#dev 2021-02-14

2021-02-14 UTC
samwilson and [tantek] joined the channel
#
[tantek]
what is Tanzawa?
#
Loqi
It looks like we don't have a page for "Tanzawa" yet. Would you like to create it? (Or just say "Tanzawa is ____", a sentence describing the term)
#
[tantek]
also nice Quill screenshot!
KartikPrabhu joined the channel
#
cambridgeport90
This seems like something new that I've not heard of until now.
KartikPrabhu joined the channel
#
cambridgeport90
I just looked it up and couldn't find anything.
#
jacky
!tell [chrisaldrich] good question. I want to say that I try to go in the order of h-feed, json-feed then rss
#
Loqi
Ok, I'll tell them that when I see them next
#
jacky
b/c from my experience, rss feeds tend to be the most variable (either super stripped down to force click throughts or more verbose with fluff squeezed in to encourage engagement back to a site - like a CTA)
#
jacky
I kinda wonder if there's a "programmatic" way to pick such a feed for people
#
jacky
maybe the social reader renders a post it finds in three feeds and allows the user to pick the most ideal one (and that's then used to determine what feed to use going forward?)
#
aaronpk
Together does that, there's a thing in the microsub spec for feed previews
#
cambridgeport90
I think the question is which feed is the most detailed? And depending on the site configuration, a few could be available. I have both the default RSS and then the JSON Feed plugin for Wordpress installed, a well, so bringing up my site in a reader would probably fetch both?
#
jacky
cambridgeport90: it should!
#
jacky
aaronpk: I'm thinking about something that might even -hide- the feed types used and allow for people to like "swap" between different rendering forms to see which one works 'best'
#
cambridgeport90
I never know which is the best one to use; still learning, hence why I'm following the dev channel.
#
jacky
like there might be a bit of a blip for some in their ATOM feeds than what you can garner from a h-feed (for example, reply counts)
#
jacky
cambridgeport90: I think that people _shouldn't_ have to know what feed to use
#
jacky
I think this was discussed in here before
#
jacky
but I do want to see if it's possible to do that in a deterministic way for people
#
cambridgeport90
I haven't tried Koype yet, though I intend to.
#
jacky
like I do plan to expose like feeds for tags on a post as well as a feed for the replies to said post as well as a feed for the _type_ of post (that's like 2 + n)
#
jacky
to have a reader drop a post URL (or one's homepage) would be an interesting feat in determining what to show
#
jacky
cambridgeport90: heh, it's far from ready imo
#
cambridgeport90
One of the things we need to look into is easier ways for people to implement everything they might want to on their site, and how to get people going. Somehow I think that newer protocols are going to take precedence here soon. I see Gemini being one, and ActivityPub being used more. even people who know nothing about computers can understand the
#
cambridgeport90
AP concept whenever I explain it.
#
jacky
!tell cambridgeport90 I agree to a degree. Things like Gemini are a bit akin to a 'reset' of the Web so that might be harder
#
Loqi
Ok, I'll tell them that when I see them next
#
jacky
but yeah things like Fed Bridgy can help with the AP space of things
#
@jkup
I added Webmentions to my site today. Reply to this tweet and your replies should show up on my site! https://jonkuperman.com/webmention-playground/
(twitter.com/_/status/1360786533280800775)
#
@jkup
↩️ Is your blog working with webmentions? Or the Twitter search API?
(twitter.com/_/status/1360793389826990084)
gbmor joined the channel
#
@gramsco_
↩️ It looks nice anyway! I tried to look at docs about webmentions but I got a headache so another time ._.
(twitter.com/_/status/1360796677062463488)
KartikPrabhu joined the channel
#
@zachleat
↩️ I’ll be honest I read this as watermelons at first and I do really like webmentions but I was still a little disappointed
(twitter.com/_/status/1360799873667395586)
KartikPrabhu joined the channel
#
jacky
So random: let's say a client provides a `me` address to my site and it doesn't match any of the ones I expect
#
jacky
should I make that request fail?
#
jacky
right now I think I will, since it's a single-user site
#
aaronpk
It's a hint, so you could also just ignore it
#
aaronpk
the trick is that you might enter something in a client that doesn't exactly match your full URL
#
jacky
hm wouldn't canonalization catch that?
#
jacky
like if I do `v2.jacky.wtf`, I'd recognize that
#
jacky
as well as just `http://v2.jacky.wtf`, I'd validate that too
#
jacky
I might be thinking too hard about this
#
jacky
is rewriting parts of his authorization endpoint to satisfy the conditional use of code_challenge bits
#
jacky
_actually_ I can see why leaving it as a hint is fine
#
jacky
like even if it's wrong
#
jacky
the client will get back the truth
#
jacky
I can see this being handy for something like micro.blog actually
#
jacky
doing a `jacky.micro.blog` versus `https://micro.blog/@jacky`
#
jacky
if someone put in the former, then the latter can be accurately returned if micro.blog prefers to display the former
#
jacky
also aaronpk, is there a way to contribute to https://indieauth.rocks?
#
jacky
definitely wanna hit it with my site in the future
#
jacky
actually nvm lol
#
jacky
ah but nothing's there
#
jacky
what would be needed?
#
jacky
also doesn't expect an answer here - if this inspires tickets; he'd happily read those
#
jacky
lol re: question
#
jacky
> If the logged-in user doesn't match the provided me parameter by the client, the authorization endpoint MAY either ignore the me parameter completely or display an error, at the authorization endpoint's discretion.
#
jacky
basically _do whatever you want_
#
jacky
tbh to be a bit safe
#
jacky
I might consider restricting some scopes if someone does that on my site
nertzy joined the channel
#
jacky
wow okay this is kinda fun lol
#
jacky
like with this, I could even add an internal dashboard to see what scopes each app has been given
#
jacky
i even squeezed in a bit of tracing to see where I'm making requests from
#
jacky
might use that to alert if I'm doing it in a place that my site doesn't expect it to come from
samwilson, dhanesh, nickodd, [Grant_Cameron], silo, blade82 and petermolnar joined the channel
#
jamietanna[m]
jacky I've got my IndieAuth server using the provided `me` to autocomplete which profile to authenticate as, but I can go and change that if I want - as Aaron says, I use it as a hint
#
jamietanna[m]
jacky re IndieAuth rocks I'm also interested, when writing my IndieAuth server I'd looked at some of the scenarios/edge cases, so I'd also be interested, but feel like we need a framework like micropub rocks in place
#
jamietanna[m]
(I'm also up for helping to make it, but given my PHP knowledge is lacking I'd prefer to have something in place I can contribute to more easily)
#
jamietanna[m]
with guardrails etc
#
jamietanna[m]
I'm definitely gonna have some "audit" info to allow me to keep an eye on what's been allowed, and to see if there's anything suspicious being granted that J don't remember
samwilson joined the channel
jjuran joined the channel; ShadowKyogre left the channel
ShadowKyogre joined the channel; ShadowKyogre left the channel
#
@jkup
↩️ Yeah it's a bit messy. I'm using https://brid.gy/ to poll Twitter and turn replies into webmentions and then https://webmention.io/ to receive them
(twitter.com/_/status/1360970469483180034)
#
@jkup
↩️ Yeah it's a bit messy. I'm using https://brid.gy/ to poll Twitter and turn replies into webmentions and then https://webmention.io/ to receive them
(twitter.com/_/status/1360970469483180034)
#
@CharlieCroom
↩️ Dumb question. Twitter doesn't support sending webmentions right? So is there some service that shims that part?
(twitter.com/_/status/1360970163928109060)
ShadowKyogre joined the channel
#
@jkup
↩️ yeah I actually have a URL question that I need to learn about today! But as for malicious links yeah it could be an issue.. http://webmentions.io does give me an easy dashboard to delete and ban though!
(twitter.com/_/status/1360972402365243395)
silo joined the channel; ShadowKyogre left the channel
#
@Sean_Waite
↩️ Hugo on Azure or netlify with content pulled from codeberg and then federated with activitypub and/or webmention? Hm. Turtles all the way down, need to do more reading
(twitter.com/_/status/1360981944671096838)
[Grant_Cameron], [kimberlyhirsh] and ShadowKyogre joined the channel; ShadowKyogre left the channel
#
jeremycherfas
Can anyone tell me how to find where modern::perl is on my machine? I get a new to me error of `Can't locate Modern/Perl.pm in @INC ` which seems to have arisen after installing Anaconda for a Python course I am doing. Anaconda has definitely messed with $PATH, but I don't know where modern::perl was before.
ShadowKyogre left the channel
#
jeremycherfas
Ahah. I found it, in `user/perl5/lib/perl5/Modern` Is there a way to tell the program -- there's only one currently that needs it -- to find it there, without editing $PATH?
ShadowKyogre and [KevinMarks] joined the channel; ShadowKyogre left the channel
#
@bsookman
XML Schema found to be a literary work and copyright infringed in UK Software Solutions Ltd & Ors v 365 Health and Wellbeing Ltd & Anor [2021] EWHC 237 (IPEC) http://www.bailii.org/ew/cases/EWHC/IPEC/2021/237.html
(twitter.com/_/status/1360221251164250120)
#
[KevinMarks]
Not sure where this should be linked from https://blog.alexgleason.me/run-your-own-server/
#
petermolnar
"Cloud hosting has made it easy for anyone to run a website." Did it? Is it actually easier to run a cloud "powered" website than it was to upload a bunch of files to an ftp server?
shoesNsocks1, ShadowKyogre, silo and [snarfed] joined the channel; ShadowKyogre left the channel
#
[snarfed]
sounds like he’s comparing to commercial or otherwise larger web sites on a physical server, eg 20y ago, not FTP on consumer grade shared hosting like dreamhost etc
shoesNsocks1, [tw2113_Slack_], ShadowKyogre, silo and samwilson joined the channel; nickodd and ShadowKyogre left the channel
#
[KevinMarks]
hm. downloaded my twitter archive, but they've changed the format and included all kinds of PII and DMs so I can't just bung it on my server like I used to
[tw2113_Slack_] joined the channel
#
[tw2113_Slack_]
what a bunch of jerks
[chrisaldrich] joined the channel
#
[KevinMarks]
also includes 9.6M of twitter emoji (ie all of them, not stripped for the ones I actually use)
Seich joined the channel
#
samwilson
I'm finally getting around to importing all my Flickr photos into my website – where their times will be in UTC and so ordered correctly at last
#
vilhalmer
the twitter archive was in a different format every time I downloaded it
#
Loqi
samwilson has 1 karma over the last year
#
aaronpk
samwilson++ nice! I still have so many photos in Flickr that I need to deal with
#
aaronpk
It's a good feeling to get those imported
ShadowKyogre joined the channel; ShadowKyogre left the channel
#
samwilson
yes, it feels great to have everything in one place. the import process is going rather slowly though! about 7 seconds per photo :( and I've got 23000 to go
#
aaronpk
oh wow haha
#
aaronpk
i think the rest of my foursquare checkins is next on my list
#
samwilson
I'm adding syndication link back to them on Flickr, but I guess at some point I'll stop paying for Flickr and they'll all go away
#
samwilson
yeah I think I'll do Twitter after this, although that sounds more painful than the flickr API
#
GWG
I need to continue my import process for photos, which is one by one
#
samwilson
that's how I was doing it, but then I decided to spend 5 hours a week for the rest of my life tinkering with bespoke photo album software :-P
#
aaronpk
you can do it from the twitter export you can download instead of dealing with their api
#
aaronpk
then you just read JSON files from disk
#
samwilson
I briefly looked at that, but I want to get tweets that I replied to as well (for context)
#
samwilson
maybe that's a second pass though