#indieweb 2017-10-29

2017-10-29 UTC
tantek, Ruxton, hs0ucy and [eddie] joined the channel
#
[eddie]
!tell aaronpk: question, what caused you to post podcasts on their own site but not on yours?
#
Loqi
Ok, I'll tell them that when I see them next
snarfed joined the channel
#
[eddie]
So I have a philosophical question for anyone interested: When you have multiple sites, do you post as "original" on both, do you post it on one and do a "repost" on another site? A couple practical examples:
#
[eddie]
1. I would like a microsite for my podcast, but I also want the content on my identity website. So should I post on both? Or do I post on the microsite and then repost on my site? Feels weird to not have it be original on my site, but reposting on the site FOR the podcast also feels strange
#
[eddie]
2. I am working on a website for my family. Something that could contain photo albums, travel logs, etc for my wife and I. Some of the content might come from my site, some might come from my wife's site.
#
[eddie]
For example 2 I feel good with a repost on the family site because it is displaying something for a different audience from my original website
#
[eddie]
But I don't feel super confident in that. I've tried searching the wiki for brainstorming but couldn't figure out what to really search and didn't really find anything
#
snarfed
!tell j4y_funabashi hmm, sorry for the trouble! not sure what's up, it worked for my account just now. facebook API seems slow right now though, make sure it fully loaded?
#
Loqi
Ok, I'll tell them that when I see them next
#
Loqi
snarfed: [chrisaldrich] left you a message 3 days, 17 hours ago: For fed.brid.gy & WordPress, if one's feed isn't on their homepage, but somewhere like example.com/blog/, then one should use example.com/blog/ instead of just the domain for creating the <head> atom feed w/ granary?
#
Loqi
snarfed: j4y_funabashi left you a message 6 hours, 49 minutes ago: I am looking at the demo app for granary, trying to look at my facebook feed but pressing GET returns no results, do I have to fill in activity_id?
#
tantek
[eddie] I think there are different ways of cross-posting and perhaps it would be better to gather / document examples first
#
snarfed
!tell j4y_funabashi (and no, no activity_id required.) if it still doesn't work, try revoking granary's access on https://www.facebook.com/settings?tab=applications and then r login on granary
#
Loqi
Ok, I'll tell them that when I see them next
#
tantek
for example, sometimes cross-posts are just syndicated copies, not reposts
#
tantek
that is, the original author puts the post in both places. that's more syndication than reposting. reposting typically has the implication of someone else reposting something
#
tantek
(though Twitter I think enabled retweeting yourself?)
#
tantek
[eddie] here's an example - see top (most recent) posts on microformats.org
#
tantek
from different authors, cross-posted from their personal sites. not "reposted"
wolftune and EmreSokullu joined the channel
#
[eddie]
tantek: ohhh that’s interesting. That makes sense that syndication would be different than reposting.
#
[eddie]
I think one of the challenges in researching this one is that it doesn’t feel common inside of Silos and I’m not sure what to research besides them. Lol
snarfed joined the channel
#
tantek
group blogs
#
snarfed
!tell chrisaldrich re bridgy fed + url paths + atom...i honestly don't know. :P bridgy fed currently federates responses, but not (really) accounts or posts. obviously a big whole in the broader UX of making indieweb interact w/fedsocnets!
#
Loqi
Ok, I'll tell them that when I see them next
#
snarfed
!tell chrisaldrich i'm all for bridgy fed helping with that part too, but i'm hoping the community can help with research to see what works. want to help? https://github.com/snarfed/bridgy-fed/issues/14
#
Loqi
Ok, I'll tell them that when I see them next
#
Loqi
[snarfed] #14 docs for following and delivering posts, both directions
#
tantek
snarfed, it's one way right now right? Responses but not backfeed?
#
snarfed
tantek: two way!
#
[eddie]
tantek: ahh good point! Because that’s essentially what my family site would become.
#
snarfed
definitely still finding inconsistent edge cases, but fully implemented and tested at least the most obvious ways. https://fed.brid.gy/#sites
EmreSokullu, tbbrown, renem, wolftune and eli_oat joined the channel
#
Loqi
[superfeedr] "The ongoing saga of attempting to connect micro.blog to Known" by Jeremy Cherfas on 2017-10-26 https://stream.jeremycherfas.net/2017/the-ongoing-saga-of-attempting-to-connect-microblog-to-known
eli_oat, jjuran and tantek joined the channel
#
tantek
is there a common name or phrase for the problem the Oatmeal recently documented? https://twitter.com/oatmeal/status/923250055540219904
#
tantek
the cartoon is captioned "Reaching people on the internet."
#
tantek
but really the problem is one where the silo doesn't show you things from (at least some) pubilshers without making the publishers pay
#
tantek
pay for reach as it were
#
tantek
and a quick web search reveals that there's enough use of that phrase that it should be known
#
tantek
what is pay for reach?
#
Loqi
It looks like we don't have a page for "pay for reach" yet. Would you like to create it?
#
tantek
pay for reach is a business practice by [[Facebook]], [[Twitter]], and other social media [[silos]], whereby they charge publishers (ask them to pay) to have their posts be shown to users on their site, even if the users have already opted-in ([[followed]], subscribed) to posts from those publishers.
tantek joined the channel
#
tantek
what is the facebook algorithm
#
Loqi
It looks like we don't have a page for "facebook algorithm" yet. Would you like to create it?
#
tantek
Facebook Algorithm is [[Facebook Master Algorithm]]
#
tantek
algorithm << Facebook Master Algorithm
#
Loqi
ok, I added "[[Facebook Master Algorithm]]" to the "See Also" section of /algorithm
wolftune joined the channel
#
tantek
pay for reach << Facebook Master Algorithm
#
Loqi
ok, I added "[[Facebook Master Algorithm]]" to the "See Also" section of /pay_for_reach
#
tantek
pay for reach << antipatterns
#
Loqi
ok, I added "[[antipatterns]]" to the "See Also" section of /pay_for_reach
#
tantek
AMP << 2017-10-26 Ethan Marcotte: [https://ethanmarcotte.com/wrote/seven-into-seven/ Seven into seven.] <blockquote>Significant features and changes require the approval of AMP’s Technical Lead and one Core Committer, that list seems exclusively staffed and led by Google employees</blockquote>
#
Loqi
ok, I added "2017-10-26 Ethan Marcotte: [https://ethanmarcotte.com/wrote/seven-into-seven/ Seven into seven.] <blockquote>Significant features and changes require the approval of AMP’s Technical Lead and one Core Committer, that list seems exclusively staffed and led by Google employees</blockquote>" to the "See Also" section of /Accelerated_Mobile_Pages
EmreSokullu, friedcell and oodani joined the channel
#
oodani
Hi, I just implemented distributed IndieAuth on 00dani.me, but IndieAuth.com doesn't seem to be able to handle it? After approving the auth request on my site and redirecting back to IndieAuth.com, I get a message saying "Something went horribly wrong! I'm sorry, there's not much other information available. You should probably file an issue: https://github.com/aaronpk/IndieAuth.com/issues." I've no idea what could've gone wrong - I haven't filed an issue
#
oodani
yet since there's basically no information to put into it. As far as I can tell I've implemented all the endpoints correctly, and they played nicely when testing in development with HTTPie. Any ideas? Is there an IndieAuth implementation that'll give me a little more debug info, perhaps?
#
Loqi
aaronpk: [eddie] left you a message 6 hours, 20 minutes ago: question, what caused you to post podcasts on their own site but not on yours?
#
aaronpk
try signing in to telegraph.p3k.io
#
KartikPrabhu
oodani: welcome. hopefully aaronpk would be able to resolve your issue
#
oodani
Ooh, Telegraph's worth a try. (I already tried Quill, but it doesn't let me auth since I haven't got micropub done yet. :P )
#
aaronpk
i can check the server logs on IndieAuth.com tomorrow but I'm not at my computer right now
#
oodani
My implementation rejects authentication from Telegraph, fails because it doesn't have a <link rel="redirect_uri"> set up like IndieAuth.com does. Oops.
#
oodani
I guess checking for <link rel="redirect_uri"> isn't really a good idea yet, since only some clients actually use it?
#
aaronpk
Ooh nice. Yea unfortunately that's kind of an experimental property right now
#
oodani
How do most implementations verify the client_id then? :o
#
oodani
Just switched off my redirect_uri verification and authenticating with Telegraph works perfectly. Hmmm.
#
KartikPrabhu
aaronpk: it is kind of weird that clients use experimental properties
#
KartikPrabhu
which then trips up people testing it
#
oodani
I can't authenticate with woodwind.xyz, because it thinks my authorisation endpoint is at woodwind.xyz/auth/indie instead of 00dani.me/auth/indie. That seems like a bug in Woodwind, honestly.
#
aaronpk
KartikPrabhu: it's more that the guide for clients is stricter than most clients are actually implemented
#
aaronpk
oodani: that does sound like a bug in woodwind. Are you using a relative url for your endpoint?
#
oodani
Yep, since that's what Django produces by default. <link rel="authorization_endpoint" href="/auth/indie" />, specifically. I include a <base> everywhere, so relative URLs should be resolvable without any trouble.
#
KartikPrabhu
aaronpk: hmm yeah that sounds like a problem for people trying to test their implementation (such as this case). is there a indieauth.rocks?
#
KartikPrabhu
oodani: yeah that does seem to be a bug with woodwind
#
aaronpk
Oof I don't know if anything involving <base> resolving would be considered not any trouble
#
aaronpk
but it should at least be resolving using the url it retrieved from
#
KartikPrabhu
shouldn't most "parsers" repsect the <base> tag?
#
oodani
My <base> is just <base href="https://00dani.me" />, it's there so that you can properly parse relative URLs in local copies of my pages.
#
KartikPrabhu
oodani: yup that seems correct to me
#
oodani
So even clients that don't know about <base> should be okay here, since they go to the live site anyway.
#
aaronpk
It's late, I've gotta go to bed
#
oodani
Timezones!
#
oodani
Thanks for your help so far. :)
#
KartikPrabhu
oodani: thanks for reporting the weirdness!
[jeremycherfas] and EmreSokullu joined the channel
#
Zegnat
Good morning IndieWeb
jeremycherfas joined the channel
#
jeremycherfas
Good morning IndieWeb
#
Zegnat
How has your known testing gone, jeremycherfas? All peachy?
#
jeremycherfas
Timezones on chat.indieweb.org do not know about the end to daylight savings in Europe.
#
jeremycherfas
Very peachy indeed. I just wish there were a flag for whether or not to POSSE. Not everything deserves to be everywhere.
#
jeremycherfas
Had the first opportunity to use my automated bad comment delete script in the wild this morning and it failed. Because in all my testing, I have it the full filename, with extension. But the email that alerts me that there is a comment does not include the extension. Simple enough fix, but now need to wait for another opportunity to use it.
#
jeremycherfas
Zegnat, do you know how the bridge from Slack to IRC handles replies in a thread at Slack? It seems like they are ignored.
#
Zegnat
I do not know how they are handled, I think I’ve never even logged into Slack
#
Zegnat
Also, re flags for POSSE, isn’t that what the syndicate-to property in Micropub is for?
#
jeremycherfas
No prob. I'll find my reply there and repost here. I seem to recall a problem like this before.
#
jeremycherfas
Yes it is; but the hook up between micro.blog apps and Known seems to be hard-wired. Everything from one of the apps goes to Stream.
nitot joined the channel
#
jeremycherfas
Ah; chat isn't logged, of course. I replied to your post about loudness, zegnat, as follows: Fascinating, especially the logging part. I have a dB meter on my phone for audio reasons but I’ve never thought to log.
#
Zegnat
Oh, yeah, that showed up in IRC too
#
Zegnat
But if you wrote it in Slack as a reply to a specific message, that information gets lost: https://imgur.com/GR8fjMw
#
jeremycherfas
OK. I use Slack to catch up from when I quit IRC, so if there are things that happened a while ago, I make them replies in a thread. That's obviously not such a good idea.
#
oodani
Update from earlier: I ended up keeping the rel="redirect_uri" verification as an optional extra rather than mandatory-for-all-IndieAuth-clients. https://00dani.me/articles/4/lemoncurry-1-5-0-now-with-indieauth Pretty proud of it honestly.
#
Loqi
[Danielle McLean] lemoncurry 1.5.0: now with indieauth!
#
@kevinmarks
“The web is and will always be the most popular mobile operating system in the world—not iOS or Android.” https://medium.com/swlh/browsers-not-apps-are-the-future-of-mobile-c552752ff75 #indieweb
(twitter.com/_/status/924564071097163776)
#
Loqi
[superfeedr] "“The web is and will always be the most popular mobile operating system in the world—not iOS or Android.” https://medium.com/swlh/browsers-not-apps-are-the-future-of-mobile-c552752ff75 #indieweb" by Kevin Marks on 2017-10-29 http://known.kevinmarks.com/2017/the-web-is-and-will-always-be-the-most-popular
#
Zegnat
oodani, it is not so much an experimental feature, as it is something that doesn’t exist in the spec apart from a small FAQ-like section: https://indieweb.org/authorization-endpoint#Redirect_URI_verification
#
Zegnat
Wouldn’t surprise me if most people haven’t yet realised registering redirect URIs is a thing
#
oodani
Yeah, that's where I got the info about using it. :P Aaron described it as experimental, but I guess if it's not actually in the spec yet then that's fair.
#
Zegnat
Ooh, you have a u-pronoun! Time to document!
loicm joined the channel
#
oodani
I got the idea from your h-card, actually. :3
#
Zegnat
Darn, I was just typing up to ask where/when you found it. And was hoping you may have come up with it yourself ;)
[kevinmarks] joined the channel
#
oodani
I think I found it while searching around for alternatives to the unbelievably terrible p-sex/p-gender-identity combo?
EmreSokullu joined the channel
#
Zegnat
As far as I know, I am the first to introduce u-pronoun and give arguments as to why the old p-pronoun-{case} is undesirable.
#
Zegnat
The brainstorm is here (where I will add your site as an example): http://microformats.org/wiki/h-card-brainstorming#Pronouns
#
Loqi
h-card Brainstorming
#
oodani
Yep, I read your argument as to why p-pronoun-{case} doesn't work and I totally agree. u-pronoun linked to basic examples of usage is so much more usable.
#
Zegnat
Documented below my example code :) http://microformats.org/wiki/h-card-brainstorming#Pronouns Feel free to edit if you think I’ve presented your use and opinions wrong.
#
Loqi
h-card Brainstorming
#
oodani
Looks good. :)
#
Zegnat
Great! It is a good prompt to get a link to Pronoun Island on there. It is relatively popular on Twitter and Tumblr in my experience. (Though utterly English-centric once again.)
#
oodani
I like it a lot, although it really ought to support more languages.
#
Zegnat
What is Pronoun Island?
#
Loqi
It looks like we don't have a page for "Pronoun Island" yet. Would you like to create it?
#
Zegnat
Pronoun Island is a website listing personal pronouns with usage examples.
catsup joined the channel
#
Zegnat
Pronoun Island << [https://github.com/witch-house/pronoun.is/issues/66 Internationalisation] has been a long-standing issue.
#
Loqi
ok, I added "[https://github.com/witch-house/pronoun.is/issues/66 Internationalisation] has been a long-standing issue." to the "See Also" section of /Pronoun_Island
#
Loqi
[garbados] #66 Internationalization?
#
Zegnat
Pronoun Island << IndieWeb example: [https://00dani.me/ 00dani.me]
#
Loqi
ok, I added "IndieWeb example: [https://00dani.me/ 00dani.me]" to the "See Also" section of /Pronoun_Island
#
Zegnat
That’s enough documenting for now, I think
#
oodani
I didn't realise Loqi could do this stuff. That's pretty cool.
#
Zegnat
Hopefully enough documenting to sneak a mention of pronouns into the weekly podcast (hint-hint schmarty)
#
jeremycherfas
My main pronoun beef is silly people who insist that singular they is not allowed.
#
oodani
Same, those people are obnoxious.
#
jeremycherfas
Most prescriptivist language people are. Almost by definition.
#
oodani
Pretty much.
#
Zegnat
Singular they has snucked into many dictionaries though, its usage goes a long way back. So I see no reason for prescriptivists to ignore that.
#
oodani
Oddly enough, schema.org handles gender better than standard microformats2. It's… probably the only thing schema.org has ever done well?
#
oodani
Instead of having a "biological sex" and a "gender identity" there's just one gender field, which can be schema.org/Male, schema.org/Female, or freeform text if neither of those work. And that's really good, actually. Which is weird because schema.org sucks.
#
Zegnat
The original hCard was basically just vcard codified to HTML. Any limitations come from the vcard spec. Theoretically, everything is freeform text in mf2.
#
Zegnat
Also note that the vCard reference spec allows for freeform gender: https://tools.ietf.org/html/rfc6350#section-6.2.7
#
Zegnat
So that pretty much matches what schema has exactly. It might just be badly described in the h-card spec
#
oodani
In the h-card spec it's two separate fields, p-sex and p-gender-identity, since the gender field in vCard has two components.
#
oodani
Dropping p-sex and changing the field to *just* be the freeform text part would be ideal, honestly.
#
[kevinmarks]
Well, the fields are optional. So use p- gender-identity
#
Zegnat
Yep, only use p-sex if you for some odd reason must classify your sex with a single letter.
#
Zegnat
Honestly, I only have it on my site because I am a completionist.
#
oodani
I'm not really a fan of the *name* p-gender-identity either. It could just be p-gender.
#
Zegnat
That I agree with
#
[kevinmarks]
The problem with schema's approach (inherited from opensocial, so mea culpa) is that having free text with 2 predefined values encourages engineers to neaten things up by just using the predefined ones
#
[kevinmarks]
Especially those who think in terms of enums
#
oodani
Oh damn. Good point. :(
#
Zegnat
Then again, I also use the class name “p-identity-manness”, so I can’t comment on what names are good and not ;)
#
Zegnat
Honestly, it should only be freeform. That’s what e.g. MetaFilter does
#
[kevinmarks]
When we were adding this to vcard a few years ago, gender identity was a more common term
#
oodani
Ah. Not so much now.
#
[kevinmarks]
Which is a sign of progress, I think.
#
Zegnat
I think a shift to simplifying all those terms is relatively recent though. Much how I am seeing more and more people abolish “preferred pronoun”, as it should just be “pronoun”. (Not a preference, but a statement of fact.)
#
oodani
Yeah, I'm very much in favour of that change.
#
Zegnat
Same. Although as someone who never gets misgendered anyway, my stake in this is incredibly low.
#
oodani
I'm kinda leaning towards using a field "u-gender", akin to u-pronoun? Since that enables linking to a gender page with more info.
#
[kevinmarks]
Sarah Dopp's post to the vcard mailing list really helped shape this https://www.ietf.org/mail-archive/web/vcarddav/current/msg01778.html
#
[kevinmarks]
And if you read some of the replies you can see how we got the current compromise
#
Zegnat
I’ll have to make some time to dive into that, thanks [kevinmarks]!
#
oodani
"Sex is straightforward" is an astonishing way to respond to a post pointing out that intersex people exist, honestly.
#
Zegnat
oodani, I think I prefer a textual representation for gender, so not sure how I feel about u- in that context.
#
[kevinmarks]
Adopting some 1970s medical record standard with numeric sex was the original proposal, before we got involved. And this is only 7 years ago
#
oodani
Jeepers.
#
oodani
Zegnat, hmm, a textual representation of gender is good, but also, you *have* already got a gender page on your site. :P
#
Zegnat
Yes, that one has a p-name though, so it all parses into a single textual representation too ;)
#
Zegnat
And really, my gender page is one big experiment, haha
#
Loqi
nice
#
oodani
Fair enough. :3
#
Zegnat
I’ve seen so little adoption of Sam Killermann’s Genderbread Person, it saddens me a little. Because it does make people think whenever I show my gender page.
#
oodani
Honestly I'm a little uncomfortable with it, mostly because "biological sex" is one of the axes. I really think the whole idea of "biological sex" is something we should be moving away from.
#
Zegnat
Yeah, it is a bit of a weird one. I think it is meant to convey physical characteristics. And as part of the model it can make for a good contrast to expression.
#
[kevinmarks]
Do we need a new term for what twitter does? "marketing gender"?
#
[kevinmarks]
Where they assign you a gender based on what they want to sell you things?
#
oodani
"assigned gender at tweet"?
#
Zegnat
Do you have an article on them doing that, [kevinmarks]?
#
Zegnat
Also, oodani, you might like this one: http://wiki.zegnat.net/sexes - my go to table for whenever someone tells me biological sex is binary.
#
Loqi
[Sam Killermann] A Guide to Gender
#
Zegnat
That’s not at all the title of the page, but, thanks Loqi?
#
oodani
Oh wow, I definitely like that table.
#
Loqi
Liam Adelson
#
Loqi
Liam Adelson
#
[kevinmarks]
Maybe "targeted gender" is the term
#
Loqi
gender examples
nitot joined the channel
#
Zegnat
Thanks for the links [kevinmarks]
#
[kevinmarks]
They now let you see and edit your assigned gender
#
Zegnat
Now I remember why I do not like editing the microformats wiki: recaptcha.
#
Zegnat
Also, apparently there is an IndieWeb tool to check your Twitter network’s gender: https://www.proporti.onl/
#
@martijnvdven
@jkphl are there any plans towards a Leaders Summit in Berlin yet? 5 people have stated an interest on-site. https://indieweb.org/Planning#Planned
(twitter.com/_/status/924592373245599744)
DanC, tbrb, eli_oat, leg, loicm, nitot, hs0ucy, EmreSokullu, wolftune and snarfed joined the channel
#
sknebel
oodani: re woodwind login, bug seems to be here: https://github.com/kylewm/flask-micropub/issues/5
#
Loqi
[sknebel] #5 Handle relative links to endpoints
#
oodani
Yep, looks like that's the spot. :o
loicm, nitot, snarfed, raretrack and botka joined the channel
#
raretrack
trying to set up an indiewebified Wordpress instance to use in parallel with my Known instance
tantek joined the channel
#
GWG
raretrack, need help?
hs0ucy joined the channel
#
raretrack
i'm just going through the indiewebify.me tests at the moment - webmentions test doesn't seem to work at the moment
#
tantek
raretrack, might get better help on debugging webmentions in #indieweb-dev
#
raretrack
thanks tantek, i'll try there if I stay stuck
#
aaronpk
why do i always struggle figuring out how to display webmentions
#
tantek
aaronpk because webmentions are plumbing, and trying to figure out how to display plumbing is not only really hard, it's the wrong question!
#
aaronpk
i meant how to display comments
#
tantek
also sounds like #indieweb-dev :)
#
aaronpk
nah purely talking about the visual aspect of it
#
tantek
that's still talking about creating the design
#
tantek
you're talking about how do you create the design, rather than how do you use it
#
tantek
creating content -> #indieweb, creating design/UX/code -> #indieweb-dev is kinda how I was splitting it
botka and snarfed joined the channel
#
raretrack
so think I've got webmentions working now - my WP test post https://testblog.raretrack.uk/test/test-post/ picked up my reply from my Known instance at https://raretrack.uk/2017/reply-test
#
Loqi
[admin] Test post
botka, loicm, jeremycherfas, eli_oat, nitot, tantek and raretrack joined the channel
nitot, [miklb], loicm, friedcell, wolftune and sebsel joined the channel
#
@schmarty
Your &lt; 10min update on the #IndieWeb community! This Week in the IndieWeb audio edition for Oct 21st - 27th, 2017. https://martymcgui.re/2017/10/29/163907/
(twitter.com/_/status/924738083538591744)
#
@huffduffer
This Week in the IndieWeb Audio Edition • October 21st - 27th, 2017 https://huffduffer.com/schmarty/438960
(twitter.com/_/status/924738929143828482)
#
schmarty
your interview is in this one, tantek :}
[keithjgrant], friedcell, bengo, snarfed and nitot joined the channel
#
oodani
How should shortpermalink URLs be marked up in an h-entry? It looks like using <link rev="canonical"> is a fairly standard way to indicate them on a page-wide basis, but that doesn't work on a feed page.
#
oodani
An extra u-url with no u-uid? That doesn't indicate that the URL's the short one, though.
#
oodani
Perhaps the URL's physical length is a better way to indicate that. :P
snarfed and tantek joined the channel
#
tantek
oodani: rev=canonical is deprecated and discouraged: http://microformats.org/wiki/rel-canonical#rev-canonical - for more markup discussion, let's follow-up in #indieweb-dev
#
Loqi
rel-canonical
#
oodani
Fair enough.
wolftune and snarfed joined the channel
#
Loqi
rel="shortlink"
#
snarfed
rel-shortlink++
#
Loqi
rel-shortlink has 1 karma
#
oodani
Good to know. :3