#kylewmoh man this is very confusing. I want to get authorization credentials from Wordpress.com, but later also want to use it as an identity provider ... should I create two different "Applications" for those two purposes?
#kylewmbecause the SiloPub Wp.com application only allows one redirect_uri to be used, and I don't know how to use the same one for those two very different purposes...
j12t, Sebastien-L, KartikPrabhu, fkooman and ttepasse joined the channel
#tbrbpetermolnar: I think it went pretty well actually, most people seemed to take something from the weekend and there was a lot of good hacking done on the sunday
#LoqiWelcome, indie-visitor! Set your nickname by typing /nick yourname
petermolnar and wolftune joined the channel
#mxstbrAs Lukas Rosenstock suggested on twitter, (https://twitter.com/LukasRosenstock/status/625645331821273088) I'll share a project of mine here, pgp.asc. (http://pgpasc.org) The idea is to decentralise public PGP keys by storing them on your own server, in the root folder, in a file called pgp.asc. That way, to get a public PGP key you just have to go to recipientdomain.tld/pgp.asc and you'll get his key. E.g. if you go to http://m
#mxstbrpetermolnar Whichever one you want to share
#petermolnarbut how will the mail address be discovered?
stream7 joined the channel
#mxstbrpetermolnar: it's assumed the sender already has it/has an easy to find it (e.g. a mention on your webpage)
#voxpelliif the e-mailer claims to be the same user as site X and you on site X finds a public key that can verify a signature included in the e-mail then that verifies that claim
#petermolnar( by the way I had /pgp for this already, but not as plain text; the pgp.avc does make is more reasonable )
#aaronpkmxstbr: for the GPG support for indieauth.com, I let people put their GPG key wherever they want, including on a different hostname, and just link to it with a rel=me attribute from their home page
#voxpelliit and webfinger – I still find them a bit interesting, but lack of momentum around them makes me not give them a lot of attention
#aaronpkthe benefit is you can put the GPG key wherever you want, regardless of what backend storage mechanism you're using. and if you want, you can even put it on a different server
#aaronpksome people link directly to their key on keybase.io that way
#voxpellitoo bad there doesn't seem to be a mime-type for such keys – otherwise a <a rel="me" type="a-relevant-mime-type"> would have been kind of neat
#mxstbraaronpk yeah, but then you need a webpage to parse. Having the pgp key at /pgp.asc, no matter if you have a website or not has the benefit of not needing a webpage.
#aaronpkthat feels like an edge case not worth optimizing for
#aaronpkespecially since it still requires web hosting
#aaronpkwhen I have seen people who have a domain but don't have a website there, it's because they don't have hosting for the domain and there is no A record for their domain
#mxstbrDon't feel like that's necessarily true, one of my domains (mxs.is) is hosted on a virtual host that is solely a link shortener — no homepage there.
#mxstbrMaybe it's just me though, that's possible. :-)
#aaronpkwhy would you want your pgp key on your short domain tho?
#aaronpkI'd love someone to help with design for some of my projects like indieauth.com and webmention.io :) I mostly use bootstrap and do as little as possible haha
#snarfedi tried teacup against my test tumblr just now, and i think teacup worked, but http://snarfed.tumblr.com/ doesn't show it. maybe because teacup doesn't pass content?
#snarfedalso jonnybarnes minor feature suggestion, i first tried logging in with snarfed.tumblr.com and it errored. maybe consider adding http:// automatically in that case.
#KartikPrabhusnarfed: yeah weird. because it is the same code as on my website which seems to get your webmentions fine afaik. will track the problem
#snarfedjonnybarnes: another bug report, when i clicked on Refresh Syndication Targets, it errored and then kept erroring on every page until i cleared cookies. just fyi
#fkoomansnarfed, it has the ability to view and post to private 'spaces', so you have to specify the URL of the 'space' you want to post to. Only if you are a member of the group associated with the space you can post...
#Jeenafkooman what problems do you have with it? Most of the things I had was that I had to change something in the php.ini to get it running, perhaps I can help?
#fkoomanJeena, thanks, I'll fiddle some more first :)
#kylewmsnarfed: thanks for the testing!! yeah the micropub radio button thing is a problem i've had before too. i need to fix it so that it has to submit before you can click "authorize micropub" or something like that
#aaronpkJeena: you shouldn't need to be on port 80
#aaronpksomeone logged in with their domain on port 8081 this weekend
jciv joined the channel
#Jeenaweird, I couldn't do it, it would just hang or something
#kylewmsnarfed: also thanks for the kind words about the instructions -- i borrowed a bunch from bridgy for blogs, especially the direct links to Customize HTML type pages
#aaronpkJeena: well until sunday one of the servers was blocking outgoing traffic except 80 and 443
#aaronpkparticularly if you've written either a micropub client or server
rascul joined the channel
#Jeenajeez, Facebook has such a shitty UI, something a friend of mine commented showed up in my stream, I wanted to find it, but I can't on his timeline it doesn't show up and in my stream it's gone too
fkooman joined the channel
#voxpelliaaronpk: so far I have assumed "content" to always be HTML – good with a generic approach for that
#aaronpkvoxpelli: if you assume it to be html, how would you expect to be able to type un-escaped things in a client? like if you're typing literally <blink> and you want the <> to show up in your note?
#aaronpkcause the other way this could go (which diverges from the microformats<->micropub mapping) is to say that clients should send escaped content
#voxpellithat would make all params e-* essentially
KartikPrabhu and tilgovi joined the channel
#voxpelligood to be able to differentiate between when one wants a e-* and when one wants a non e-* parameter – so +1 on your suggestion
#voxpellithen there's "q" for GET requests and "mp-action" for POST requests that decides what kind of request it is and everything else regarding that request is relative to the kind of action, right?
#voxpellisuch a setup is very flexible and future proof
#KevinMarksI'm trying to get 304 working with unmung hovercards, but hitting all kinds of oddnesss
#KevinMarkswhich wasn't showing up when I just dis a naive read
#aaronpkyes, as well as "access_token" for the post body
#voxpellibut +1 on cweiske – having clients fetch data from the individual page removes much of the point with fat pings from the hubs
#cweiskeaaronpk used to have fat atom/rss feeds which my reader only had to fetch once. now, even with push, my reader has to request each article by itself
#KartikPrabhuno idea what in hfeed2atom is causing my server to go nuts
#voxpellione way to solve that is to put the resolving in the hub – I believe Superfeedr has played around with some such things
mxstbr joined the channel
#tantekcweiske - that's an interesting observation re: # of HTTP requests
#voxpelliand the image data is often much worse than the Open Graph data – because honestly – how many web devs test the sites they design in RSS-readers vs tests them at Facebook?
#aaronpkright, but then clients go fetch the canonical html anyway to get the full article
#voxpelli(suprisingly few even tests them at Facebook!)
#voxpellicweiske: not just ads, statistics are at least as important
#voxpellisites want to know how many readers they have – and that very question was up in this channel as well – that people here wanted to know how many readers they would lose if they dropped their RSS-feed
#aaronpkit just occurred to me that this is analogous to podcast distribution
#voxpellione difference is that audio (and photos) are static – harder to put a strain on a server when it's raw bandwidth rather than dynamic generation
#cweiskemost podcasts have multiple formats, it would make no sense to put them all in the same feed
#voxpelliand it can be cached efficiently compared to an article page
#cweiskeif only one format is then used by the client anyway
#cweiskebut the HTML in the feed *will* be used, because there is no other format that delivers the same data
#aaronpkinteresting observation re: dynamically generated pages vs static assets
#voxpellibut it would probably be a problem if a podcast started doing PuSH and had thousands of clients all trying to download an episode at once
#kylewm"most podcasts have multiple formats" -- citation needed?
#voxpelliif one subscribes to a podcast through eg. iTunes – do Apple perhaps "fat ping" the audio by mirroring at their own servers and pointing clients there?
#kylewmhuh, I'm pretty sure I've only ever seen that kind of thing on archive.org before
#M-KeganIsn't this mainly a bandwidth problem though? In the ideal world every rss feed would spit out all the info you need, but that's not the point of pinging the client like this though imo
#GWGI am wondering if anything in the IndieWeb community would make for a good podcast
#M-Keganvoxpelli, sure that would help with thundering herds
#voxpelliand the fact that it is static means that if one have a big audience one probably puts the audio on a CDN
#M-Kegancweiske, I don't get why this is any different between rss feeds and pubsub. If the server is going to send me a crap ton of data, I'd rather it didn't.
#voxpelliso – the conclusion here can perhaps be that if everyone were to switch their sites to jekyll then we could skip fat pings altogether – right? :)
#cweiskebecause there is no other format that delivers the same data
#GWGtantek, I think I may need to replace zachdonovan for a while
#M-KeganThe core data needs to be supplied, but that's it imo. Send too much stuff and you also start increasing latency as my client desperately tries to download older content whilst newer stuff builds up.
#ben_thatmustbemeaaronpk: i think that makes sense, and is exactly what i do already
#voxpelliwell, just like you download a podcast when you want to hear it you can download the content of an article when you want to read it – the only really necessary stuff is URL, date – and a title is kind of nice :)
#aaronpkshould that be sent in the fat ping as data URI?
#kylewmsnarfed: does Bridgy use/handle refresh tokens for blogger?
#aaronpk(i'm illustrating a point here, not actually suggesting doing this)
#voxpelliit's amazing what people sometimes put in their metadata
#M-Kegancweiske, yes but my point is that the diff may be large.
#M-KeganEg connectivity losses and the next time you ping you get lots of backlog.
#voxpelli(eg. found someone that put their entire blog posts within an og:details – one then hopes that one has put an upper limit on the size of that field within ones parser and data store :P)
#cweiskewhich "core meta data" are the ones you would expect in a feed?
#M-KeganThere needs to be enough information for that client to make an informed decision on whether they want to "click" and get more info.
#voxpellidate because you need to know if it's new and relevant and title isn't really necessary but most presentations need it – kind of a notification first design approach as I believe tantek has been pushing for :)
#M-KeganIt may be a title, a short summary, it will depend on the content.
#rhiaroI didn't read backlog, but just wanted to mention that I had no idea indieauth for login was a thing until I saw tbrb implement it at the end of last week, and omg I want it
#Loqimicropub chaining is when you have one micropub endpoint itself use micropub to post to another micropub endpoint, like to POSSE a post to silos, or to a community site or other commons https://indiewebcamp.com/micropub_chaining
#ben_thatmustbemeinterestingly i can set my micropub endpoint to syndicate different content now to twitter
#LoqiSalmentions are a way to pass comments upstream by sending a webmention from a reply post to the original post when the reply recieves a comment https://indiewebcamp.com/salmention
#KevinMarkshm, how do I mark up inline native comments?
#KevinMarkshm, now I'm editint he template and nothign is happening
#tanteklocal comments are comments that are made directly (locally) on a post in the UI of the site serving the post, instead of posting a [[reply]] post on an indieweb site and sending a webmention.