2015-08-20 UTC
colintedford1 joined the channel
# 00:02 colintedford1 snarfed: brid.gy presents: When APIs Attack!
# 00:02 colintedford1 snarfed: brid.gy presents: When APIs Attack!
# 00:05 colintedford1 Is there an echo in here?
# 00:05 colintedford1 Is there an echo in here?
# 00:09 colintedford1 My utterances from /today?beta client seem to have one, so I'm gonna hush.
# 00:09 colintedford1 My utterances from /today?beta client seem to have one, so I'm gonna hush.
mdik joined the channel
mdik_, mdik and colintedford1 joined the channel
# 00:33 colintedford1 KevinMarks: I did have a different log page open in another tab
# 00:33 colintedford1 KevinMarks: I did have a different log page open in another tab
# 00:33 colintedford1 But apparently closing it didn't help
# 00:33 colintedford1 But apparently closing it didn't help
# 00:34 colintedford1 ...or maybe it did
# 00:34 colintedford1 ...or maybe it did
# 00:35 colintedford1 *relurks*
# 00:35 colintedford1 *relurks*
glennjones joined the channel
# 00:37 kylewm watching someone code sounds about as exciting as watching someone fish
# 00:42 endi there'd probably be more swearing involved than fishing
# 00:43 kylewm haha I don't know, fishing is pretty frustrating
# 00:46 kylewm ha, just found a weird bug in Woodwind. when it goes to show a reply to a post on werd.io, it looks to see if it has previously fetched the post so it doesn't have to fetch it again
# 00:46 kylewm but if that post was previously fetched via RSS then it's a little bit uglier than the h-entry version
mdik_ and mlncn joined the channel
# 00:56 kylewm I regret trying to explain that, but at least it's fixed now (for new posts)
# 01:11 kylewm !tell Oliver__G are you still having problems with Quill & silo.pub?
# 01:11 Loqi Ok, I'll tell them that when I see them next
mdik, glennjones, tentonbricks, mdik_, lukebrooker, keroberos, djwesto, glennjones_ and KevinMarks_ joined the channel
# 02:28 pokr twitter is updating ''recommended'' users despite all cookies blacklisted. yalla overt fingerprinting
GWG, uwe, KevinMarks__, KevinMarks_ and mdik joined the channel
# 03:00 Loqi Ok, I'll tell him that when I see him next
# 03:02 colintedford !tell kylewm: re: "as exciting as watching someone fish": esp. if they're working with angler.js
# 03:02 Loqi Ok, I'll tell them that when I see them next
uwe, todrobbins, mdik_, mdik, gRegorLove, eschnou, endi, KevinMarks__, cweiske, snarfed, loic_m, Jihaisse, friedcell, KevinMarks_, alexhartley, tvn, djwesto and mapkyca joined the channel
# 07:45 cweiske finally managed to patch up PEAR's Mail_mime and Crypt_GPG libraries to be able to send out pgp-encrypted emails
jcbsnd joined the channel
# 07:49 cweiske i'm amazed that there are no php libs to send out pgp encrypted mails
# 07:49 Loqi cweiske meant to say: i'm amazed that there were no php libs to send out pgp encrypted mails
eschnou joined the channel
sanduhrs, KevinMarks, j12t, adactio, djwesto and LauraJ joined the channel
j12t, friedcell, KevinMarks_, uwe, mlncn, eschnou, elf-pavlik, LanceyWork and indie-visitor joined the channel
# 11:13 Loqi Welcome, indie-visitor! Set your nickname by typing /nick yourname
indie-visitor_ and frzn joined the channel
# 11:27 Jeena hm so in Edinburgh they have a HWC the week it is not in Portland/SF? It kind of would be cool if we could coordinate all of them on the same day all over the world :D
# 11:33 Zegnat Jeena: I think rhiaro wanted to do Edinburgh every week rather than every other week, at least now when starting up
Jihaisse joined the channel
eschnou and j12t joined the channel
# 11:48 rhiaro Jeena, Zegant: yeah we're weekly, but also Thursdays because Wednesday just isn't an option
adactio and MrClaw joined the channel
friedcell and cweiske joined the channel
# 12:10 indie-visitor Em... a little confused.
# 12:11 LanceyWork !tell aaronpk it looks like Loqi will only greet the first indie-visitor in chat. since everyone afterwards has to use a different username (indie-visitor_), Loqi doesn't recognize them
# 12:11 Loqi Ok, I'll tell them that when I see them next
# 12:12 indie-visitor test nick name
# 12:13 geng Just settle the domain and login problem.
# 12:14 geng But it seems that I don't need to have a sever managed by myself, just a main page will be enough?
# 12:14 LanceyWork yeah, as long as you have a webpage somewhere that you can put rel=me links on
adactio joined the channel
# 12:25 geng It seems that I totally got it wrong....
j12t and sanduhrs joined the channel
# 12:36 geng You know... the server from digital ocean, ddns, website... it seems that I should use heroku or github.io
# 12:37 geng to make it simple
eschnou, nedman, nedorito and ttepasse joined the channel
# 13:29 Loqi [mention] Jeena wrote a post that linked to an event: Homebrew
fourtonfish and scor joined the channel
j12t joined the channel
# 14:20 voxpelli Jeena: aaronpk: that was a weird looking mention there – no URL or anything?
# 14:21 voxpelli !tell colintedford The issue tracker discussion was never really documented on the wiki I think – did take some very sketchy notes for myself – I think one could do some interesting things there, especially with Brid.gy-like features
# 14:21 Loqi Ok, I'll tell them that when I see them next
eschnou joined the channel
fourtonfish_, benwerd, eschnou and indie-visitor joined the channel
# 14:40 tentonbricks Mornin', folks!
benwerd joined the channel
LauraJ, _fran, j12t and endi_ joined the channel
alexhartley joined the channel
# 15:57 tentonbricks So, slack. Super cool concept. Completely useless in my office. So here I am, in a chat room of 1. Looks pretty, at least :)
# 15:58 tentonbricks Well, yes, here in IRC.
# 15:58 tentonbricks But this isn't slack. I have a slack instance up and running just to see what the fuss is all about, but an office that has zero need for it.
gRegorLove joined the channel
# 15:59 GWG tentonbricks: We just have a slack gateway
cleverdevil and wolftune joined the channel
j12t joined the channel
# 16:10 tentonbricks KevinMarks_: Here's the deal. We use email, which works fine for what we need. I work in local government HR, so there's not a lot of collaboration that's needed between specialists. We are each our own team of 1, essentially.
# 16:10 tentonbricks I have a bad (?) habit of trying to nose myself into the tech space, when my work is pretty far removed from that world, for the most part.
# 16:12 LanceyWork i get really excited about new tech i find like slack and immediately want to set it up for everything
# 16:13 tentonbricks LanceyWork: Exactly. My biggest hurdle, I think, is the fact that I am the low man on the totem pole here, so I have next to zero possibility of influencing change or moving us out of the mid-20th century.
# 16:14 cleverdevil I'm planning on creating an `h-review` focused plugin for the Known CMS.
# 16:16 LanceyWork why not make it an h-entry and use some distinct markup within it to denote that it's a review?
uwe joined the channel
snarfed joined the channel
# 16:40 Loqi Ok, I'll tell them that when I see them next
# 16:45 rhiaro tbrb has been struggling with silly typos in his code
# 16:45 rhiaro and is deciding how to setup his site, thinking about githubpages to start
# 16:45 rhiaro KevinMarks: as in, he's outside the door right now?
# 16:46 rhiaro tbrb is on crutches, but nonetheless more mobile
stream7 joined the channel
# 16:47 rhiaro I couldn't see anyone I didn't recognise in the main foyer just now..
# 16:52 rhiaro I'm sorry KevinMarks, andrewmarks was flyering for the Festival, I'm going to have to ask him to leave
KevinMarks__ and KevinMarks_ joined the channel
fourtonfish joined the channel
j12t joined the channel
# 17:00 Zegnat URLs on a flyer? I hope they were short, hate typing those things on my mobile whenever I see posters or get flyers
# 17:04 rhiaro tbrb is discussing possibilities of multi user micropub endpoints, who has talked about this before?
KevinMarks__ joined the channel
# 17:06 rhiaro andrewmarks is distraction tbrb with IRL conversation right now (rubbish, right?!)
# 17:06 aaronpk there should be no problem with a multiuser micropub endpoint
# 17:06 aaronpk the token that is used in the micropub request should identify the user making the request
# 17:07 aaronpk so the endpoint can do whatever it wants based on which user is making the request
# 17:09 tbrb aaronpk: I guess the question then runs to what clients would support posting to a micropub endpoint which isn't defined in the user's site
# 17:09 rhiaro yeah, if the user wanted to post from a client that wasn't the site they wanted to post to, the client needs to discover the right endpoint
# 17:10 rhiaro I guess easiest is just if the site whose endpoint you're posting to is also the client
# 17:10 tbrb I guess there could be a couple of ways of doing that - either having clients allow you to override which endpoint you want to post to (just keep autodiscovery as a default maybe) or support discovering of multiple endpoints from a person's site
# 17:11 tbrb The latter would also allow you to discover what sites a person posts to as well, be that good or bad
alexhartley joined the channel
# 17:14 aaronpk yeah, I think most people were thinking they didn't want to have to link to all the places they post on their home page
# 17:14 aaronpk I think the problem is that so far nobody has had an actual need for this yet, so nobody has tried to build any of the pieces
andrewmarks, alexhart_ and stream7_ joined the channel
# 17:16 rhiaro I guess an alternative would be to be able to tell a client which site you want to publish to, and it sends there instead of to your own endpoint
alexhar__, KevinMarks_ and nedsauce joined the channel
# 17:21 camerongray Wait, rhiaro, Homebrew webstie club on today?
# 17:22 camerongray So yes I have
# 17:22 camerongray I've fucked up FreeBSD, I'll leave it to be future Cammy's problem
# 17:23 rhiaro KevinMarks: not quite - this is instead of, not as well as
# 17:31 camerongray rhiaro: Well, I'll be along in a bit
# 17:31 camerongray Waiting on servers
alexhartley joined the channel
# 17:32 aaronpk the closest use case for me needing a multiuser micropub client is posting to things like company blogs or community blogs
# 17:32 camerongray What are other people planning?
# 17:32 camerongray Leaving office now
rMdes joined the channel
# 17:48 Zegnat rhiaro: I enjoyed your LD/mf2 post today, nicely done.
yakker joined the channel
# 18:01 camerongray rhiaro: at side door, can you let me in please?
mlncn and uwe joined the channel
# 18:20 moredhel have funsies :) wish I was there
# 18:20 moredhel may do a little coding from here to feel better
JasonO joined the channel
Kongaloosh joined the channel
# 18:39 Jeena So next week we'll have a Homebrew Website Club in Gothenburg again. And I should really add Events to my website, this is getting ridiciolous.
# 18:41 voxpelli tbrb: aaronpk: I have the need for multiple micropub endpoints for my identity, using temporary hackish identities now
# 18:42 kylewm voxpelli: do they differ by client_id? or something else?
# 18:42 voxpelli (Adding a rel-me mailto: and sometimes a separate URL with that that's only meant for discovery)
JasonO joined the channel
# 18:43 voxpelli kylewm: I've multiple sites I post to, but only a single IndieAuth identity
# 18:43 aaronpk and your intent is to have the same identity across these different sites?
# 18:44 voxpelli aaronpk: that I want to post to sites without connecting them to my identity/social graph – pretty much the multiuser thing tbrb mentioned
# 18:44 aaronpk so you *don't* want to tie your main identity to these sites?
# 18:45 voxpelli But my Micropub is endpoint still. Rey much expects me to authenticate myself
# 18:46 Loqi voxpelli meant to say: But my Micropub is endpoint still very much expects me to authenticate myself
# 18:46 aaronpk if I wanted to post on example.org and not tie it to aaronparecki.com, i'd just go sign in to a micropub client as example.org
# 18:46 voxpelli And if I were to sign in to example.org as well, how would it know who I was?
# 18:48 voxpelli It's always preferable to authenticate with a URL that represents ones identity I think
# 18:48 aaronpk but you just said you *don't* want it tied to your identity
# 18:49 aaronpk can you describe the end result you are looking for? this is too much plumbing speak that is confusing things
# 18:50 aaronpk there are like 3-4 different situations that all fall under "multi-user" and they are all very different
# 18:51 voxpelli One simply provides two pieces rather than one: A URL of something one wants to edit and a URL that identifies oneself
# 18:51 Loqi voxpelli meant to say: One simply provides two pieces rather than one: A identifier of something one wants to edit and a identifier that identifies oneself
# 18:52 aaronpk e.g. "I want a blog post on companyblog.example.org that says the author is aaronparecki.com, and I want to authenticate to clients as aaronparecki.com"
# 18:52 Loqi gives aaronpk a blog post on companyblog
# 18:52 voxpelli I want to get access to my news site by identifying myself as myself
# 18:52 aaronpk vs "a blog post on companyblog.example.org that says the author is aaronpk.example.org and I don't want any assocaitoin with aaronparecki.com"
# 18:53 voxpelli Just like I get access to the IndieWeb wiki by identifying myself as myself
# 18:54 voxpelli Does that matter? It could be The Verge, Techcrunch, whatever?
# 18:54 aaronpk i mean i just don't understand what you're saying
# 18:55 aaronpk so you mean you want to post at a shared blog like techcrunch having identified as yourself?
cweiske and JasonO joined the channel
# 18:57 voxpelli aaronpk: pretty much, wasn't that what you all discussed earlier?
# 18:58 voxpelli so as one of the points there was that no-one has had the need, I pointed out that I do
# 18:59 aaronpk can you share the URLs of the sites you are posting to? or is that private?
# 18:59 voxpelli Yes, but the client support is slim so I can only do it with my own client
# 19:00 voxpelli Well, voxpelli.com is one of them and while that is technically part of my social graph, it isn't in any way that IndieAuth supports
# 19:01 aaronpk why isn't that something that indieauth supports?
# 19:02 voxpelli aaronpk: because IndieAuth requires all silos to be one step away, mine are two steps away
# 19:02 aaronpk ok, cause that has little to do with micropub clients
# 19:03 voxpelli But the same solution works for that shortcoming as well as for multiuser blogs as I then can pick another entry point into my social graph to auth with
JasonO joined the channel
# 19:05 aaronpk so do you have an examle of a multiuser blog you are trying to post to?
# 19:06 voxpelli How do you identify whether a blog is single or multi-author?
# 19:07 voxpelli Until the second author post its first post you can't
# 19:07 aaronpk whether there are posts that say they are authored by different people
# 19:07 aaronpk voxpelli.com says "Hi, I'm Pelle Wessman and this is my blog." so i'm pretty sure that is not multiuser
# 19:08 aaronpk that's different tho, since you still end up being the author of those articles
# 19:11 voxpelli Jeena: Next week you won't be alone in having HWC in Sweden it looks like ;)
# 19:12 aaronpk i'm looking for actual examples and use cases for this so that there is something more to talk about other than we think we need this thing
# 19:14 aaronpk so what you're talking about I wouldn't call a multi-user blog. you want other people to help make edits to your posts.
# 19:18 aaronpk if your site provides an interface for editing your posts, you don't need micropub at all to let others help make edits, since they can just sign in as themselves.
# 19:18 aaronpk now if you're talking about having other people use generic micropub clients to make edits to your posts, that's interesting and that involves micropub
[tantek] joined the channel
# 19:18 [tantek] aaronpk++ for keeping pushing for concrete use cases!
# 19:19 voxpelli I use my indie-identity to log into eg. the wiki, to many indie readers – to basically all things indie – except for when I want editorial access to a site, then I have to use the indie-identity of that site?
# 19:20 [tantek] Just like it's up to IWC wiki to let you edit with whatever identity you sign in with.
# 19:20 [tantek] Nothing new needed for that. Just the other site has to keep an ACL or allow any indieauth sign in etc.
# 19:20 aaronpk where it gets to be a new thing is if for example we wanted people to be able to use Quill to edit the wiki
# 19:21 aaronpk so in that case, I would want to start by talking about what is the desired user flow for people doing this
# 19:22 voxpelli if I want to use a generic Micropub client to edit an arbitrary blog by authentication as myself, then that would require additions
# 19:22 voxpelli not in the protocol itself, but in how the client does its discovery
# 19:22 voxpelli [tantek]: I've asked for that? And I've used that with my own clients?
# 19:24 [tantek] Thus for any other sites that want to allow similar editing as IWC wiki, they should implement it like IWC wiki first. Direct web ui to sign in and edit as whoever.
# 19:24 [tantek] Before worrying about solutions that require new protocols etc.
# 19:24 aaronpk voxpelli: what micropub clients do you want people to use to help you edit your posts?
# 19:26 voxpelli there are no Micropub-clients that supports edits I think so that's a bit premature, I admit
# 19:26 aaronpk that's my point, this is too early to talk about changing the protocol to support a thing that isn't even supported in single-user mode
# 19:27 aaronpk hmm just noticed poetica's "send to wordpress" feature... wonder if I could talk Blaine in to adding "publish via micropub"
R2ZER0, ben_thatmustbeme, squeakytoy2 and Kongaloosh joined the channel
# 19:30 voxpelli I'm talking about clients having additional siscovery steps, just like you, tbrb and rhiaro did before
# 19:30 Loqi voxpelli meant to say: I'm talking about clients having additional discovery steps, just like you, tbrb and rhiaro did before
# 19:31 aaronpk yeah I just think that should be driven by actual real-world use cases
# 19:31 voxpelli additional or alternate – where one can either discover additional sites to publish to or authenticate with alternative identities
# 19:32 aaronpk I mean I'm definitely interested in it, but haven't actually had a need for it myself yet
# 19:32 voxpelli well, I feel that I have use cases for it, but apparently you see it another way
# 19:33 aaronpk until you have a client that can edit, you can't have a use case for having that client support multiple users
bltavares joined the channel
# 19:33 voxpelli the edits was just a minor little side line of the discussion – the main case is for many users to be able to add content to many sites
# 19:34 voxpelli the way I have it set up right now I'm authenticated as myself in Editorial.app and I'm allowed to edit all of my sites through that one authentication
reidab joined the channel
# 19:36 voxpelli the many-to-many relationship between identities and access to sites
mattl and squeakytoy2 joined the channel
# 19:37 voxpelli we should probably drop the discussion though, getting nowhere, I will just continue down my trail on my own instead
# 19:37 aaronpk I realize this is important, I'm just trying to ground it with actual examples
jonnybarnes and uhhyeahbret joined the channel
# 19:43 Zegnat I think aaronpk and voxpelli are coming from two different angles here? voxpelli would like to solve his problem with an indieweb solution, and aaronpk wants to document a solution for others on the indieweb? Thus: documentation first versus implementation first
# 19:44 aaronpk eh, I'm mostly trying to get voxpelli to actually describe his problem in concrete terms, rather than jumping to describing the problem with a possible solution
# 19:47 voxpelli "As personal user X I want to be with a generic client be able to publish a post to larger site Y, to which I've been given permission to post"
# 19:48 Loqi voxpelli meant to say: "As personal user X I want to with a generic client be able to publish a post to larger site Y, to which I've been given permission to post"
# 19:49 aaronpk interestingly, you can do that with Quill right now, assuming site Y supports it
# 19:49 aaronpk you'd sign in as "site Y" on Quill, Quill would discover that site's authorization endpoint and direct the user to it
# 19:49 aaronpk site Y would create an authorization endpoint that then *authenticates* the user, asking them to sign in as their own identity
# 19:51 aaronpk after site Y returns the user to Quill, Quill would see that user logged in as "site Y", and when it makes the micropub request to site Y's micropub endpoint, that endpoint would be able to tell which user was actually making that request
# 19:52 voxpelli yeah, the micropub client wouldn't really see me as a personal user
# 19:52 aaronpk there might be a UX reason that it needs to, but I don't know what that is yet
# 19:53 voxpelli if "personal user Y", my colleague, uses my computer to publish something but forgets to log out – then we're in for an awkward ride – would be impossible to know as what author something were going to be published
# 19:55 voxpelli and if the micropub client were to want to have personal settings connected to their account, then that would require it as well
# 19:56 voxpelli I guess some indie-readers are kind of micropub-clients in that they can micropub-publish interactions, but also use the same authentication to save which feeds one follows
minsky joined the channel
# 20:00 kylewm would there be any value to Woodwind letting you authorize micropub with a different identity than you are signed in with?
# 20:01 Jeena voxpelli, I won't? Will there be another one?
# 20:01 kylewm right now it just uses the same user URL, but it certainly doesn't have to
# 20:02 voxpelli kylewm: perhaps rather to let one add a different site to publish to and discover the micropub-endpoint from that one instead – depends on what Woodwind publishes I guess
# 20:03 kylewm voxpelli: yeah that's what i mean, when you go into the Settings page for micropub, instead of clicking "Authorize Micropub", it would have another input for user url
# 20:03 aaronpk I was thinking it would be more useful if the reader were more like tweetdeck, where you can have multiple feeds from different accounts
# 20:07 Zegnat Jeena, I am considering if I can do 1 HWC every 2 months, but no promises.
# 20:08 kylewm there couldbe another authorization_endpoint that takes an existing user url and authorization code and re-confirms that they are who they say they are
# 20:08 voxpelli kylewm: micropub has a way to confirm a user has access
# 20:09 kylewm so you register with the site beforehand, like with the wiki -- you login and create an account and the wiki stores an auth code for you
# 20:10 Jeena Zegnat, that would be cool, perhaps we can make a fond where people can send money in for things like that
Leeky joined the channel
# 20:10 colintedford Oh wow, I run a group Wordpress site, and the flow aaronpk described makes me feel freer to use something else in the future
# 20:10 Loqi colintedford: voxpelli left you a message 5 hours, 49 minutes ago: The issue tracker discussion was never really documented on the wiki I think – did take some very sketchy notes for myself – I think one could do some interesting things there, especially with Brid.gy-like features http://indiewebcamp.com/irc/2015-08-20/line/1440080488081
# 20:11 Jeena we had that in the SELFHTML community back then, there were people who had more than enough money and there were those studenrs withoud money but wanted to attend the yearly meetups so we payed their expenses, it was quite cool
# 20:11 GWG colintedford: Did I hear WordPress?
# 20:11 GWG I really should take tantek's advice and flag for it
# 20:13 kylewm I was basically just thinking, the site would remember the relationship... "multiuserblog.com/profiles/kylewm is identified by kylewm.com", then it would delegate the authorization_endpoint bit to indieauth
# 20:13 kylewm so you'd sign into the micropub endpoint with multiuserblog.com/profiles/kylewm, but it would know that your ID is your personal domain
# 20:14 Zegnat Jeena: the fund thing might be cool, but how feasible is it for weekly/bi-weekly meets?
# 20:17 Jeena even if it only works once a month or every other month it would still be cool
# 20:19 gRegorLove GWG or other WordPress users: What theme directory sites do you recommend? All I remember is "avoid woo themes"
# 20:21 colintedford There was a study a while back showing that free themes from other places very often contained malicious code.
# 20:23 gRegorLove Yeah, I'll definitely be vetting and likely minimizing whatever theme I get.
# 20:24 GWG gRegorLove: The official repo is what I use
Lancey joined the channel
# 20:27 colintedford making a todo note. don't have cite handy & should get back to work anyway
tantek joined the channel
# 20:32 voxpelli colintedford: I would rather see "issue" being generalized into a "todo"
j12t joined the channel
# 20:36 colintedford voxpelli: OK. It sounds like maybe there needs to be a little more discussion around generalizing.
# 20:38 colintedford & then we can make a new page when/if we have some consensus about generalizing
almereyda and fourtonfish joined the channel
# 20:56 gRegorLove It's half about a Buzzfeed journalist that doesn't do *personal* email anymore
# 20:57 gRegorLove And half people's voicemails they left the show about "Email Debt Forgiveness Day" on April 30th.
wilfredh, mlncn, KevinMarks__ and KevinMarks_ joined the channel
# 21:27 tantek interesting anecdotal observation re: different forms / variations of POSSEing
# 21:28 tantek I get way more interaction on the FB POSSE copy of my post when I directly POSSE from my site via Bridgy Publish, instead of the POSSE to FB via Twitter POSSE copy.
niamu joined the channel
# 21:32 snarfed colintedford: probably often yes, since it skips the 140 char twitter middleman
# 21:35 tantek Unknown if this is because of the difference in appearance between how the FB POSSE copy is created, or if FB shows Bridgy Publish posts more often than Twitter posts.
# 21:35 tantek gRegorLove: yes they show up slightly differently, and the Bridgy Publish posts are *editable* whereas the cross-posts from Twitter to FB are *not* editable (no idea why)
# 21:39 snarfed twitter exposes analytics to users; i wonder if fb does
# 21:44 tantek one limitation (why I can't automate this with Bridgy Publish yet)
tantek joined the channel
# 21:46 tantek * Twitter -> FB preserves whitespace (e.g. linebreaks)
# 21:46 tantek * Bridgy Publish indieweb -> FB collapses whitespace (including linebreaks into normal spaces that collapse with other spaces).
# 21:46 snarfed tantek: oh no, really? that's sad, and not intentional
# 21:46 tantek does anyone else use Bridgy Publish to POSSE notes to FB?
# 21:47 snarfed didn't bridgy publish to fb preserve whitespace at some point?
# 21:47 tantek I have memories of editing past posts for this purpose too
# 21:48 tantek past FB POSSE copies created by Bridgy Publish that is
# 21:48 snarfed ah, by notes you mean posts, not facebook notes (a specific different product feature of theirs)
# 21:48 snarfed yeah. it didn't originally preserve whitespace; i added it later
# 21:50 tantek ooh just got a Twitter failbot - don't see those too often
frzn joined the channel
# 21:51 kylewm snarfed: tantek: the issue is that tantek uses literal \n's for whitespace right? whereas Bridgy expects html and converts it to plaintext
# 21:52 tantek sending literal \n's to Twitter works to preserve whitespace there, and then Twitter likely sends the same literal \n's to FB which works to preserve it there
# 21:52 tantek just saying Bridgy should handle it similarly
# 21:52 snarfed oh huh, maybe? if tantek uses e.g. css white-space: pre*, bridgy definitely doesn't include a css rendering engine :P
# 21:53 snarfed oh, you give it plain text. very different use case
# 21:53 aaronpk cause twitter isn't getting your post text from your html
# 21:54 aaronpk the problem is with html you don't know if the whitespace is intentional or not
# 21:54 kylewm I mean tantek, you're giving bridgy an e-content right? how's it supposed to know your post isn't really html?
# 21:54 snarfed yeah css support is very unlikely to make into bridgy publish
# 21:55 snarfed good point tantek, i see what you mean. if the e-content has meaningful whitespace, the css doesn't matter
# 21:55 snarfed man i could have *sworn* this worked at one point
# 21:56 snarfed never mind. yeah, it did work, just probably with html only
# 21:56 snarfed kylewm: i'm guessing yes. maybe there's an option
# 21:56 snarfed ok time to move this to the issue, at least for me :P
# 21:57 tantek (hence why I filed the issue before bringing it here :) )
# 21:57 snarfed oh no, feel free to keep discussing here, i'm just going to start capturing
# 21:58 tantek kevinmarks - false. I've never had meaningless whitespace show up in my Twitter POSSE copies of my notes
# 21:58 snarfed i expect KevinMarks means other people's post, not yours
# 21:58 snarfed basically, to determine *whether* whitespace in an e-content is meaningful, i still need to parse css and handle selectors...?
# 22:00 KevinMarks o_O a note vs a not-note, if 'note' is defined as whitespace significant
# 22:01 snarfed interesting! man, now i'm curious how many different real world notes implementations actually use meaningful whitespace as per that recommendation
# 22:01 aaronpk there's more to it than that. you have to ignore whitespace and html tags when comparing name and content
# 22:02 tantek (though I'd be willing to expand that to empty h-entry names as well - if someone is willing to document real world use-cases)
# 22:02 tantek Kevinmarks - so? the point is they saw the value, and added it.
# 22:02 tantek Twitter also didn't used to auto-link @-names, URLs etc.
# 22:03 tantek you're not seriously using "ignored ... for years" as any kind of justification for anything right?
squeakytoy joined the channel
# 22:03 kylewm for the record, I really dislike the idea of codifying twitter's non-html-whitespace convention as our best practice
# 22:03 kylewm it seems like that comes from the Twitter via SMS days
# 22:03 tantek aaronpk, there's no "have to ignore whitespace and html tags when comparing" because those are already taken care of (preserving whitespace, and text only property value) by the mf2 parser
Leeky joined the channel
# 22:04 tantek KevinMarks - happy to analyze examples of what you are claiming is "odd" or "often"
# 22:05 tantek no - not interested in rehash of your arguments. provide URLs to examples that back up what you're saying.
# 22:06 tantek since I have already done so (and do so with many / most of my notes that I publish and POSSE)
# 22:06 KevinMarks if I parse an h-entry and content==name, I should set whitespace-pre when displaying it?
snarfed joined the channel
lukebrooker joined the channel
# 22:13 tantek.com edited /note (+202) "/* Indieweb whitespace thinking */ note POSSE whitespace preserved by Twitter API, link to Bridgy Publish issue" (
view diff )
# 22:13 gRegorLove My notes do not really preserve whitespace. I use an auto-<p> script that converts two new lines into a <p> and one new line into a <br />, but that's all.
# 22:14 gRegorLove I think it collapses 3+ new lines into two, so there's not extraneos <p> or <br>
# 22:14 Loqi gRegorLove meant to say: I think it collapses 3+ new lines into two, so there's not extraneous <p> or <br>
# 22:19 gRegorLove I have thought it would be cool to have more robust whitespace support in my notes, but it's not a high priority item.
j12t joined the channel
# 22:28 kylewm.com edited /note (+158) "/* IndieWeb Examples */ add me, so I can officially complain about white-space: pre-wrap as a policy" (
view diff )
almereyda joined the channel
# 22:33 tantek kylewm: I don't understand the resistance to posting notes on one's own site (instead of Twitter)
# 22:33 tantek e.g. adactio was strongly against it for a long time (years), because he didn't care about tweets or something. and then at some point he sharply flipped.
# 22:35 kylewm or the mike monteiro argument about not collecting all your fingernail clippings....
# 22:36 tantek I'm willing to bet that *everyone* that switches from Twitter posting to personal site note-posting starts writing *noticeably* better quality notes
# 22:36 KartikPrabhu if you put your fingernail clippings into Twitter then why not on your own site?
# 22:36 tantek thus the clippings argument does not apply in practice
# 22:36 gRegorLove oh oh. Is there a term for wikifying before tantek prompts? :)
j12t joined the channel
# 22:37 snarfed tantek: i wonder if the whitespace thing is specific to notes at all...?
# 22:38 tantek snarfed - like many things here, I based in on a silo "innovation"
# 22:40 snarfed but i'm mostly ignorant/agnostic of post types, so i may just not be interpreting right
# 22:42 snarfed my notes definitely have meaningless whitespace, sadly, and i expect i'm far from alone in that
# 22:43 kylewm so here's a question -- why don't notes that have meaningful whitespace and no HTML use p-content instead of e-content?
# 22:44 kylewm i'm so confused why you want to convert urls to <a> and <img> (unlike twitter) but are opposed to converting \n's to <br>'s
# 22:44 tantek kylewm: why? typical silo note implementations auto_link and sometimes auto_embed too
# 22:45 tantek twitter does auto_link URLs. I don't see the confusion
# 22:45 kylewm so the whole rationale is that twitter and facebook use white-space: pre-wrap?
# 22:46 tantek kylewm: it is one of many silo "innovations", very much related to UX
# 22:46 tantek above and beyond all the backend protocol fights that the blogosophere / federated social web folks wasted time on
# 22:46 KartikPrabhu tantek: I think kylewm is asking if url -> <a> is ok then \n -> <br/> should also be ok
# 22:47 kylewm twitter doesn't autolink urls in the API though -- they give you the plain-text note. the autolinking is purely for display
# 22:47 kylewm so if you are following their example, you should use p-content and expect consumers to autolink it
# 22:47 tantek kylewm - just like the mf2 parser output JSON gives you both the hyperlinked e-content HTML and the plain text "value"
# 22:48 tantek kylewm - nope - my auto_link function is better than theirs
# 22:48 kylewm thanks KartikPrabhu that's exactly what i was trying to articulate
# 22:49 kylewm tantek: and my autolink function is better than yours (includes new TLDs), so let me do the autolinking
# 22:50 tantek kylewm: that's my point! having both the HTML and the "value" in the mf2 JSON output lets you the consumer decide!
# 22:50 tantek (presumably in Woodwind - e.g. if it's a note, then use the "value" and auto-link
# 22:50 kylewm that's not my point though, I want your content in all its html-formatted glory, so it looks nice in woodwind
# 22:51 tantek in either case you're inserting <a> or <img> tags into the text
# 22:52 aaronpk I actually don't like inlining img tags or videos, I prefer to show them at the bottom of the post
# 22:52 aaronpk so my autoembed finds URLs it recognizes, then puts those at the bottom after hte post text
# 22:52 kylewm yeah mine show up at the bottom of the note, more like how Facebook does it
# 22:53 aaronpk okay then I don't like auto-embedding, only link preview
# 22:53 kylewm tantek: poppping back up to "<br/> substitions covered ages ago" -- ok awesome, how can I as a feed reader/Bridgy Publisher distinguish between the two cases
# 22:54 rhiaro I noticed the quality of my posts is different
# 22:55 tantek but do you *only* write to Twitter by POSSEing?
# 22:55 rhiaro Unless it's a reply I have to crank out quick on my phone
# 22:57 tantek kylewm: note that <br/> substitution is insufficient for white-space preservatin
# 22:57 Loqi tantek meant to say: kylewm: note that <br/> substitution is insufficient for white-space preservation
# 22:58 tantek so it's not really a reasonable alternative to consider
# 22:58 Loqi tantek meant to say: so it's not really a workable alternative to consider
# 22:58 kylewm why do I need to preserve more whitespace than linebreaks or one space?
# 22:59 tantek kylewm: leading white space on a line is common
# 23:00 kylewm honestly i'm really surprised you are arguing against HTML
# 23:00 kylewm html has certainly been sufficient for anything I've ever wanted to post on my site
# 23:01 tantek kylewm: same reason people like writing / storing markdown instead of HTML
# 23:01 tantek kylewm: it's a matter of once you enable the feature, you find you use it more
# 23:01 tantek once Twitter starting preserving whitespace, and I noticed, and coded support for it on my site, my notes got far more interesting in that regard
# 23:02 tantek I even went back and added line-breaks to a few older posts that were inline lists
# 23:02 kylewm I get that, I preserve whitespace in my tweets too
# 23:02 KartikPrabhu kylewm: tantek: I think the issue with this argument is that you guys are not addressing the authoring of notes. Usually notes are not authored in HTML
# 23:03 tantek KartikPrabhu: no - the link I keep pasting to /note DOES address authoring
# 23:04 tantek KartikPrabhu: that's my point. haven't really heard any new arguments that the wiki page doesn't cover.
# 23:04 tantek so far I feel like all I've been doing is responding by walking through fragment links on the wiki
# 23:04 tantek at least we have more diverse examples documented now
# 23:04 kylewm tantek: no, you said one of the fragment links on the wiki wasn't valid because it didn't cover leading whitespace
# 23:05 kylewm if there are two possible ways to format notes, that's ok, i'd just like to be able to distinguish between them
# 23:06 tantek ah - it only documents the "wouldn't handle multiple sequential space characters. " problem with the <br/> approach
# 23:06 kylewm if there's only one valid way to format notes, then i need to think about if i want to argument about it anymore
# 23:06 tantek that makes sense - and per documentation of those notes permalinks - multiple ways of formatting note whitespace do seem to exist
# 23:07 KartikPrabhu I don't see the issue with mulitple ways of formatting. What is the use-case of distinguishing them?
# 23:07 tantek I'm not sure any argument for/against "validity" from any point of purity etc. makes any sense
# 23:07 kylewm e.g., see tantek's notes with whitespace in woodwind are all collapsed
# 23:08 tantek kylewm: could a heuristic as dumb as if content has line-breaks AND content DOES NOT have <br/> then white-space:pre-wrap ?
# 23:09 kylewm KartikPrabhu: try adding white-space: pre-wrap on .content in woodwind, and you'll see the issue
# 23:09 tantek KartikPrabhu: I believe the problem is that white-space:pre-wrap may add TOO MUCH linebreaks etc. on other notes
# 23:09 tantek notes that already have <br>s in their e-content
# 23:12 gRegorLove I wondered why tantek's notes were collapsed in Woodwind. Makes sense now.
# 23:13 kylewm tantek: yeah that heuristic sounds like it would work
# 23:14 snarfed sadly i think most heuristics would fail on my notes
# 23:14 snarfed they often have many different html elements, many extra line breaks due to authoring, and other oddities
# 23:14 tantek.com edited /note (+584) "/* Note Display */ try a Displaying Whitespace section per different methods notes publish whitespace" (
view diff )
# 23:14 snarfed (since meaningful-whitespace-in-notes was news to me today)
# 23:15 tantek snarfed, I'm guessing that you still have <br>(s) in such notes?
# 23:16 snarfed so if we expand it to trigger on any html, then maybe...but i expect that's too broad for others
# 23:16 tantek not sure about blockquote - that's a good question
# 23:17 tantek snarfed: do you have any notes with <blockquote> markup?!?
# 23:17 snarfed anyway. none of this is an itch for me personally. just saying my notes flagrantly disobey that meaningful whitespace recommendation
# 23:17 snarfed tantek: sure! many. e.g. current top h-entry on my front page
# 23:17 snarfed oh god, who knows. i really avoid thinking about post types. not really interested in them
# 23:20 snarfed this just makes me even more reluctant to add much implicit-post-type logic to bridgy publish
j12t joined the channel
# 23:20 gRegorLove Not sure about thumbs up, given the FB association with "liking"
# 23:20 snarfed if i can't motivate myself to learn and grok them :/
# 23:20 tantek likely not going to be needed (much?) unless/until Bridgy Publish start supporting POSSEing to Tumblr
# 23:22 tantek snarfed - better question, do you have any posts with <blockquote> WITHOUT <p> or <br> ?
# 23:23 tantek the presence of in-reply-to, u-like-of, and p-rsvp is pretty explicit markup. not really implying anything there.
# 23:24 snarfed nah i meant the part that chooses between e-content, p-name, and p-summary
# 23:24 snarfed i think i've successfully managed to please no one with the current logic :P
# 23:24 tantek snarfed - really? the current logic seems to work ok except for the whitespace problem ;)
# 23:25 snarfed nah i get a constant trickle of people asking for changes, or for why their posts don't publish like they should
# 23:25 snarfed no matter thoughy. one silver lining of all this is, if kylewm gets interested enough to implement any of these heuristics in mf2util or similar, i'll happily blindly use them
benwerd joined the channel
# 23:29 Loqi Ok, I'll tell them that when I see them next
# 23:33 tantek colintedford: is posting issues on your own site a personal itch for you?
# 23:34 tantek the IRC discussion is nearly all theoretical handwaving
# 23:34 colintedford But I hand-POSSE to/backfeed from Github, & would love for people to own their own issues.
# 23:35 tantek what in particular do you hand-POSSE to/backfeed from Github?
benwerd joined the channel
# 23:46 tantek.com edited /issue_tracker (+1395) "challenge the mostly theoretical issue tracker discussion, while citing actual real world attempts at indie publishing of code, issues, replies to issues" (
view diff )
# 23:48 cleverdevil I'll add it to the various wikis tomorrow, I've gotta run for the day shortly.
# 23:48 tantek it looks like the h-card inside the review is just an image "https:\/\/cleverdevil.io\/gfx\/users\/default-00.png" with no name?
mdik joined the channel
# 23:50 tantek (that's some very hard-earned karma right there)
# 23:51 cleverdevil thanks, time to run, I'll update the wikis later on this evening, if I can get the kids down to bed :P
snarfed1 joined the channel
snarfed joined the channel
KevinMarks joined the channel