#[snarfed]mblaney ah, bridgy needs the same web site in your FB profile that you IndieAuth with. looks like you authed with mblaney.xyz, and your FB profile has unicyclic.com
#mblaneyawesome! I'm getting a 429 at the moment but will have a look at it later
#mblaneylooks like I've been fetching content from facebook-atom but the html structure is different on mbasic
#mblaneyso I need to update my own rules about where to pull the content from
[tantek] joined the channel
#@fluffy↩️ The comments on my comic are probably the first thing I stop caring about, especially since most of my interactions come in the form of Webmentions these days anyway. (twitter.com/_/status/1362957674464223240)
#[snarfed][fluffy] saw your bridgy issue. sorry for the trouble!
#[snarfed]bridgy has supported webmention discovery via header pretty much forever. the issue you’re seeing is that it requires a 2xx response, so it’s giving up on your 401.
#[snarfed]aaronpk i looked through the spec a bit but couldn’t tell whether endpoint discovery should still succeed on non-2xx response. do you know?
#sknebelno spec justification, but I'd argue that if the receiver chooses to link an endpoint from a 4xx page, they have a reason to mean that
#aaronpkYeah interesting that we didn't think to mention it in the spec specifically
#aaronpkbut I don't see any reason discovery should be ignored on a 4xx response like sknebel said
#[snarfed]i guess yeah. i just expect that happens a lot right now, and if i start attempting them all, 99% of those attempts will also fail
#sknebel(I guess "login required", request-origin-dependent "blocked for legal reasons", ... are cases were a human might very well see a post, reply to it, but their WM sender then might get an error page)
#sknebelI guess that's something you could collect numbers on for a bit?
#[snarfed]yeah. and the numbers probably still don’t matter, i should just make all those attempts anyway
#aaronpki just haven't used facebook in so long i forgot what i did with that post
#[snarfed]yeah that’s what i’ve seen w/most of the browser extension users so far 😆
[tantek] joined the channel
#Loqisebsel has 3 karma in this channel over the last year (9 in all channels)
#[tantek]sebsel++ thanks for the review and test. merged
[tw2113_Slack_], jeremych_ and [fluffy] joined the channel; nickodd left the channel
#[fluffy][snarfed] ah, makes sense. But mentions to 401s will be very important for supporting private posts in general. Could there be an exception for that? For that matter I feel like any error page which declares an endpoint should still be consisted a valid target.
#[fluffy]Like consider a mention to a page that returns an error due to a transitory problem
#[fluffy]Wouldn’t want that mention to be lost forever!
#LoqiUsing rel=me on a hyperlink indicates that its destination represents the same person or entity as the current page, which is a key building-block of web-sign-in and IndieAuth https://indieweb.org/rel-me
#jackythis could be the thing I was looking for re: linking to a feed that's like just the Webmentions of a particular post 👀
#jackybut yeah, I kinda don't see the extra value-add _just yet_
#[KevinMarks]well, ages ago bradfitz built a crawler for all the xfn values, so you could query it for related accounts and friend links etc too, but it got deprecated
#jackybut what does finding 'related' accounts actually do?
#[KevinMarks]so you'd have photos on flickr and short notes on twitter and a blog and music plays on last fm and so on, and they did actually use the rel=me links
#[KevinMarks]right, but they have all stopped being curlable now
#[KevinMarks]plus a lot of detail on how it actually needs some messy code to canoniclize the links
[chrisaldrich] joined the channel
#btremIs there a place to get demo webmentions json to test rendering? What I see on webmentions.io is json for a single url, but the code I borrowed from sia.codes gets webmentions for a site and filters, and it's screwing up, and I really need test data.