#dev 2021-01-10
2021-01-10 UTC
[chrisaldrich] and [dmitshur] joined the channel
# [dmitshur] woohoo, thanks for reviewing the grant_type PR Aaron!
# [dmitshur] I can once again sign in to indieweb.org without any temporary workarounds, it feels good.
[snarfed] joined the channel
# lahacker i wonder what others' auth/token endpoint UXes look like; keep a running list of client_id/redirect_uri/scope/date and provide a Revoke button?
# lahacker right.. thanks
# lahacker i've got the simplest case of sign in/out working cross-browser using the `identity` WebExtension API if anyone's interested: https://github.com/angelogladding/liana/blob/main/background/auth.js
# lahacker wondering if i can't turn that file into a library
# lahacker k i'll take a whack at decoupling and packaging it and let you know
# lahacker yeah it's IndieAuth
[tw2113_Slack_] joined the channel
# lahacker does anyone know of any indieweb projects with i18n?
[tantek] and treora joined the channel
[schmarty], [tw2113_Slack_] and robindrake joined the channel
# [tantek] worth reading: https://www.w3.org/2021/01/09-social-minutes.html
nickodd, ShadowKyogre, leg, oodani, swentel, KartikPrabhu, jdp, sebbu and kira joined the channel; ShadowKyogre left the channel
dhanesh, [Raphael_Luckom], [KevinMarks], shoesNsocks, KartikPrabhu and nickodd joined the channel; ShadowKyogre left the channel
# jamietanna[m] Anyone used Postman before? I'm building a thing for the Postman hack around IndieAuth (2020 updates) and Micropub and may be looking for some testing for folks who support a couple of specific extensions
[tantek] and KartikPrabhu joined the channel
KartikPrabhu, shoesNsocks and jamietanna joined the channel
# jamietanna Hmm, it may actually not work as well as I'd hoped, I didn't realise folks usually kept their q=config requring authentication
# jamietanna Looking at the first three folks I was gonna test with, they're all blocked - you, Aaron and Manton
# jamietanna https://micropub.spec.indieweb.org/#discovery shows the example with a token, and says it must allow the same tokens, but doesn't say whether it _must_ be authenticated. I guess most folks do as it's simpler?
# swentel I guess https://micropub.spec.indieweb.org/#authentication-0 enforces it
# jamietanna Ah yep :(
# jamietanna May have been something I'd gone non-standard on then, as I felt it didn't need to be required for a lot of the config
# jamietanna That's gonna make this much harder to test / implement :D
[snarfed], [Rose] and [schmarty] joined the channel
# jamietanna well, aside from the authentication issue, it seems to work pretty nicely, and generates me a handy Postman collection with all my post types and queries supported :D
# jamietanna I'll probably get a blog post up tomorrow about it, as I'm not sure I'll have a chance today - been good fun working on it
# jamietanna[m] But it anyone has an unauthenticated config endpoint gimme a shout so I can test my tool!
# jamietanna[m] I think it'd be good to revisit, even if it's a "MAY not require authentication" to allow folks to relax it if they want
# jamietanna Yeah I think things like that and q=syndicate-to or what `q`s / `post-type`s do I support
# jamietanna IMO there's not much sensitive in q=config
# jamietanna but happy to let folks keep it secure if they want to
# [Raphael_Luckom] it would be interesting to see a temp graph superimposed on a sentiment-analysis graph. Are they correlated?
# @15VirginiaLee PRE ORDER SALE 99c! #preorder #sale #kindle #nook #kobo #ibook #izzymatthews #Sale #improbableodds #indieauth (twitter.com/_/status/1348359152284950532)
paulrobertlloyd joined the channel
# paulrobertlloyd Been thinking about consuming/producing mf2 and JF2 in Indiekit, and which format to use when: https://getindiekit.com/indiekit/decisions/0003-store-post-data-as-jf2
# paulrobertlloyd Curious is anyone has felt similar pain juggling the 2 formats?
# [Raphael_Luckom] interesting, hadn't noticed jf2
# [Raphael_Luckom] I haven't gotten to the point of adding mf2 support to my site yet, but I assumed that I'd use that markup in my html. jf2 adds an interesting twist.
# [Raphael_Luckom] https://jf2.spec.indieweb.org/#jf2feed
nickodd left the channel
# [Raphael_Luckom] One thing that seems a little awkward about mf2-in-html is the "first microformat on a page wins" dynamic. That makes me think that maybe what I'd want is for each mf2-document to have its own unique URL, which doesn't seem to be how they're usually used. I've been thinking about making that the default when I get around to implementing it--each microformatted item would be published standalone on a url only intended for
# [Raphael_Luckom] machine-readability, in addition to being incorporated in whatever human-readable content I'm writing. In that type of system, there would be no tension between publishing both mf2 and jf2-versions of everything, just like you could offer both rss and atom feeds of blog posts.
# [Raphael_Luckom] If that's the expectation, the inability to distinguish between multiple same-type microformats on a single html page seems like a pretty big usability challenge, unless there's a way around it I haven't seen yet.
# paulrobertlloyd I’m currently publishing a JF2 feed for my site for the reasons outlined here: https://getindiekit.com/indiekit/decisions/0002-share-publication-state-using-jf2
# paulrobertlloyd But that’s intended as a private feed, almost
# paulrobertlloyd Indiekit doesn’t publish JF2, but does use it internally for populating templates etc.
# aaronpk for example take a look at this page which describes how to find the author of a post, where there is absolutely no assumption that there is only one microformat on the page https://indieweb.org/authorship-spec
# [Raphael_Luckom] Those are interesting, and yes, they address part of what I was talking about. But when you're sending webmentions, I think you've still got the issue from this earlier comment: https://indiewebcamp.slack.com/archives/C1PA11USK/p1610226286085400
ShadowKyogre and [KevinMarks] joined the channel; ShadowKyogre left the channel
# [KevinMarks] This is where fragments in the url can help
# [Raphael_Luckom] I don't understand
[Tim_Nolte] joined the channel
# aaronpk now if you send a webmention with a source URL https://example.com/#entry-2 the receiver knows to look for the object with that URL on the page
# [Raphael_Luckom] right. What I got from what you said before was that, since I don't control the webmention-receiving code, the only way that I can make sure the right thing is parsed is to make sure it's first on the page.
# [Raphael_Luckom] As a proxy for "what to expect," is there a standard that says that webmention receivers _should_ handle url fragments?
# aaronpk i guess there's some notes here https://indieweb.org/comments
# [Raphael_Luckom] I get that it's not reasonable for me to ask for every independent implementation to do what I think is reasonable. But I also don't want to make structural decisions based on an informal assumption.
# aaronpk also that sentence was carried over from the original draft from 2014 and never revisited really https://indieweb.org/wiki/index.php?title=comments&oldid=2320
# [Raphael_Luckom] I think that's the only issue I have. If url fragments are definitely included in "the URL", then I don't think there's anything left for me to argue about.
# [Raphael_Luckom] maybe that's already the natural way to read that, idk
[chrisaldrich], jdp and [snarfed] joined the channel
# [snarfed] bridgy has a couple feature requests for this: https://github.com/snarfed/bridgy/issues/681 , https://github.com/snarfed/bridgy/issues/445
[tantek] joined the channel
# [KevinMarks] would this be worth a spec iteration, or some new tests in webmention.rocks?
# aaronpk there's even a few test cases available! https://authorship.rocks/
# aaronpk it looks like /comment-presentation is a redirect to /comments ?
# aaronpk the "how to receive likes" bit is on a different page https://indieweb.org/like#receive_and_markup_incoming_likes
# GWG So, would it be better if we use /responses ?
# aaronpk at the very least an immediate first step could be to add links on /responses to the section on the page that talks about how to receive each type
ShadowKyogre left the channel