#GWGTell me if there is anything else I can stick in. I've put in a lot for this next version.
#GWGThe time fields set up for all sorts of activities.
#GWGI still need to add a bunch more in terms of options. I have featured images, original publish date, etc.
arlen, squeakytoy and [aaronpk] joined the channel
#[aaronpk]Everyone keeps recommending webmention.herokuapp because it has a JS widget that actually renders comments, not just an API like <http://webmention.io|webmention.io ;-)
#aaronpkreviewing the open issues on webmention.io, it looks like a few of them i'm blocked on because they are actually issues with the Ruby mf2 parser :-(
#kylewmaaronpk: think the mf2 parser can be fixed or easier for someone to write a new one from scratch?
#aaronpkprobably can be fixed, but I haven't really looked at the guts of it
#kylewmi got the impression at one point that it was vocabulary-aware
#aaronpkkylewm: you did something recently where you wanted an indieauth client to get a URL back that was different from what the user entered, right?
#kylewmi wanted to give my indieauth my url as multiusersite.com and it would return multiusersite.com/profiles/kylewm as "me"
#aaronpkwhat's to stop your multiusersite.com returning a "me" url of "aaronparecki.com'?
#aaronpkdoing this opens up this security hole. we'd need to ensure that an indieauth client only allows the server to return a "me" url on the same domain. problem is there is no technical way of doing that.
j12t joined the channel
#kylewmI *think* it only opens up the security hole if you cache the indieauth/token endpoints between requests
#kylewmif you rediscover the endpionts with the new "me" url, then your request for verification or aaronparecki.com with aaronparecki.com's token endpoit will fail
#aaronpkthis also means that proper handling/validation of the "state" parameter is *super* important, so that you can't trick clients in to logging people in without the client expecting it
#aaronpkwell one easy thing if you don't want to switch to JWT auth codes is you can delete them after they're used once, or you can set an expiration time and then every time a new one is generated, delete all the old ones that expired
#kylewmah yeah definitely makes sense. i will poll the Known folks for interest
catsup joined the channel
#aaronpki'm a huge +1 for known being its own auth server and token endpoint instead of relying on indieauth.com
#[shaners]<@U03QR2B3P kylewm: if either or both if you are willing to work on the mf2 ruby parser, I'll work with you on it and get it released. I still have release credentials, I think.
#[shaners]The singular/plural seemed like a good idea and for our internal uses it was, but clearly it turned out not to be in the longer run.
#[shaners]We could break compatibility and release a 3.0 (to comply with SemVer).
Pierre-O, j12t, snarfed, kousu[xmpp] and kousu_xmpp joined the channel