#indiewebcamp 2015-09-15

2015-09-15 UTC
shiflett and snarfed joined the channel
#
snarfed.org edited /original-post-discovery (+33) "/* 24hr time window */ counterexample" (view diff)
mblaney joined the channel
#
snarfed.org edited /original-post-discovery (-493) "/* Brainstorming */ drop unnecessary recap" (view diff)
#
snarfed ok i think https://indiewebcamp.com/original-post-discovery#Brainstorming is up to date with our current real world experience from working on https://github.com/snarfed/bridgy/issues/51
#
Loqi snarfed: tantek left you a message 5 hours, 28 minutes ago: is "subpost" a thing? or are you trying to make it a thing? per your usage: http://werd.io/2015/publish-on-your-own-site-reflect-inwardly#comments http://indiewebcamp.com/irc/2015-09-14/line/1442256633622
#
snarfed there are four heuristics, and all have counterexamples we've hit in the real world with bridgy, so none are ideal
#
snarfed !tell tantek re "subpost," meh. obvious expansion of subtweet. i haven't really put any thought into it.
#
Loqi Ok, I'll tell him that when I see him next
#
mblaney what is indiechat
#
Loqi #indiechat is an IRC channel on freenode.net for off-topic chat for the IndieWebCamp community https://indiewebcamp.com/indiechat
#
mblaney thanks loqi
#
Loqi you're welcome
#
snarfed so the current state of a reliable original post discovery algorithm that works across people and silos is...weak at best
#
snarfed ah well!
tantek, mblaney and snarfed joined the channel
#
snarfed (barring syndication links, of course)
#
@IPFSbot RT @_crossdiver: I love pinning @neocitiesweb sites to my local #IPFS node. The #indieweb just keeps getting cooler. (twitter.com/_/status/643583034512568320)
#
mblaney what is IPFS?
#
Loqi It looks like we don't have a page for "IPFS" yet. Would you like to create it? http://indiewebcamp.com/s/102r
#
kylewm snarfed: hey do you know if GAE's urlfetch does if-none-match/if-modified-since kind of stuff?
#
mblaney IPFS is https://ipfs.io
#
loqi.me created /IPFS (+50) "prompted by mblaney https://indiewebcamp.com/irc/2015-09-14/line/1442277159551 and dfn added by mblaney" (view diff)
#
Loqi ok
#
snarfed kylewm: huh probably! it shouldn't mess with things too much; should just pass headers back and forth directly
#
snarfed the list of prohibited request headers is documented, and won't include those
#
snarfed we use etags with bridgy/granary
tantek joined the channel
#
kylewm snarfed: oh so that is probably the opposite of what i want :) i want it to do that stuff on its own magically without me having to cache responses in my app
#
kylewm snarfed: back to your thoughts about original post discvoery, does bridgy already take into account "already found a POSSE copy on this service"?
[snarfed] joined the channel
#
[snarfed] kylewm: interesting! no it doesn't
#
[snarfed] oh but yes, urlfetch will often use 304/etags itself to not refetch
#
kylewm oh, phew
#
[snarfed] but then it'll still give you the original response unless you specifically include those request headers
#
kylewm woot
#
kylewm snarfed++ thanks
#
Loqi yay!
#
Loqi snarfed has 140 karma
#
[snarfed] don't quote me on that though. it's a service outside of gae (and it's also been many years :P)
#
kylewm well, i mean they say it leverages the infrastrcuture of google
#
kylewm so presumably if it's not doing etags and stuff, it's because it's doing something better/smarter
#
[snarfed] and also it has to be careful to respect your own request
#
[snarfed] this is for forkontext?
#
kylewm yeah, GAE seems like a good fit for that
#
kylewm it's just hosted mf2util with caching :)
#
[snarfed] so i was thinking about that. with a js parser, and if the user's site sets CORS headers, could you do it entirely in client js?
#
kylewm sure
#
kylewm that's a big if
#
[snarfed] the second one? eh
#
[snarfed] with forkontext, they'd use it from server side?
#
kylewm thinking jsonp
#
kylewm it'd fetch the parsed mf2 json with jsonp, and then format it however
smus joined the channel
#
kylewm also, I think I want stuff to be cached indefinitely in case the target site goes away
#
[snarfed] right, good idea
#
kylewm it sort of hurts my head to think about how aggressively to cache in the middle there... like maybe every incoming request to generate an outgoing request
#
kylewm s/to generate/should generate
#
Loqi kylewm meant to say: it sort of hurts my head to think about how aggressively to cache in the middle there... like maybe every incoming request should generate an outgoing request
#
kylewm network topology stuff is kind of counter-intuitive
#
[snarfed] one thing is that req'ts for archiving are mostly disconnected from req'ts for caching
#
[snarfed] eg you can archive and still try to refresh each time
#
[snarfed] and caching req'ts are also mostly disconnected from fetching strategy
#
[snarfed] eg you can always request with an etag if you have one, even if you decide your own cached page is expired
#
kylewm oh right
#
kylewm trying to refresh on every request seems like overkill
scoates joined the channel
#
[snarfed] eh. straw man: cache with 24h timeout, but fetch every time async (ie trigger background task)
#
[snarfed] you should talk with KevinMarks, he's been tuning exactly this kind of behavior for his hovercards
#
Loqi +1
#
kylewm cool, will do
#
kylewm (i like the async refresh idea, hadn't considered that)
frzn and [kevinmarks] joined the channel
#
[kevinmarks] Have a look at the unmung code for how I do it
#
[kevinmarks] I'm using memcache though, not datastore
#
[kevinmarks] I serve from cache, then schedule an async refetch and parse
#
[kevinmarks] Also, I set expiry and etag last modified myself so browsers can cache the hovercard
shiflett, tantek and [shaners] joined the channel
#
[shaners] I'm back!
[kevinmarks] joined the channel
#
[kevinmarks] hm http://deep.design/mobile-first/?utm_content=buffer676b6&utm_medium=irc&utm_source=indiewebcamp.com&utm_campaign=webmention
frzn, Kongaloosh, KartikPrabhu, sensiblemn, reidab, jacus, halorgium and snarfed joined the channel
#
KartikPrabhu why all that utm stuff?
#
tantek (yeah seriously)
#
Loqi tantek: snarfed left you a message 3 hours, 16 minutes ago: re "subpost," meh. obvious expansion of subtweet. i haven't really put any thought into it. http://indiewebcamp.com/irc/2015-09-14/line/1442276435492
#
snarfed welcome back shaners!
#
KartikPrabhu also that article does not make sense.
#
KartikPrabhu assume "mobile first" means something it isn't and then criticise it
#
tantek KartikPrabhu: TBH I ignored it due to the trendy (meaning unestablished) TLD
#
KartikPrabhu I am a fast skimmer so finished that article in 2 mins instead of the "recommended" 16
#
KartikPrabhu must write post about stupid long-form post trends like "reading time"
#
kylewm tantek: nTLD;dr?
#
KartikPrabhu lol
#
tantek kylewm++ lol
brianloveswords joined the channel
#
Loqi kylewm has 238 karma
jacus joined the channel
#
kylewm honest question, would you have ignored if it were on medium.com?
#
KartikPrabhu it looks like medium on custom domain
#
tantek I do ignore most medium.com URLs its true
#
tantek most posts there seem to be kinda longwinded and pompous - pretty design slapped on mediocre (fluffy) writing for the most part
#
KartikPrabhu new age saying "Don't judge a post by its TLD"
#
KartikPrabhu though most Medium posts are exactly as tantek described
#
tantek Most Medium posts aren't actually medium-form writing - they tend to be medium-form concepts bloated to long form writing.
#
KartikPrabhu I have noticed this with blogging too. Some how people feel the need to write long things when they think of blog posts
#
KartikPrabhu notes++ for that reason
#
Loqi notes has 1 karma
#
tantek KartikPrabhu: indeed!
#
tantek longnotes++ for that reason - longer than Tweets FTW!
#
Loqi longnotes has 1 karma
#
tantek Also I've been having very good responses/interactions with "long notes" when they're POSSEd fully intact into FB (e.g. using Bridgy Publish)
#
tantek so much so that I'm definitely tempted to use Bridgy Publish for any POSSE to FB automatically when the length of the note is > tweet length
#
KartikPrabhu yes I have a lot of trouble condensing long notes to fit Twitter's mandates
reidab joined the channel
#
tantek I kind of gave up on that and just tried to keep the first sentence / part that goes to Twitter interesting enoughto make the permashortlink worth clicking
sensiblemn, halorgium and minsky joined the channel
#
@neocitiesweb RT @_crossdiver: I love pinning @neocitiesweb sites to my local #IPFS node. The #indieweb just keeps getting cooler. (twitter.com/_/status/643638571509747712)
#
tantek anyone else here POSSE replies outside of their original post context?
tantek and mdik_ joined the channel
#
tantek like a reply using your own site to a comment on FB, but then POSSE your reply to Twitter and prefix it with the @-name of the person you're replying to on FB
#
KartikPrabhu I do that with likes and disguise them as normal posts on the silo
#
tantek whoa interesting. have you documented this practice anywhere
#
KartikPrabhu no. but here is an example https://kartikprabhu.com/notes/like-hypocrites-guide-public-speaking
#
tantek Thanks. I'd just prefer to make up something new
#
tantek *not to make up something new*
#
tantek the last time I deep thought about this I wrote up this I think: http://indiewebcamp.com/Twitter#POSSE_Replies_to_Twitter (specifically the subsection original lacks POSSE tweet )
#
KartikPrabhu right so mine fits in with "reference the indieweb in-reply-to URL as a reference inline "
#
tantek oh neat!
#
KartikPrabhu those suggestions seem pretty sound to me
loic_m joined the channel
#
GWG !tell pfefferle Could you have a look at https://core.trac.wordpress.org/attachment/ticket/33735/33735.diff and tell me if you think it would offer a better way to solve the Webmention Notifcation problem?
#
Loqi Ok, I'll tell them that when I see them next
lukebrooker, cweiske, lewisnyman, LauraJ and pfefferle joined the channel
#
Jeena I need to come up with some strategy on how to get more people from outside of our company to attend HWC's here in Gothenburg.
#
cjk101010 are there any web people at your company?
#
cjk101010 besides you, of course
eschnou and friedcell joined the channel
#
pfefferle good morning all
#
Loqi pfefferle: GWG left you a message 2 hours ago: Could you have a look at https://core.trac.wordpress.org/attachment/ticket/33735/33735.diff and tell me if you think it would offer a better way to solve the Webmention Notifcation problem? http://indiewebcamp.com/irc/2015-09-14/line/1442295296171
friedcell, interactivist and squeakytoy joined the channel
#
KartikPrabhu Jeena: why not start by asking the usual IRC question IRL? "Do you have/want a personal website to post all your web-things?"
petermolnar joined the channel
#
Zegnat Jeena: I might try travel down again for the October 1st one, assuming there is going to be a HWC then.
#
Zegnat Also, good morning IndieWeb!
#
Jeena KartikPrabhu, the people I asked it in IRL answered "no" with a questionmark in their face ^^
#
KartikPrabhu Jeena: then maybe point them to http://indiewebcamp.com/why and if they are still not interested then "give up" (that's what I did)
#
Jeena I have to handle everything with care, the people who are coming are already negative to the whole web-thing (most of them are embeded programmers) so I try to help them with normal websites like blogs, etc. it's a long way to any indie stuff here ;)
#
Jeena and by indie stuff I mean the things we write about in the wiki like webmentions, etc.
#
Zegnat Jeena: I guess getting them started with Known as a blog might be a good idea? It would instantly have them connected to the IndieWeb
pfefferle joined the channel
#
KartikPrabhu Jeena: I empathise. Most people I know don't know/care about owning their web-presence. IMO this stems from the the idea that either silos will be around forever or they don't really care about the things they post on silos. In either case I don't know of any good argument to convince them so <shrug>
#
Jeena yeah, hehe one of them has wordpress already but is not really interested in the whole social media thing, he just wants to post blogposts, the other started writing a small CMS in Haskell, but didn't show any social media interest either ;)
#
KartikPrabhu not having social media is fine but enabling website-to-website comments might be of interest to them
#
KartikPrabhu In fact I have found that photographers and designers are more inclined to own their data
#
Jeena they already have a comment form
#
Jeena yeah
#
Jeena that is why I am trying to find out where to find people who are interested but just haven't heard about IndieWeb yet
#
KartikPrabhu specially photographers who want to keep their photos and avoid absurd censorship rules of the silos
#
Zegnat photographers are an interesting demographic as they are very concerned about their copyrights. But I have also noticed a lot of them turn out to be very non-technical outside of their niche
#
Zegnat they require a lot of help
#
KartikPrabhu Zegnat: yes. hence photographers are a nice demographic for expanding indieweb tools to be more user-friendly
#
KartikPrabhu unfortunately not a lot of photo hosting on own domain tools exist AFAIk
#
KartikPrabhu bridgy Flikcr backfeed is a good start though
#
KartikPrabhu also photogrpahers care more about the visual design of a website than "programmers"
glennjones joined the channel
#
Zegnat Actually there are a few solid CMS that work well for photography
#
Zegnat Indexhibit comes to mind - though I am biased as I spent a summer working for them
#
Loqi [mention] someone wrote a post that linked to http://indiewebcamp.com/webmention https://webmention.io/notification/lGUEBfZBbgeT9i1SGdRZow
#
Zegnat They are mostly advertised as "portfolio" CMSes though, and could probably grow if they wrote guides specifically to help photographers
#
KartikPrabhu Zegnat: yes but not a lot of them are indieweb friendly
#
Zegnat To be fair, KartikPrabhu, almost no CMS is ;)
#
KartikPrabhu I present to you, Known
#
Loqi [mention] Pelle Wessman and ben thomas payne wrote a post that linked to http://indiewebcamp.com/ https://webmention.io/notification/jtn4FOoBunA-JdFDbtz_7Q
#
KartikPrabhu Zegnat: but that is my point. Indieweb has the opportunity to attract the photograpers by providing them with a indieweb photography solution
#
Zegnat Got ya. I guess the problem is finding someone selfless enough to invest decent time into developing tools they themselves might not even use. Most of us aren’t photographers. And that goes against the scratching of your own itches idea
#
KartikPrabhu I do a lot of photography and photo hosting is on my itch-list, but it is enough work that I have never started it
#
KartikPrabhu as a fallback I host my photos on Google Photos :P
LauraJ joined the channel
#
Zegnat Has anyone here requested to be part of the Let’s Encrypt beta?
#
Zegnat Their first cert is out now, and they will be starting with beta testers the coming months: https://letsencrypt.org/2015/09/14/our-first-cert.html
sanduhrs joined the channel
#
petermolnar I did, no reply yet
#
petermolnar although it will be an interesting new era of phishing
#
petermolnar and it's going to cause us much trouble
#
Zegnat Interesting thought, petermolnar. You mean because we have trained people to "trust" websites with a "green lock"?
#
petermolnar training is one thing
#
petermolnar the other thing is that SSL in theory was something checked
#
petermolnar I mean in theory you should not have been able to get a valid, trusted root ca published ssl cert for, for example, a bank's registered domain
j12t joined the channel
#
voxpelli !tell KevinMarks I can manually remove mentions and more is in the works: https://github.com/voxpelli/webpage-webmentions/issues/5 and https://github.com/voxpelli/webpage-webmentions/issues/28
#
Loqi Ok, I'll tell them that when I see them next
lukebrooker, friedcell, j12t, lewisnyman, frzn, LanceyWork, LauraJ, eschnou and pfefferle joined the channel
#
@reinergaertner ...ich wünschte, Indie wäre auch morgen (und ich hoffe weiter) ... #indieweb https://twitter.com/oler/status/639423821771091968 (twitter.com/_/status/643753276442288128)
tvn, ttepasse, j12t and hs0ucy joined the channel
#
@pfefferle RT @oler: Endlich ist der Mitschnitt online! Felix @diplix Schwenzel: Indie war gestern https://www.youtube.com/ #nebenan #indieweb (twitter.com/_/status/643760567346438145)
friedcell, cuibonobo, nitot, j12t, acegiak, lewisnyman, tantek, mapkyca, fourtonfish and mlncn joined the channel
#
tantek hey all - what do people think of doing an IndieWebCampSF 2015-11-14..15?
LauraJ joined the channel
#
tantek or Cambridge for that matter?
#
GWG tantek: Keep me posted on either. But Cambridge is closer for me.
#
tantek got it
scoates joined the channel
#
GWG tantek: So I think I'm interested.
#
tantek GWG cool - as participant or co-organizer?
#
tantek.com edited /2015/planning (+184) "edit list of potentials for rest of year, another possible date for SF, more sublisting, prompting for dates" (view diff)
#
tantek GWG, could you add yourself accordingly (and perhaps with dates) to https://indiewebcamp.com/2015/planning#Planning items that apply to you?
#
tantek.com edited /2015/planning (+31) "/* Planning */ fix dates and explicitly note when AEA dates are" (view diff)
snarfed joined the channel
#
tantek Does anyone else implement "On This Day" functionality on their site?
#
tantek At its simplest it's just a way to show posts on the same day but from all past years. From there you can add fancy presentation/UI to make it look nicer / more timeline-like / more navigation etc.
#
tantek Theoretically you can do this for any "day"
#
tantek Facebook supports it for *today* for you personally at https://www.facebook.com/onthisday/
#
tantek just wondering if anyone implements that themselves, or has an itch to do so
#
tantek worth documenting on the wiki?
j12t joined the channel
#
@dissolve333 #indieweb @kevinmarks, know of a micropub client geared at livetweeting / liveblogging? (https://ben.thatmustbe.me/note/2015/9/15/9/) (twitter.com/_/status/643794130854764544)
#
kylewm does anybody have an idea the order of magnitude size of some of the shared twitter block lists?
#
kylewm nm, looks like 10,000 for the one i have in mind
shiflett, wolftune, shiflett_ and mlncn joined the channel
#
@dissolve333 #indieweb i really should make hashtags auto set that category (https://ben.thatmustbe.me/note/2015/9/15/32/) (twitter.com/_/status/643812003706900480)
eschnou and gavinc joined the channel
#
@dissolve333 #devopsdaybos #indieweb "Open Networker" what I used to call "Friend Collector" (https://ben.thatmustbe.me/note/2015/9/15/44/) (twitter.com/_/status/643819270397292544)
snarfed, mlncn, elf-pavlik, cleverdevil, tantek, chreekat, eschnou, sensiblemn and KartikPrabhu joined the channel
#
GWG tantek: Depends. If it involves venues...I'm not tackling that. But if I can help out, I will.
#
tantek GWG - that's ok - put that in as a caveat then. Explicitly list what you *can* help out with.
snarfed joined the channel
#
@misuba I used to think waves of "fuck Facebook for this change, I'm quitting" were pretty regular. Haven't seen one in a while, though. #indieweb (twitter.com/_/status/643850257206480896)
KevinMarks, hs0ucy and snarfed joined the channel
#
voxpelli For those that wants to follow a block list discussion: https://github.com/snarfed/bridgy/issues/473
#
voxpelli !tell KevinMarks Follow this issue for more on the block list/moderation of Bridgy and my endpoint: https://github.com/snarfed/bridgy/issues/473
#
Loqi Ok, I'll tell them that when I see them next
#
tantek voxpelli: can you link to it from /block ?
#
voxpelli tantek: sure
#
voxpelli hmm – I wonder if exposing a blocklist through a micropub query could make sense?
#
voxpelli but how would one then get realtime updates about changes to it? one would like to subscribe to it through something like pubsubhubbub...
#
KevinMarks the other thought would be to define an indieweb "block" api that we mirror through to the silo
#
Loqi KevinMarks: voxpelli left you a message 8 hours, 47 minutes ago: I can manually remove mentions and more is in the works: https://github.com/voxpelli/webpage-webmentions/issues/5 and https://github.com/voxpelli/webpage-webmentions/issues/28 http://indiewebcamp.com/irc/2015-09-15/line/1442310726594
snarfed joined the channel
#
Loqi KevinMarks: voxpelli left you a message 4 minutes ago: Follow this issue for more on the block list/moderation of Bridgy and my endpoint: https://github.com/snarfed/bridgy/issues/473 http://indiewebcamp.com/irc/2015-09-15/line/1442342094280
#
tantek what if I wanted to include someone else's block list in my own, dynamically?
#
voxpelli KevinMarks: micropub already has a way to query for information – so that would basically be building a "block api"
#
KevinMarks the challenge is delegated auth between indie services, but I think indieauth is robust enough to do that
#
tantek and then what if someone else wanted to then include my entire block list (including what I dynamically included) ?
#
tantek could we form a ... wait for it ... block chain?
#
KevinMarks right, I'm thinking more about the UI for "there is webmention abuse on my page - I want to remove this and propagate the block" idea
#
voxpelli I guess block lists would be such a kind of content where you might want to show different info depending on who accesses it – private / public and maybe even different between different private users
#
tantek block list chains
#
KevinMarks block party!
#
voxpelli has the case of advertising the possibility of "upgrading" a public page to a private page been discussed/solved?
#
KevinMarks we've talked about it sometimes
#
M-Kodo I'll endorse any tech that yields something called a block party
#
voxpelli To the public I say that I block A and B, but if eg. tantek asks I'll tell him to not trust C and D either
#
KevinMarks http://epeus.blogspot.com/2008/05/mixing-degrees-of-publicness-in-http.html
#
tantek voxpelli: ooh that's interesting
#
tantek Kevinmarks what is publicness?
#
tantek is that related to publics?
#
KevinMarks not in that case
#
KevinMarks this is the api equivalent of cookies
#
voxpelli one way could be to add a HTTP header: "Vary: Authentication" :P
#
tantek Upgrade: Authentication
#
KevinMarks where we can view the same url and see different things depending on who we are
#
M-Kodo Speaking of cookies has anyone heard if macaroons? I hear they're supposed to be better or something
#
M-Kodo Not completely sure what they are yet though
#
KevinMarks I was suggesting just WWW-Authenticate: IndieAuth on the 200 page
#
M-Kodo Heard of*
#
KevinMarks ratehr than a 401 response
#
voxpelli interesting
#
@InramsT Piwik Is An IndieWeb Alternative To Google Analytics - http:///piwik-is-an-indieweb-alternative-to-google-analytics/ https://twitter.com/InramsT/status/643858402591002624/photo/1 (twitter.com/_/status/643858402591002624)
#
voxpelli if one does that and then advertises a unique URL as the rel-self that PuSH expects then one could even get private Pubsubhubbub through ordinary hubs
#
voxpelli and same could work for block list and h-feeds and everything
#
tantek indeed
#
tantek I might even do semi-public stuff like that
#
tantek e.g. public crawlable vs. public anyone with IndieAuth credentials (no whitelist)
#
voxpelli yeah
#
kodfabrik.se edited /block (+88) "Adding link to Bridgy discussion" (view diff)
#
voxpelli Now where do we document such a brainstorm?
#
voxpelli it's basically https://indiewebcamp.com/indieweb-messaging where the 401 is replaced with a 200 and where the fetch isn't initiated by a WebMention but by consuming a block list or an h-feed or such
#
KevinMarks so is it http://indiewebcamp.com/private_posts
#
KevinMarks or are blocklists not posts?
#
KevinMarks hm.
#
KevinMarks being tantek for a moment "everything is a post"
#
tantek except events
#
tantek I mean, there's a post that's the creation of the event
#
voxpelli h-feeds are lists of posts
#
tantek but the event itself is not a post
#
KevinMarks so a block action would be a post saying "kevinmarks blocked [hitler-avatar-chap]"
#
tantek right voxpelli
#
voxpelli I'm thinking a block list is more like a private hostile XFN-list?
#
tantek KevinMarks: sure - that's like /follow posts
#
KevinMarks then the blocklist would be a h-feed of those
#
tantek in fact, exactly, the flipside
#
KevinMarks except they should likely be private
#
tantek if you model followings, you'll likely come up with something applicable to blocks
#
KevinMarks well unfollow would be the flipside
#
voxpelli the posts would be to broadcast the activity of adding or removing someone to the list – but the actual XFN-like list would be the canonical list of everyone blocked
#
tantek right
#
voxpelli so as one can import an XFN-listin a readerto populate who one follows, then one should be able to do the same to import who to block with this
#
KevinMarks so then I would webmention the block post to my upstream indieservices
#
voxpelli KevinMarks: I rather think they would discover it from your profile and import/subscribe to it
#
KevinMarks right, I was thinking about the 'get this off my page now' case
gavinc joined the channel
#
KevinMarks where a webmention may be better than a polling based model
#
voxpelli KevinMarks: Pubsubhubbub-subscription?
#
KevinMarks so, we could do public blocks realtively easily
#
@benwerd Federalist is a new platform for efficient, responsive, static federal websites. https://18f.gsa.gov/2015/09/15/federalist-platform-launch/ #indieweb (twitter.com/_/status/643862114625564672)
#
KevinMarks and if we can get this auth dance right, make it work for private ones too with the usual polling/PuSH/webmention flows
#
voxpelli if the private block list adverties a unique URL that includes a token as its rel-self, then it can do private lists as well
interactivist joined the channel
#
voxpelli and since that URL were to be only used to subscribe to the page at the PuSH-hub there is no danger if the unique URL leaking
#
KevinMarks is federalist closed source?
#
KartikPrabhu is anyone documenting this, because I can't follow at the moment
#
voxpelli KartikPrabhu: sure – will find a place for it and mention it in the issue mentioned above
#
KevinMarks oh, there we go: https://github.com/18f/federalist
#
KevinMarks "By default, the application should use local disk storage in place of a database."
#
KartikPrabhu database also stores to local disk
#
KartikPrabhu unless they mean in human readable files
#
KevinMarks they mean don't install postgres locally
#
KevinMarks though apparently the do run it on the server
elf-pavlik joined the channel
#
kodfabrik.se edited /private_posts (+1057) "Documenting the brainstorming around upgrading public pages to private ones" (view diff)
#
voxpelli KartikPrabhu: there you go – will link to it from the issue
#
KartikPrabhu nice
#
tantek voxpelli++
#
Loqi voxpelli has 49 karma
#
voxpelli And updated issue with even more thoughts: https://github.com/snarfed/bridgy/issues/473
hs0ucy joined the channel
#
kodfabrik.se edited /block (+190) "Add link to additional implementation issue and to public post upgrading brainstorm" (view diff)
#
voxpelli Core challenges I see yet, that I outlined in the Bridgy issue, is how to discover a block list and how to mark up a block list
#
voxpelli A "u-blocks" on the h-card that identifies a user that links to an XFN-like list that rather than rel-contact has a rel-blocked or something?
#
tantek no plurals in mf property names
#
voxpelli And a "u-follows" for the reverse whitelist that would link to a standard XFN-list that shows people one follows
#
tantek we dont' have h-entries - we have h-feed
#
tantek similarly, you would want a block-list
#
tantek or a block-feed
#
tantek depending on the semantics etc.
#
tantek also "blocks" is quite a generic term
#
voxpelli "u-block-list" and "u-follow-list" perhaps then?
#
tantek that's certainly an improvement
#
voxpelli "u-ignore-list" could be an alternative
#
rhiaro I read "u-blocks" as a verb not a plural
#
tantek and for experimentation, use the -x- prefix e.g. u-x-block-list
#
tantek rhiaro: and that's one of the reasons we don't have plural microformats property names
#
rhiaro as in voxpelli blocks [these people]
#
tantek right
#
tantek voxpelli: re: ignore - you may want to take a look at /mute
#
tantek not sure if you mean the same thing
#
rhiaro works that way though, right?
#
tantek what is ignore
#
Loqi It looks like we don't have a page for "ignore" yet. Would you like to create it? http://indiewebcamp.com/s/102s
#
tantek please see /mute before creating /ignore
#
tantek and either define differently (if necessary) or redirect
#
voxpelli in the current case /mute and /block would work the same – /mute is more or less a more ligh and/or temporary version of a /block
#
voxpelli and I won't be exposing any of these lists myself at the moment – just want to push things along and be ready to consume then
#
aaronparecki.com edited /friendly (+362) (view diff)
#
voxpelli (the interesting thing of a "u-follow-list" is that Bridgy could then easily expose the people one follows there as well so one could automatically import them into ones indie reader without manually adding them to ones own sites XFN-list)
#
voxpelli I think "u-follow-list" and "u-block-list" has a whitelist/blacklist kind of relation – so they go kind of hand in hand
#
rhiaro I have started posseing twitter follows when I remember: http://rhiaro.co.uk/follows But not yet generating a list of rel=follows from that
#
rhiaro Would like to, for readers to be able to discover automatically
#
rhiaro so I have follow posts, but not a follow list, though the list can be implied from the posts
#
rhiaro making it explicit is probably more useful for search/discovery
#
tantek.com edited /specifications (+806) "braindump Specs to iterate on in rough order of simpler/shorter first" (view diff)
#
elf-pavlik u-like-list and u-read-list and u-watch-list and ... https://github.com/w3c-social/social-vocab/wiki/Verbs---owl:Class-vs.-rdf:Property#collections
#
voxpelli rhiaro: yeah, unless one has a filtered h-feed that only contains follow activities and where one posts unfollows as publicly as one posts follows a list is really needed
#
tantek elf-pavlik: nope - based on real world implementation needs
#
voxpelli (I would _never_ publicly announce that I'm unfollowing someone)
#
tantek voxpelli: people publicly announce unfollowings on Twitter all the time
#
tantek just search for unfollowed
#
kylewm those people are usually jerks :P
#
tantek they might not announce ALL their unfollowings
#
tantek but they do
#
tantek announce some of them, sometimes to make a political statement
#
voxpelli tantek: I know – some people like being rude I guess :)
#
tantek e.g. being disappointed with someone's support of a trolling group for example
#
tantek or politicians saying bad things
#
tantek there are reasons to give public feedback like that
#
tantek it's not all drama (I know, I know, Twitter ;) )
#
voxpelli yeah, I just don't like it, it's not me
#
rhiaro voxpelli: right. I think about generating the list as kind of a 'side effect' of follows and unfollows posts
tantek joined the channel
#
rhiaro Read post: http://rhiaro.co.uk/2015/08/1438731932 Don't yet have /reads for a list of them all. But yeah, lists of relations with other things (people, posts, documents) could be generated from create/remove (eg. follow/unfollw) type h-entries
shiflett joined the channel
#
@kevinmarks RT @benwerd: Federalist is a new platform for efficient, responsive, static federal websites. https://18f.gsa.gov/2015/09/15/federalist-platform-launch/ #indieweb (twitter.com/_/status/643881925359222784)
cleverdevil joined the channel
#
kylewm I'm just a little bit amused that the block list conversation got so complicated because snarfed pushed back on adding a little complexity to bridgy
#
@alenevince RT benwerd: Federalist is a new platform for efficient, responsive, static federal websites. https://18f.gsa.gov/2015/09/15/federalist-platform-launch/ #indieweb (twitter.com/_/status/643885323303432192)
interactivist, cleverdevil-, eschnou and tantek joined the channel
#
tantek indeed kylewm - it really is up to the receiving site to handle blocking etc. anything else is value added by services in the middle
#
tantek and if you're delegating your webmention handling to another service, including displaying, then presumably you're ok with whatever their level of blocking (or lack thereof) that's supported.
lukebrooker joined the channel
#
tantek btw to all: would appreciate feedback on this list: https://indiewebcamp.com/specifications#Specs_to_iterate_on - in particular regarding rough order of simpler/shorter first, and any feedback about "yes that spec is essential to my implementation" to "no, that spec sounds hypothetical that no one needs". thanks!
lukebrooker joined the channel
#
@tedcurran Can I get an AMEN, @Google, @Twitter? | It's called #RSS http://blog.superfeedr.com/it-is-called-rss/ #openweb #indieweb (twitter.com/_/status/643904197486710784)
#
KartikPrabhu tantek: add votes below list?
#
KartikPrabhu s/list/list item
#
Loqi KartikPrabhu meant to say: tantek: add votes below list item?
#
kartikprabhu.com edited /specifications (+322) "/* Specs to iterate on */ my votes" (view diff)
snarfed and Jihaisse joined the channel
#
tantek thanks KartikPrabhu - that works
Unifex, snarfed, lukebrooker, scoates_, snarfed1, scoates, tantek and lukebroo_ joined the channel
#
tantek FB changed their /hovercards! cc: KevinMarks
#
tantek they dropped all the buttons!
snarfed joined the channel
#
snarfed.org edited /specifications (+470) "/* Specs to iterate on */ OPD, syndicate by reference/content" (view diff)
#
KevinMarks they look the same to me tantek- screenshot?