2013-09-27 UTC
tantek, lukebrooker, tpinto, ozten, caseorga_ and singpolyma joined the channel
tilgovi and jmsmcfrlnd joined the channel
# 00:35 jmsmcfrlnd ok, I think I am in indiewebcamp irc channel.
# 00:38 jmsmcfrlnd I am using irc from yaaic, don't know what to put for ident. thinking I need to read docs.
jmsmcfrlnd, christoffer__, caseorganic and singpolyma joined the channel
# 01:07 tantek tommorris - re: that post of yours "I don't want an app for your media site"
# 01:08 tantek I'm trying to figure out how to capture this "proprietary app for media site" anti-pattern
lukebrooker and singpolyma joined the channel
# 01:10 tantek I tried searching for that post and couldn't find it
# 01:11 tantek and see, I've been calling this the Kanye-UX anti-pattern
# 01:12 tantek "Imma let you finish, but this native app is the best app ever!"
# 01:13 abrereton i've got to let you know that you made me laugh out loud
christoffer__ joined the channel
# 01:22 poppy i wish i knew how people thought that was a good idea, but i guess there's a lot i don't understand
# 01:23 poppy xda-developers does it, which boggles the mind
caseorganic joined the channel
# 01:57 tantek poppy - they want an icon on your home screen
# 02:08 poppy tantek: likely, but with xda i'm not sure that makes sense
# 02:10 bret Ima let you get to dat site, but we spend way to much money on this shitty app to not telly you about it
# 02:12 bret one thing I found out when doing my own site. Mobile browsers are actually super duper fast if you don't like 10Mb of JS that has nothing to do with UX
# 02:13 bret faster than native apps that have to pull in internet data
# 02:13 bret like the iOS app store app... loading..
squeakytoy, singpolyma, christoffer__, jernst, tpinto, jmsmcfrlnd, tantek and ozten joined the channel
squeakytoy, christoffer__, josephboyle, icco, spinnerin and singpolyma joined the channel
# 04:50 abrereton wow talk about wrong window and *really* out of context
tpinto, spinnerin, icco, spinnerin_, singpolyma, eschnou, cweiske, LauraJ and tantek joined the channel
christoffer__, singpolyma, tpinto, dkordic_, eschnou, scor, andreypopp, gjones and caseorganic joined the channel
singpolyma, tpinto, vrypan, scor, LauraJ, josephboyle, caseorganic, barnabywalters, gjones, friedcell, bnvk and pfefferle joined the channel
gjones, Nadreck and LauraJ joined the channel
# 13:25 michel_v are these cuneiform squares, or are they really exotic characters that vanilla OS X can't display?
singpolyma, squeakytoy, gjones and barneybook joined the channel
# 14:36 barnabywalters great to see others having the same thoughts about owning comments and responses on your blog
# 14:37 barnabywalters pity they decided to add an unnecessary layer of siloed data over that though
# 14:37 barnabywalters I’ll ping them about indieweb commenting, see if they’d be up for adding mf2 support
# 14:38 barnabywalters also a potential POSSE destination if anyone decides it has enough value to automate
# 14:39 barnabywalters it’s a pity everything in their UI works with js and no pushState — I like permalinks
# 14:39 barnabywalters interesting that their post frame doesn’t indicate anywhere that this is a reply to something
# 15:02 pfefferle would be nice to finally move the repo to go on in developement even if someone is on vacation or somewhere
# 15:02 barnabywalters there are a bunch of issues which it looks like sandeep fixed with a5e6785 but didn’t manually close — I’m going through and closing them
bnvk joined the channel
bnvk_ joined the channel
scor joined the channel
# 15:10 barnabywalters is this what GH milestones are for? issues we want resolved for a particular version?
scor joined the channel
# 15:14 barnabywalters pfefferle: I’m not as up to date with webmention issues as you, can you go through and assign the v0.2 ones to that milestone?
# 15:16 tommorris barnabywalters: Bit too much legacy there. OpenID doesn't work great with Wikimedia's SUL (Single Unified Login) system. When you log into Wikipedia, you are actually logging into about 700 sites at the same time.
# 15:24 barnabywalters apart from not specifying the target URL, and overloading with a bunch of untrustable on-the-wire data about the post
_6a68 joined the channel
# 15:28 tommorris TrackBack's RDF stuff is pure nonsense. RDF-XML inside HTML comments
# 15:29 tommorris It's worse than conditional comments. It's cargo-cult semantics.
singpolyma, ozten and josephboyle joined the channel
# 15:41 barnabywalters looking closer at trailed.io, they’re actually kinda breaking the web by wrapping it in their chrome and giving everything new permalinks which don’t link back to the originals with rel-canonical
# 15:42 pfefferle "HTTP Link: header or link element on the linked page" is the ":" right?
LauraJ joined the channel
# 15:51 barnabywalters they’re roughly split into bureaucracy, simple copy fixes and contentious issues
# 15:52 barnabywalters simple text fixes just need someone to add an initial dump and then people to improve it
bnvk joined the channel
# 15:58 barnabywalters sandeeps commit says 404 can be used to delete content, indiewebcamp wiki says otherwise
# 15:59 barnabywalters IMO this whole deletion thing is overthought, as the updating spec actually covers it
# 16:00 barnabywalters if there are issues with updating a comment to empty or “this post was deleted” in the wild, then we’ll figure out a solution
# 16:02 barnabywalters RE errors: are there any fail conditions where machine-readable errors will allow a sender to resolve them automatically?
# 16:03 barnabywalters e.g. "Specified target URL does not accept webmentions" is a sender error as sandeep pointed out, but making it machine readable isn’t going to allow the implementation to fix it automatically
# 16:04 pfefferle the problem is, that errors are part of the 0.1 spec… it would be dump to remove them for 0.2 and re-add them perhaps in 0.3… so we should be sure about that...
# 16:05 barnabywalters we should apply the selfdogfooding filter to potentially unproductive issue threads too
# 16:06 barnabywalters “if you haven’t implemented the technology on your own site, others are allowed to ignore complaints or suggestions”
# 16:10 pfefferle not really, because we introduce rel-webfinger instead of the uri
# 16:11 barnabywalters so BC implementations can run the regex then optionally check for the old rel? makes sense I suppose
# 16:12 barnabywalters the whole strict regex thing seems to make things easier for consumers at the expense of publishers
# 16:13 barnabywalters all webmention implementations I can think of do so anyway to get mf2 data out
benprew joined the channel
pfefferle, tilgovi, pfefferle_, benwerd and jschweinsberg joined the channel
singpolyma joined the channel
pfefferle and tantek joined the channel
# 16:51 barnabywalters RE webmention errors, adactio’s human-friendly implementation has changed my thoughts futher
eschnou joined the channel
# 16:52 barnabywalters so ideally success and error pages should always be HTML pages with friendly copy on
jernst_ joined the channel
# 16:57 tantek barnabywalters - adactio says he just has a check in his code
# 16:58 tantek if the webmention comes from his own site (e.g. via the form), the return result does a redirect back to the source page
singpolyma, adactio and plevin joined the channel
# 17:16 aaronpk barnabywalters: maybe we should start using the indieweb fork of webmention for issues
# 17:17 barnabywalters I’m also sending pull reqs to the converspace repo, should probably have forked the indieweb one
# 17:18 bret what has sandeepshetty been up to? he was around so much for a while now I never see him :(
# 17:21 aaronpk barnabywalters: feel free to re-open issues on the indieweb fork, we can keep going there
eschnou joined the channel
# 17:24 aaronpk barnabywalters: also webmention.org is hosted on github pages, so if you can push to the gh-pages fork on converspace it'll update the site :)
# 17:24 benwerd nobody's doing full reply-context in idno as far as I know
# 17:26 barnabywalters aaronpk: bleargh, so much copy+pasting. RE GH pages good point. not sure I want to mess with it too much without sandeep’s signoff
# 17:27 jernst_ hi tantek, sorry missed your hi earlier
# 17:28 aaronpk of course the real solution would be to get sandeep to just move the repo over to the indieweb org. then all the issues would be transferred with it.
# 17:36 aaronpk we've got the updated error response stuff in 0.2 right?
# 17:37 bret maybe we should put it on a path if sandeep hasn't okayed it yet, and its his domain
# 17:37 aaronpk all the versions should probably live at 0.1 and 0.2
# 17:39 bret barnabywalters, just make a file called 0.2.md or 0.2.html
# 17:41 aaronpk barnabywalters: I don't even think this is jekyll, it's just HTML
# 17:41 bret i think aaronpk is right, hes using the page generator
# 17:41 bret so just create a folder called 0.2 and put an index.html inside of it
# 17:44 aaronpk oh yeah it's using the page generator cause it's that default theme
# 17:44 bret and the content is generated off of the readme or something?
# 17:45 bret i think you can start messing with the raw html though
# 17:46 bret not sure if its the right way to do it but it works
# 17:46 aaronpk I push stuff up to github pages all the time that's just html stuff
# 17:46 barnabywalters so I’m not worried about pushing the 0.1 folder, what about the changes to README though? will the GH pages site automatically update?
# 17:48 bret barnabywalters, actually no, you have to use the page generator again and reload the readme
spinnerin joined the channel
# 17:48 aaronpk you can also just copy/paste stuff into the html file
# 17:50 barnabywalters github has managed to do something very impressive with gh-pages. they’ve managed to make publishing static HTML complicated
# 17:50 aaronpk gh-pages by itself is very simple. any html files in the gh-pages branch are published.
# 17:50 aaronpk but it's all the stuff around generating that html that's complicated
# 17:50 bret there are a ton of subtleties with the generators yeah
# 17:59 bret barnabywalters i think you hit the nail on the head, the way the tool works is interesting
# 17:59 bret you can just overlay on top of whatever you use
# 18:01 aaronpk i'm going to have to move/delete the indieweb fork in order to make room for sandeep to move his
pfefferle joined the channel
# 18:05 aaronpk ok, indieweb fork deleted so sandeep can transfer ownership
# 18:18 pfefferle aaronpk: What do you think about making the gh-pages the default master for the webmention spec and Jekyll for rendering the HTML
# 18:20 pfefferle And you can also add canonical urls for the different specs
# 18:21 aaronpk bret: if you want to take a stab at it, you can do it in your own fork of the repo, then we can pull in the changes later
# 18:21 aaronpk also once it's moved it'll update the ref automatically
# 18:21 pfefferle And if you use the gh-pages as master you don't have to copy everything around
# 18:23 bret aaronpk, let me finish up these tickets for work and then ill get started
Nadreck and LauraJ joined the channel
# 18:24 bret has the readme drifted from the markdown readme I wonder?
eschnou joined the channel
cweiske and pfefferle joined the channel
# 18:50 pfefferle aaronpk: You could do the same for the indieauth spec btw
# 19:03 cweiske does barnabywalter always publish but reports and comments on his website first?
_6a68 joined the channel
# 19:08 bret cweiske, not sure what you are asking?
# 19:09 cweiske my question is if he does this for every single comment and bug report
# 19:09 bret oh hrrmm… i dont know, but aaronpk and him were discussing POSSE to gh issues
# 19:10 bret i wonder if they do it through the gh api
# 19:12 cweiske at least did I get mails from my pingback server that he posted the notes
benwerd_ and pfefferle joined the channel
Chat6534 joined the channel
# 20:22 bret ok, going to change buildings and work on the jekyll port
pfefferle joined the channel
# 20:24 cweiske may I also return a 200 OK to a webmention request when I already processed it?
# 20:24 cweiske "202 accepted" should only be for to-be-processed-later requests
barnabywalters, singpolyma, benwerd and andreypopp joined the channel
benwerd and pfefferle joined the channel
# 20:46 aaronpk i'll try that later. don't have time to re-shuffle it right now
vrypan and singpolyma joined the channel
# 21:08 barnabywalters github strips out the class attribute but not all the unnecessarily complicated microdata attrs
# 21:09 barnabywalters I was hoping I could trick GH into publishing mf data by giving it mf data. apparently not :(
benwerd and tilgovi joined the channel
# 21:34 aaronpk barnabywalters: that sounds more like a limitation of github than a benefit of microdata
# 21:35 barnabywalters aaronpk: very true. although I can see why it’s safer to allow attributes which won’t affect styling
# 21:36 barnabywalters the GH API is a mixed bag. really easy to post to, loving the basic token auth option
# 21:37 barnabywalters but response vocabs are really inconsistent. mixing owner/user (author), name/title (always name)
caseorganic, _6a68, XgFgX, caseorga_ and andreypopp joined the channel
# 22:02 barnabywalters vrypan: nice post! great to see your static site + indieweb comments setup documented
# 22:06 vrypan I also try to get a text snipet from the source url, by going up the DOM tree and getting the <p> that contains the lin to target.
# 22:06 vrypan I'm not sure if this is the best way, but it seems to work nicely in practice.
# 22:06 barnabywalters vrypan: the best way of doing it would be to parse the HTML for microformats and try to find an h-entry
# 22:08 barnabywalters (I made a PHP one and know python so could probably help out if you’re up for it)
# 22:09 vrypan I'll let you know if I try. Probablyt will, but I'm not sure when...
singpoly1a, singpolyma and andreypopp joined the channel
# 22:27 jernst_ hey Loqi, which time zone are you on?
# 22:29 jernst_ this ephemeral being, never sleeps I'm sure, and doesn't know what a morning is either, I think it just picked that phrase up from somewhere
# 22:29 bret are the github pages templates in their own repo somewhere?
tpinto and singpolyma joined the channel
# 23:14 bret aaronpk, looking for feedback on how to structure the jekyll site a bit
# 23:28 bret If we use jekyll, we have to make some decisions about how to deal with the readme vs index page
shaners joined the channel