#jackyI was trying to think of something myself that could "stick out" in the raw format of my entries
#aaronpkyou should be able to use a regular @mention of a nickname and it can look it up using the search
#[cleverdevil]Agree. Just haven’t gotten around to it yet.
#aaronpkI don't want to have to do anything special, so I made mine work that way. it does require that every nickname is unique, but it;s my own personal nickname cache so that's fine
#snarfedeg silo profile pictures are global: each owner owns theirs, so when they change it, everyone sees it
#snarfedvs eg phone contacts are local (generally ish) : you own the data you enter, eg nicknames, phone numbers, etc. they don't update when someone changes their phone #
#snarfedupdating from someone's h-card is global. nick caches inside CMSes will probably default to local, but could also update.
#snarfedfor an NNCaaS though, it's an interesting design decision where to fall on the local vs global spectrum
#jackyhm, I was opting to let upstream handle resource-stuff (like images) but names be held locally
#jackynow the issue of "refreshing" a card comes into play
#jackyit's probably easier if there's an existing "relationship" of sorts so you can just push to "friends" about your h-card updates
#jackyotherwise if it's a one-off then you got someone's info that'd eventually go stale
#jacky(not otherwise but another case - this can kinda spiral)
#jackyI just see consent (if someone didn't want you saving their photo) and disclosure ("hey I updated my name/bio/nickname")
[tantek] joined the channel
#[tantek]updates disclosure/notification already works via WebSub
#[tantek]e.g. if you want updates on my h-card, you subscribe to my home page via WebSub and grab the h-card whenever there's an update
#snarfedyeah permissioning nick cache adds a huge chunk of additional complexity. we already do eg webmention without that, i'd be inclined to punt on it for nick cache
#[tantek]I believe "cache" obviates something which is otherwise public
#[tantek]definitely interested to hear of exceptions though!
#aaronpkI think of mine more as an address book really
#snarfed"nick cache" isn't really an ideal name anyway, it's basically an address bok
#aaronpkwhether my address book is smart and can automatically update from someone's website is a different story
#aaronpkgoogle contacts works that way actually. it's a local address book, but when you connect an entry to someone's google+ profile then it starts automatically updating things.
#snarfedright. all address books are the prior art there, plus modern ones like gmail that auto update profile pictures, other info
#snarfedsilos are kind of the equivalent prior art for wholly global address books
#snarfedi don't know if [cleverdevil]'s is wholly global, but you could imagine one that is, basically just a proxy cache for fetching and parsing h-cards that insulates you from some people's sites being slow or down
#jackyhas no idea if that's the emoji or just boxes
#snarfedalso parsing h-cards to build an address book/nick cache requires running eg representative h-card algorithm, probably others, which starts to make it reasonable to offload to a service (sometimes0 to avoid implementation complexity
#[tantek]you might be building more of an address book yourself, however I specifically named "nickname cache" to indicate the very minimal implementation that is: URL for someone (and cache of their h-card from there) + an optional *local* @-name you assign (in case their h-card lacks a 'nickname' property and just their first name alone isn't enough etc.)
#jackygoing back to using xray as a way to abstract this a bit
#[tantek]snarfed, it's not an "ideal name" for what you have in mind, because you have something different in mind. I named it because it precisely indicates just what it does, no more
#[tantek]and is A LOT less work than an "address book" by any measure
#snarfedoh yeah i definitely don't mean that we'd reimplement traditional address books
#[tantek]it's used to provide auto-suggest of names, friendlier display names for domains etc.
#snarfed(actually i take that back, it does mention fetching h-card once)
#snarfedright! i've been talking about it here as including features like those, not just the pure plumbing
#[tantek]"address book" is by default 100% local / static / dead info (zero updates) because that's how they all work in practice, and if you auto-update something in someone's address book, they will be surprised.
#[tantek]if you're building something with any notion of updates, I'd suggest picking a different name, especially if it's going to be a user-facing feature
snarfed joined the channel
#[tantek]"address book" has too many legacy assumptions
#[cleverdevil]I’m definitely just tinkering at the moment FWIW 😀
#snarfedyeah, tradeoff. baggage but also self explanatory for everyone. meh.
#[cleverdevil]Not sure what the answer is. Want to build some things, try, fail, and repeat 😉
#[tantek]explanatory for everyone, because it has baggage 😉
#aaronpkI disagree, there are plenty of examples of startups that provide "smart" address books
#snarfed[cleverdevil]: sure! it's global, right? ie you don't have "users," you serve everyone the same data, right?
#snarfedslowly backs away from the semantics debate
#aaronpkand I think the implication that an address book is a private thing is useful here
#[tantek]aaronpk and they've all confused users and died
#aaronpkvs trying to be an "indieweb directory" or whatever
#[tantek]address books are where information goes to become stale (as people move, change phone numbers etc.)
#[cleverdevil]I see the argument for it not being global.
#[tantek]I mean if you want to build something purely global you can build an h-card cache. then perhaps an h-card index. then perhaps world wide person search >:D
#snarfed[tantek]: yeah that "local" baggage for address book is a good point. maybe "friend list?" to bring in the modern silo connotation where everything is auto updated?
#[eddie]Yeah I do feel like “friends list” or “contacts” is a good public facing name for this feature
#[eddie]Although I don’t like the association that the person has to be considered a friend
#[eddie]I might lean towards contacts, as gmail is widely understood and used that
snarfed joined the channel
#[eddie]And people might not be your “friend” but they are people you are contacting if you’re storing their info for webmentions
#snarfedyeah. "contacts list" or "contacts" are candidates too
[asuh], snarfed and [davidmead] joined the channel
#[tantek]sad thing about Gmail is it is weirdly static too. Google Contacts is one of the biggest messes of Address Books, and feel useless to bother cleaning up
#[tantek]also it's a misnomer for these purposes as rarely are you ever directly "contacting" the folks that you would have in a cache of this sort
#[tantek]"contacting" is certainly not the most common use-case
#[tantek]eddie, you're right about the problematic association of "considered a friend" re: friends list
snarfed joined the channel
#[eddie]Is replying to them or @-mentioning them not considered contacting them?
#[eddie]I would consider @-mentioning aaronpk as contacting him
#[asuh]is the term “friending” a bad connotation for a “friends list”? I get the semantic argument between friend and contact, but i personally like the term “friend” as it’s, well, friendly! family can be friends, new people i meet become friends.
#[tantek]asuh, in the context of social networks/media, the term "friend" has acquired some hollowness. originally is someone your "friend" or just your "friendster", today it's your "friend" or "facebook friend"
snarfed, treora, [davidmead], [tantek], [asuh] and [eddie] joined the channel
#[eddie]tantek: how strange. In certain situations I very much expect a response when I @-mention someone. Not all the time, but sometimes!
#[eddie]For example, when I was launching My Url Is, I wrote an unlisted post and mentioned both you and aaronpk. aaronpk responses, you didn’t 😆 probably clarifies this entire conversation about @-mentions and responses!
#[eddie]No, I joke. I did follow up with you, and it was because your twitter @-mentions fill up your “webmention inbox”
#[eddie]But in all seriousness, there were times on Facebook, I would post in someone’s wall, which is essentially @-mentioning them, or tweet someone a question and I have expected a response the same as I would a text message
#[eddie]Hmm interesting. That must be the essential ingredient where we differ. I think of contacting someone has being either public or private, not being required private
#[tantek]there is no "inbox" because it's not private
#[tantek]notifying may be a secondary effect, but the first effect is a public performance
#[tantek]it necessarily changes the meaning of what is being said
#[tantek](and I think usually for the worse, hence Twitter argument machine etc.)
#[eddie]That’s fair. I guess if I think about it, my post was unlisted and if we had private indieweb posts figured out at the time I probably would have made it private
#[eddie]Which then leads to the discussing that wouldn’t a “nickname cache”/“contact list” also be used when sending a private mention when that is worked out?
#[eddie]Thus swinging the discussion back to the original axis of a nickname cache being a form of a contact list?
#[eddie]And it also brings up the interesting idea that should private webmentions receive more UI prominence? Since based on your analysis Private is more authentic
#[eddie]2:30 EST means I should go to sleep. 😴 night all!
KartikPrabhu and [kevinmarks] joined the channel
#[kevinmarks]See "snitch tagging" for mentions as performance.
#[kevinmarks]This public nicknames cache sounds a bit like the old blogrolling.com
tglobe, tglobe_, swentel, barpthewire, [kevinmarks] and deathrow1 joined the channel
#Zegnat!tell [cleverdevil] I was going to make a feature request for letting pseudonym.cleverdevil.io also search p-nickname (gives me Zegnat), but then realised I am not in the db anyway, haha
[Rose], deathrow1 and [jgmac1106] joined the channel
#[jgmac1106]I still like the idea of opt-in global, I may be on ten networks but only want IndieWeb folks to mention me in four... I opt in and decide what networks to display
#[jgmac1106]Or alternatively I opt in and it caches my h-card from my website where I control networks I list
#ZegnatThing is, you can’t stop people from mentioning you on the other six networks. If those identities of yours are public, I might chose to mention you on those as well
#[jgmac1106]no but it can be a cue to the community that says hey don't mention me here....I just think folks who have faced online harrassment or other problems might have different view than "well it is online and already public so we will are maing it wasier for people to @ mention you"
#[jgmac1106]some people may not want it to be easier to @ mention them,, friction providing a tiny bit of safey
#[eddie]jgmac1106 as Zegnat said, people will do what they want in private nickname caches. However, a global approach like [cleverdevil]’s is literally just parsing people’s h-cards so if someone only wanted mentioned in 4 by limiting their h-card or rel=me to those 4, [cleverdevil]’s approach honors that
#[eddie]So I do think that global approaches that use [cleverdevil]’s tactic is the way to go, for the reasons you outlined, jgmac1106
jeremych_ joined the channel
#jgmac1106Great idea!! and I still have this vision of the opt-in IndieWeb search engine where people could decide what post types get crawled and then we could ask stuff like Hey loqi what is the most popular listen this month?
#ZegnatI think current crawler opt-out technology is pretty accepted. robots.txt et al. So I wouldn’t worry too much about starting crawling for a search engine
#[jgmac1106]but go back to idea opt-in more inclusive always than opt-out
eli_oat1, [kevinmarks], [schmarty], [davidmead], tantek, [jgmac1106] and eli_oat joined the channel
#[schmarty][jgmac1106] looks like you're using aaronpk's webring links rather than your own!
#Loqi[schmarty]: jgmac1106 left you a message 2 hours, 5 minutes ago: for some reason I can’t get my indiewebring profile recognized: https://jgregorymcverry.com/
#jgmac1106I must have copied his when I was looking at how he styled his microcast club ring or something
#jgmac1106but the webring, indieweb directory, publich h-card cache could bring everything full circle in one nice community package
tglobe_ joined the channel
#[jgmac1106]I join a ring and I can choose which networks to share so indieweb ring that is microblog and IRC but my web literacy ring is just Twitter, so when I join my
[tantek] joined the channel
#[jgmac1106]I can choose which of the five networks to add to the rings directory
#petermolnarhttp://manu.sporny.org/2012/mythical-differences/ "Microdata 1.0 itemref: RDFa Lite 1.1 not needed - Used to copy-paste a piece of data and associate it with multiple things." - that's the complete opposite, itemref was to avoid copy-paste.
Bastinkie joined the channel; Bastinkie left the channel
#petermolnareh, this whole RDF thing is just making me feel miserable
calumryan, [jgmac1106], tglobe_, KartikPrabhu, tglobe, eduardm, [schmarty], ichoquo0Aigh9ie, tantek, [kevinmarks], eli_oat, djmoch, petermolnar and [tantek] joined the channel
#[tantek]petermolnar, worth mentioning why specifically if you can link to something on /RDF