#dev 2024-08-25

2024-08-25 UTC
aaaa, terf, bugliker, GuestZero, [tantek], [KevinMarks], IWSlackGateway, Lars-Christian, [dougbeal], robalex, [dave], [snarfed], [manton], [Joe_Crawford], ttybitnik, [schmarty] and gRegor-mobile joined the channel
#
gRegor-mobile snarfed: https://gregorlove.com/2024/03/followed-will-norris/
#
[tantek] gRegor++ for another follow post example!
#
Loqi gRegor has 27 karma in this channel over the last year (123 in all channels)
#
[schmarty] [aaronpk] updating the indieweb webring for http://indielogin.com + the json metadata doc. can you add `https://xn--sr8hvo.ws/auth/id` to the list of acceptable `client_id`s? i can also just pass in the homepage, which should already be in the allowlist.
#
aaronpk [schmarty]: done
#
Loqi aaronpk: vikanezrimaya left you a message 21 hours, 22 minutes ago: IndieAuth spec, Section 4.2.1, makes a conflicting claim about whether `client_uri` MUST be rejected if not a prefix of `client_id` or merely elicit a warning if it is not on the same domain. I suppose this should be added as errata, what's the process for this?
#
aaronpk vikanezrimaya: that's probably a bug from the latest update that made some related changes, i can just publish a new version
#
aaronpk good create day project if i can make it down there later this afternoon
robalex joined the channel
#
Loqi aaronpk has 45 karma in this channel over the last year (135 in all channels)
#
[schmarty] aaronpk++ thanks! still seeing `This client_id is not registered (https://xn--sr8hvo.ws/auth/id)` maybe a cached response? 🤔
#
[schmarty] also: hope you can make it down here!
#
[schmarty] (aaronpk: using the original client ID url works, so i'm guessing i can just try again later w/ the `/auth/id` path if it's cached, or something will need kicked. no big deal.)
Lars-Christian and bterry joined the channel
#
[snarfed] [tantek] https://github.com/themattharris/RelMeAuth/issues/8 re https://tantek.com/relmeauth/ 😆
#
[schmarty] anyone with rel=me auth (but not an indieauth authorization server) configured on their homepage want to test out relmeauth support on the indieweb webring? https://xn--sr8hvo.ws/
#
sebbu invalid_session
#
sebbu The session was missing data. Ensure that you are initializing the session before using this library
#
sebbu on indielogin.com it supports only mail & github, on indieauth.com it support mail, github, gitlab & gpg, but both works
#
sebbu (with my domain sebbu.fr)
#
sebbu [schmarty]
#
[schmarty] thanks! taking notes right in indieweb-meta for IWC PDX but will check out it. appreciate you trying it!!
#
sebbu i commented out my auth/token endpoint long ago since the (standalone file) projects got stalled/abandonned/unmaintained, and i put rel=me on my website for tons of sites (mainly to have them validate my own site on my profile on them)
#
sebbu maybe we should support activitypub/mastodon for relmeauth ;)à
#
sebbu ;=)
#
sebbu (and atproto/bluesky!)
AramZS joined the channel
#
[tantek] [snarfed] oh no I need to do some work to update that
#
[schmarty] thanks all for patience with my broken demo! my opinion of hugo's greatest crime: `execute of template failed at <time $item.published>: error calling time: unable to parse date: 2024-06-24 17:16-0700`
#
[tantek] RFC3339--
#
Loqi RFC3339 has -1 karma over the last year
Lars-Christian joined the channel
#
sebbu [schmarty], it's missing second
#
sebbu ISO8601++
#
Loqi ISO8601 has 1 karma over the last year
#
[schmarty] sebbu: oh i know, haha. displaying reply context info from other folks' posts on my hugo site has been a fun nightmare of date formats.
#
[schmarty] (i've got a long-shelved project to have PHP's more liberal parser clean them up. until then i wait until i notice my site is missing posts and then find and fix them manually.)
#
sebbu well, I have a long-shelved project to actually make a homepage
#
sebbu instead of having a random index with relmeauth & other meta-data, then some unindex stuff under it (dokuwiki, wordpress, etc...)
#
ptramo[d] I have a long-lived project to actually make a simple crossplatform configuration management tool to install my packages and dotfiles
#
ptramo[d] instead of playing with the homepage for it 😄
GuestZero joined the channel
#
[schmarty] ok. RelMeAuth should now work with the webring!
#
capjamesg[d] ptramo[d] I would love better Python package management.
#
capjamesg[d] 😄
#
capjamesg[d] I know that's only tangentially related to what you mentioned, but it's an increasingly constant yearning for me 😂
#
ptramo[d] capjamesg[d] not enjoying poetry?
#
capjamesg[d] I haven't used poetry much!
#
capjamesg[d] Maybe that's where I'm going wrong.
#
ptramo[d] IDK much, but I find it bearable. Note I've only used it for work for less than 2 months after virtually no python for over 10 years.
[morganm] joined the channel
#
[snarfed] hacknorris re http://hacknorris.myremotecentral.com/hacknorris , you need Content-Type
#
[snarfed] https://www.w3.org/TR/activitypub/#retrieving-objects
hacknorris joined the channel
#
hacknorris just asking. someone knows why i can't find myself on a "real" mastodon? my pseudo-profile here : http://hacknorris.myremotecentral.com/hacknorris (not a page, just content which should federate SMH…)
#
[snarfed] hacknorris you need to serve the Content-Type header, https://www.w3.org/TR/activitypub/#retrieving-objects
#
hacknorris i use apache defaults…
#
sebbu The authorization server of the returned profile URL did not match the initial authorization server
#
sebbu [schmarty], invalid_authorization_endpoint
#
hacknorris i don't have thought of doing it pseudo-static…
#
sebbu it works with https://sebbu.fr/ but not with https://sebbu.fr
#
hacknorris all of it is just literal files, so…
#
hacknorris like at me
#
sebbu my mastodon profile is https://mastodon.social/@sebbu2, and it works correctly :D
#
hacknorris "hacknorris" is a file, not folder…
#
sebbu (i don't self-host)
#
hacknorris i see
#
hacknorris anyway - gonna be tomorrow, it's 11pm elaready (server goes off too)
#
hacknorris *already
#
[schmarty] sebbu: good catch. the URL without a trailing slash is "technically" not valid, while the trailing slash is, so that makes sense. but i do have some normalization code that i think should be fixing that up. i think i'm putting it away for the day, though. glad one of them works!!
#
sebbu earlier today, just typing the domain completed it with http instead of https
#
sebbu it should check both (preferably https, then http if it's not there)
GuestZero, [aaronpk], ms_boba[d], IWSlackGateway, [snarfed], [tantek], Killeik, [schmarty], [Joe_Crawford], [dave] and claudinec joined the channel