantek it make the confusions you just experienced about the rel attribute 10x worse
# 00:02 tantek benwerd - a-ha yes! so reply-context is *the* key use-case for downstream webmentions back to replies.
# 00:02 tantek then the recipient can decide whether to update their reply (and reply-context) or not
# 00:03 tantek I wouldn't do it automatically myself - since my original reply was worded according to what was originally said. and if its meaning changed...
# 00:04 acegiak well the actual relationship between a like and the document that is the target of the href is "target" or "object" and the other information should be stated elsewhere
# 00:04 acegiak the problem comes formt he fact that the subject of the sentence is an event
# 00:05 tantek if all you want to do is treat every target as just "target" or "object", then you have no need for the rel attribute
# 00:06 tantek but those of us who want to indicate the destination as a noun relative to the source find the rel attribute quite useful for values accordingly. nearly all the existing-rel-values follow that pattern/rule
# 00:09 tantek so it's more like "object-of-noun-that-represents-a-post-that-is-the-result-of-a-verb"
# 00:10 acegiak my objection is to the inelegancy of the wording really
# 00:10 tantek ah good, alternatives are better feedback than objections :)
# 00:11 acegiak tantek: yeah, sorry objection wasn't the right word there.
caseorganic joined the channel
# 00:12 tantek though likee and favoritee solve the OPPOSITE problem
# 00:13 tantek the authors of each post are simply the h-cards in the p-authors of the h-entrys
# 00:13 acegiak yeah, I'm just saying "object-of-like" is a poor solution from a human-readable point of view
# 00:14 tantek again, an absolute criticism is not particularly helpful, but if there's an alternative to consider, that can be useful
# 00:15 acegiak tantek: I was providing that as reasoning for my suggestion of likee and favouritee or rather liker and favouriter
# 00:15 acegiak but I see now that those would seem to point to the actor rather than the action
andreypopp joined the channel
benwerd joined the channel
# 00:21 acegiak if we can tread "like" as a verb we could potentially treat "liked-by" as a verb
# 00:28 bret This concept of semantic html + web-mentions + automatic parsing and displaying is super powerfull
# 00:33 tantek bret - it's amazing what you can get done when you leave behind programmer-only DRY violation abstractions.
# 00:36 benwerd I mean, it always was the web. "The next Facebook" is a very small concept in comparison.
# 00:36 bret the next facebook should be something that prevents another facebook outbreak
# 00:36 benwerd But nonetheless, the web - the real one, with interconnectedness and ownership and decentralization - will win.
# 00:37 benwerd On that note, I'm going to shut down and head to Backspace
# 00:37 tantek.com edited /responses (+1272) "response collections and symmetric rels - just some capturing some thoughts along with criticism that there is no use-case for them y et" (
view diff )
scor joined the channel
# 00:51 tantek and with that - we have the first Federated Event post + RSVP
# 00:52 bret im hoping to polish things up tonight at backspace with ben and aaronpk and anyone else whos going to be there
# 00:52 tantek it's so easy once you have replies with webmentions working
# 00:52 tantek and assuming it's easy to add new post types to your CMS
# 00:53 bret tantek: add a link with the proper in reply to class and manually send one,
# 00:53 tantek bret - yeah - I want to make it turnkey (on the server to server side) before claiming I have it working.
# 00:54 bret but I can generate the data automatically
# 00:54 aaronpk bret: seems like it wouldn't be hard to add that to a simple shell script that does git push whatever then send webmention
# 00:55 bret right now i have a prototype to generate the command based off the post and I just ctl c ctl v it into my terminal
# 00:57 tantek aaronpk the "I'm RSVPing yes" is just fallback text
# 00:57 tantek the semantic of the post is an RSVP of "Yes" - encoded with <data class="p-rsvp" data="yes">
# 00:58 tantek I would have worded it "… is attending" myself.
# 00:58 aaronpk my instinct when creating this RSVP post on my site is to write a reply with the text "Sounds like fun, I'm in!"
# 00:58 tantek so when the event receives the RSVP webmention - it gets the semantic and displays it
# 00:58 tantek aaronpk - I think that makes sense to do RSVP + comment
# 00:59 aaronpk but it should also be possible to post just an RSVP with no comment
andreypopp joined the channel
# 00:59 bret I didn't include any note conents, just that rsvp tag
# 00:59 aaronpk for the rsvp+comment case, should I wrap the comment text in the <data> attribute?
# 00:59 bret i wonder what would happen if I included text
# 01:00 aaronpk and the contents would be empty if I didn't include a comment?
# 01:00 bret aaronpk: do what makes sense, your blazing a new path right now
# 01:03 tantek Bret you did include note contents because the plain text of I'm RSVPing "yes" is inside your <section class="e-content p-summary p-name">
# 01:05 bret Loqi: who is still in portland from IWC?
# 01:06 aaronpk unfortunatly my tweets don't show up in the indiewebcamp twitter search
# 01:08 bret It might be interesting to have a similar grid, but for specific websites
benwerd and scor joined the channel
# 01:18 bret aaronpk: it feels like indiauth is much slower since the weekend. is this true for you as well?
# 01:20 aaronpk trying to think if my changes would have made it slower
# 01:20 aaronpk I have a couple things I am going to change to speed it up a ton, but I dont think my changes would have slowed it down
# 01:21 tantek and I continue in my role of document you folks's awesome achievements first, slowly implement things to catch up second. ;)
# 01:28 bret oops, benwerd send you another webmention to see if I could add text to my rsvp, but it looks like I lack some of the required fields for an update time
# 01:29 bret i changed the conents of my original rsvp
# 01:29 aaronpk bret: I think you're publishing the full comment text as the rsvp value instead of just "yes"
# 01:29 aaronpk (I accidentally refreshed the wrong tab and saw your new comment)
# 01:31 bret i love git. I do all my development on any computer I have access to that has git
andreypopp joined the channel
EHLOVader joined the channel
andreypopp joined the channel
# 01:34 bret scratch that i will be a bit actually
# 01:54 acegiak what post should I reply to to test that webmentions are working?
# 01:54 tantek acegiak - you can try a webmention to any indiewebcamp.com URL and if it works you'll see a [mention] from Loqi here in the channel
scor joined the channel
# 02:02 tantek federated event+rsvsps was purely theoretical just 48 hours ago!!!
# 02:02 Loqi tantek meant to say: federated event+rsvps was purely theoretical just 48 hours ago!!!
caseorganic joined the channel
andreypopp_ joined the channel
# 02:05 Loqi today's #NetworkWorkingGroup #IETF #IESG #RFC cadre
# 02:07 tantek benwerd - did you implement the update protocol too?
# 02:07 tantek aaronpk - could you ask benwerd IRL if he got updates working?
# 02:10 aaronpk omfg I have got to replace this webignition crap in the php-mf2 parser
# 02:10 aaronpk "Fatal error: Call to undefined function webignition\NormalisedUrl\idn_to_ascii()"
benwerd_ joined the channel
# 02:11 tantek yeah for some reason I thought you implemented updates the same time as deletes, then I went back and checked the logs and saw that you'd only gotten deletes working so far
# 02:11 aaronpk i should really mark up these IRC logs as proper h-entries with h-cards so we can send web replies to individual lines in irc
# 02:11 tantek will be interesting to see if webmentioning an existing comment/reply gets an update
# 02:15 tantek especially one that say aaronpk commented on already - so we can go to his comment and see your old comment in the reply-context
# 02:16 benwerd_ er, my test was on localhost, but gimme a minute and I'll arrange that
# 02:18 benwerd_ I think I probably need someone else to update their comment to me and re-ping me
# 02:19 benwerd_ I can ping myself, but that doesn't seem like a fair test
# 02:21 acegiak so I sent a webmention to the microformats page on indiewebcamp.org, but looks like it got lost along the way
bizybee joined the channel
# 02:24 acegiak I'm using pfefferle's plugin so give me a minute to see if I can find logs
# 02:24 aaronpk you can try sending it manually to test the markup
andreypopp joined the channel
# 02:32 aaronpk hmm I think I can get likes working now after making the infrastructure change for RSVPs
# 02:32 tantek aaronpk - since that's such a useful webmention testing endpoint, how hard would it be to extend it to check for in-reply-to linkage and say "[comment]" instead of "[mention]"?
# 02:33 aaronpk i'm not parsing any microformats-2 markup there yet
# 02:33 aaronpk i suppose I could just do it manually since it's just one class
bizybee joined the channel
# 02:37 aaronpk OH: "this is so much better! because there are no investors!" (in reference to all the indieweb implementations popping up right now)
# 02:38 tantek aaronpk - or more appropriately - just one rel value - which is pagewide anyway
# 02:39 aaronpk oh right... but I'd have to look for both rel=in-reply-to and u-in-reply-to right?
# 02:39 tantek but the rel should only show up on permalink pages
# 02:40 tantek however the permalink pages are all you should get webmentions for
# 02:53 bret Im in that state where there is so much to do it feels useless do do anything
# 02:53 bret i think ill just work on presentation
# 02:54 tantek bret - look at your site and posts, and try to feel what aspect bothers you the most
# 02:54 aaronpk bret: that just means you need to find smaller tasks, then write them all down on the wiki page, then yo ucan work on one at a time
# 02:54 tantek aaronpk knows something about that challenge ;)
# 02:54 tantek (and how liberating it is once you've done so)
# 02:55 tantek then you create the page for it. describe it. then the "working on" section. then the itches "section.
# 02:55 bret i guess octopress takes credit for a jekyll implementation
# 02:56 bret the problem is that it HAS to be named becomes.github.io for the hosting
# 02:56 bret yeah I guess I could fork it and remove my stuff
# 02:56 tantek aaronpk - do your RSVPs automatically send the webmentions I presume?
# 02:57 tantek bret - doesn't matter - you can still call it something even if the only implementation is the one hardcoded for your site
# 02:57 bret that was the original plan but trying to design for a generic person got frustrating
# 02:57 tantek don't worry about designing / doing directory structures for a generic person
# 02:57 tantek just name it, create the wiki page and get cranking.
# 02:57 tantek then you can add abstracting / directory foo as a "get to it eventually" task on said wiki page
# 03:00 tantek cool - then just make the page on indiewebcamp.com a stub that links to that.
# 03:06 benwerd_ So now I'm wondering what it would take to write a webmention-enabled RPG
# 03:07 bret cross dimensional webmention enabled rpg
# 03:07 benwerd_ everyone gets to host a room and define the physics / rules
# 03:09 aaronpk "two players could play a chess game just by posting moves on their own indieweb sites as replies to each others' moves."
# 03:19 aaronpk bret: do you mean something like this: :<a class="p-name u-url" href="http://aaronparecki.com/">Aaron Parecki</a>
# 03:19 bret yeah and have that count for both the URL and Name
# 03:20 bret <a class="p-name fn u-url url" href="http://bret.io">Bret Comnes</a> only parsed as my name the other day
# 03:21 bret I had to add a second item: <a class="u-url" href="http://bret.io">bret.io<!-- This is here to force the hand of your MF2 parser --></a> for the URL to parse
# 03:29 bret so are we ripping the power of the @ symbol away from twitter and treating it as an arbitrary URL to someones identity we are responding to?
# 03:30 benwerd_ Trying to decide what to work on next. All my updates today have been because Tantek pointed out I was doing something suboptimally ;)
# 03:30 bret IE @person no longer referes to someone's twitter profile but rather to whichever profile they want to identify as?
# 03:31 bret sure, but the URL is the defining process? My name in RL has a lot of namespace issues
# 03:32 benwerd_ I actually did already accidentally reply to some dude called @bret on Twitter
# 03:55 aaronpk *finally* updated p3k web interface to fit on a mobile phone
# 04:00 bret aaronpk, benwerd_ do you put micro data into your list of posts?
# 04:01 aaronpk i think mine is just a list of h-entry's right now
# 04:02 aaronpk I probably won't change it unless I find a good reasoan to make it a real h-atom feed
# 04:10 aaronpk benwerd_: just noticed you're using indieauth from benwerd.com but posting from werd.io!
# 04:11 benwerd_ yeah, because my rel="me" links are all on a sub-page ... I should write a little header for the werd.io homepage
tantek, caseorganic and bizybee joined the channel
# 05:07 tantek Interesting. I'm seeing a dulicate set of posts on werd.io on mobile safari.
# 05:07 bret aaronpk: caseorganic nice meeting with you tonight! sorry for the abrupt departure
caseorganic, andreypopp and spinnerin joined the channel
# 05:13 grawity fdevillamil: Yes.
# 05:24 bret still not done but better than it was
benwerd joined the channel
# 05:34 bret sorry about leaving so abruptly, had to make it home for something benwerd
# 05:34 benwerd bret, not at all! We actually all left pretty soon afterwards
# 05:34 benwerd (nb: turned out that one of the musicians was jolie holland, formerly of the be good tanyas, which was pretty neat)
# 05:35 bret these hack sessions are kind of hard for me to focus on anything going on around me
# 05:36 benwerd honestly, I was finding the yodeling a little distracting though
# 05:37 bret would your site parse that note text too?
# 05:39 benwerd yes - I'm saving it, but not actually displaying it right now
# 05:40 bret I kind of feeling like re-doing my whole site
# 05:40 bret i made all kinds of bad decisions when I had no idea about any of this stuff
# 05:53 bret wow I just solved a long standing DRY problem with jekyll
# 06:25 bret Can I reference an h-card as an author of an h-entry if that h-card is outside of that dom element?
friedcell joined the channel
benwerd and pfefferle joined the channel
erikmaarten, pfefferle and seyz joined the channel
sandeepshetty and pfefferle joined the channel
# 07:53 pfefferle I just read about that crud and have some concerns about the "delete"
# 07:55 pfefferle I just implemented some kind of threaded webmentions on my site… that means, that a webmention could cause a big discussion in my comment stream… if the original article would be deleted, then a) a whole discussion would be whipped out or b) the discussion wouldn't be context aware any more if i would only delete the root element
# 07:57 pfefferle the same with updates… if it is possible to update a comment on any time, it could kill a whole context
# 08:01 erikmaarten pfefferle: I think I'm leaning towards skipping the whole updating/editing/deleting of comments. As a side effect, that might also incentivize thinking before typing instead of after :)
# 08:02 erikmaarten or perhaps adding new text to old comments could be allowed, but not editing the old content
# 08:03 pfefferle erikmaarten: i think it makes sense for a short period of time… perhaps 5 min or so… but then the comment should be locked
# 08:03 benwerd The joy of the IndieWeb is that everyone can implement it in the way that makes sense for them and their site!
# 08:04 erikmaarten I think Stack Exchange deals with comments like that (although comments are not that important over there, perhaps)
# 08:04 erikmaarten benwerd: agreed that plurality is great, though in the end it may also cause a lot of confusion
gjones joined the channel
# 08:05 benwerd I think in the beginning it's worth it, for competing ideas and models for how this should work
# 08:05 benwerd best practices and de facto standards will emerge
# 08:05 pfefferle perhaps I add it to the wiki as possible issue, what do you think?
# 08:07 erikmaarten that's a good idea, then we can develop the ideas and issues there as implementations are tested
eschnou joined the channel
friedcell1 joined the channel
# 08:25 acegiak pfefferle: can you try webmentioning one of the test posts on my blog?
# 08:26 acegiak is the difference between the webmention display and the regular pingback display the use of [...] wrapped around hte pingback mention?
friedcell joined the channel
# 08:33 acegiak pfefferle: the webmention goes on the do_pings call same as pingbacks, yeah? cause that takes ages on my system for some reason
# 08:40 pfefferle i have an idea… can you update your plugin with the latest version?
# 08:40 pfefferle and search for add_action('publish_webmention_reply', 'webmention_insert_post');
# 08:41 pfefferle and add add_action('publish_post', 'webmention_insert_post')
# 08:44 pfefferle and/or try to reply on my comment using the little reply link
# 08:56 acegiak pfefferle: back (half watching the tv because our primeminister is being challenged to a leadership ballot)
# 08:58 bret updated my layout a bit. I also managed a huge amount of backend refactoring that should help speed up adding new post types
# 09:24 acegiak pfefferle: I've updated to the latest version of the plugin
# 09:56 acegiak also: we have our old prime-minister back here in australia
# 10:05 acegiak I kwwp getting errors in my apache log which are redirection loop errors on thread.php
# 10:06 acegiak I have a fair few on there for managing the network, let me try doing some spring cleaning
adactio joined the channel
# 11:21 fdevillamil I'm pushing my co maintainers to add a separate stream on our blog engine for short status since we alreadt provide a URL shortener.
seyz joined the channel
scor, pfefferle_ and pfefferle joined the channel
# 12:03 acegiak pfefferle, I seem to be having trouble with ipfefferle, I'mma hit the sack
bnvk, morrocco_mole, bnvk_, brbcoding, seyz, scor and friedcell joined the channel
friedcell1, andreypopp, bnvk, b0bg0d, tilgovi, willnorris, tantek and ozten joined the channel
erikmaarten, andreypopp and pfefferle joined the channel
pfefferle, morrocco_mole, bizybee, spinnerin, bnvk, sandeepshetty, caseorganic, andreypopp, caseorga_, benwerd, melvster, tantek, eschnou and KevinMarks joined the channel
bnvk joined the channel
eschnou joined the channel
willnorris, b0bg0d, caseorganic, tantek, bnvk, andreypopp, morrocco_mole, scor and spinnerin joined the channel
# 23:01 tantek greetings - I see there was some discussion of the proposed CRUD protocols
tantek joined the channel
# 23:16 bret is platform the right word for indieweb implementations?
melvster joined the channel
caseorganic, scor and benwerd joined the channel