#dev 2022-01-04
2022-01-04 UTC
# 
[jacky]2 it would be 🙂
# [jacky]2 I get the commenting and liking bits are what people want to add
# [jacky]2 if anyone gets inspired to make a social reader, remember Path https://indieweb.org/Path#Mobile_UI!
# [jacky]2 someone brought it up on Twitter and I was so glad I grabbed those shots
# edgeduchess[d] oh that's nice
# edgeduchess[d] I'm glad you got the UI screenshots
# [jacky]2 mobile apps are the hardest to archive
# [jacky]2 IMO
# [jacky]2 yeah that felt natural
Seirdy joined the channel
# kandr3s[m] Hello IndieWeb
# kandr3s[m] Is anyone here using Bridgy to publish to Mastodon?
# kandr3s[m] I'm trying to syndicate to Mastodon. I set up Bridgy Fed and it works fine but realized the "normal" syndication would be better for my use case. When I click "Enable publishing" on Bridgy's page it takes me to a Mastodon error page with "We're sorry, but something went wrong on our end." I tried a different instance and got the same error. I wonder if I'm missing something? Twitter publish/backfeed works without problems.
alex11, darkkirb, [tw2113_Slack_], strugee, saptaks, vj--, aspen[d], gillytech and KartikPrabhu joined the channel
# 
@kandr3s Bridgy Fed wasn’t the solution for me, trying to set up its regular webmention endpoint has not been easy with Mastodon. Looking for solutions… (twitter.com/_/status/1478237428972507137)
KartikPrabhu, maxwelljoslyn[d], jessealama, MarkJR84[d], Murray[d] and gRegor joined the channel
grantcodes[d], [Joe_Crawford], [James_Van_Dyne], eviestears, IWSlackGateway, [chrisaldrich], [jacky], [aciccarello], [jeremycherfas], jessealama, gRegor, Tommy2, [snarfed]1, [chrisaldrich]4 and IWSlackGateway1 joined the channel
# 
@aslakr ↩️ Mulig https://www.w3.org/TR/webmention/ er en bedre løsning som er mer spesifisert for ting som sletting osv. (twitter.com/_/status/1478323424237871107)
omz13, jessealama, tetov-irc, tomleo[m], jamietanna[m], PeterMolnar[m], unrelentingtech, diegov, mambang[m], benatkin and LaBcasse[m] joined the channel
# 
Murray[d] Useful for getting started on CSS schemes: https://github.com/dbox/html5-kitchen-sink
[JampiDotNet] and jessealama joined the channel
# 
Loqi GraphQL is a query language for APIs and a runtime for fulfilling those queries with existing data https://indieweb.org/GraphQL
# Murray[d] I have an opinion in that it's all I've used at my past 2 jobs and I'm completely accustomed to it now 😄 It definitely improves front-end DX in terms of just getting up and running, but it also creates overly complex endpoints (which have a habit of getting worse over time). Strict typing is nice, too, though I find it's more often skipped (and it's not like this isn't possible outside of GraphQL either)
jessealama joined the channel
# jessealama I think there are a few places where GraphQL is a good way to go. I was recently going to write a Braintree API but was essentially pushed into using GraphQL for it
# jessealama (I also prefer a "plain" API approach and consider GraphQL somewhat faddish. But I think that if you're dealing with a service that supports it, writing GraphQL queries is pretty straightforward and gives you the data you need in JSON format. I'm glad I don't have to write a service that consumes GraphQL queries and has to convert them into potentially very complex SQL queries!)
jessealama and Seirdy joined the channel
# 
[KevinMarks] One supposed value of GraphQL is that if you have a very many-keyed and nested JSON API (eg twitter's) and you're only using some values, you can construct a query that just asks for that subset, which both reduces data size on the wire, and helps the API provider understand which callers changes will effect. I'm not sure how that has worked out in practice.
# zerojames[d] manton I have used it too. I think it’s nice that you can query only the data you need and can write nested queries. But, queries can get complicated and some things like subscriptions weren’t intuitive when I tried.
Guest6 joined the channel
# [KevinMarks] the API problem is a bit like the DB normal form problem though - You can make a twitter like API where the result is a very nested json tree, or you can make more specific APIs where you then need to make multiple calls to get all the nested bits you need (separate endpoints for authors, posts and likes for example)
# zerojames[d] Yes indeed. You can architect your back end api to send whatever data you want.
# zerojames[d] Good point aaronpk.
# 
Zegnat In my experience with GraphQL: I like it as an API consumer, but not as an API provider. Last I was involved with implementing it for a bigger project, we ended up having to build so many special logic on the backend to support specific combination queries that we would have been better off just offering that same data with REST (or REST-like) rather than giving consumers the ability to mix and match.
# 
petermolnar re: See Also, BBC, and things: I have never developed anything for mediawiki, so where would one start to add a functionality to Loqi's See Also parsing to fetch the site and try to extract info, such as <title>, plus save it to archive.org on the fly?
# petermolnar I know how to do parts of it, just not for mediawiki and/or Loqi.
# 
aaronpk you can pop a URL in here to see what Loqi sees https://xray.p3k.app/
# petermolnar regardless of all the meta:og meta:twitter, actually valid LD+JSON and <title> on the BBC page
# zerojames[d] I think you can do it on the fly?
# zerojames[d] That’s it!
# petermolnar I think, but I'll double check
# zerojames[d] If you can’t I have a script to update…
# petermolnar https
# aaronpk https://indieweb.org/Internet_Archive#Trigger_an_Archive "As of some time around 2020-08 this no longer works, and archive.org returns an error that says "You need to be logged in to use Save Page Now."
# petermolnar but yes, that's the endpoint
# petermolnar it works
# petermolnar who said it isn't?
# petermolnar well, let's try
# petermolnar yep, works
# petermolnar it's rather slow though
# zerojames[d] That’s what I planned to do aaronpk.
# zerojames[d] I have a script that should hopefully archive any URL I link to on my site.
# zerojames[d] And I am considering allowing archive.org to index my site again petermolnar 🙂
# [KevinMarks] BBC went very LD at one point, and their ontology claims to have a core:slug in this graphic https://www.bbc.co.uk/blogs/internet/entries/0e8d8442-defc-39e4-85bf-e04a385c02c9
# [KevinMarks] of course the ontology URL is a 404 now
# zerojames[d] tantek 😂
# petermolnar writing one's own ontology and or vocabulary is probably one of the most fragile things to do
# [snarfed]1 kandr3s sorry for the trouble! yeah that's a long standing Mastodon bug that sadly I've never managed to drum up interest in fixing. https://github.com/tootsuite/mastodon/issues/12915
# Loqi [snarfed] #12915 reproducible HTTP 502 error in OAuth /oauth/authorize
# [snarfed]1 maybe I'll try to ramp up and fix it myself someday
# aaronpk basically the goal of xray is to map a bunch of different formats and silos into microformats JSON: https://github.com/aaronpk/XRay/tree/main/lib/XRay/Formats
# petermolnar aaronpk: I'll try to find some time for it, but I'll have to dig into existing libraries, because the last thing I want to do with my life is writing a real structured data parser
# petermolnar that shouldn't be too hard: datePublished, headline, url; and then either publisher or author for the details
KartikPrabhu and cygnoir[d] joined the channel
# 
@EssentialRandom ↩️ I haven't explored it as much as I wish I had time to, but IndieAuth is also a step in that direction that might help bridge people into the discussion https://indieauth.com/.
Though for how much I like it, I wouldn't be surprised if what you can do with wallets goes beyond. (twitter.com/_/status/1478443322783850497)
# 
@cupidhack ↩️ yeah I love the ethos here but indieauth is still totally centralized and dependent on this one service
the advantages of wallet based authentication are a persistent identity and login mechanism that doesn't require any one service to manage your credentials (twitter.com/_/status/1478445883045789696)
[schmarty] joined the channel
# 
[schmarty] 2022: the year IndieAuth.com finally retires??
# edgeduchess[d] that's my tweet FYI
# edgeduchess[d] (the first one)
# [schmarty] not really sure how to reply that indieauth.com is _a_ service and not _the_ service.
# edgeduchess[d] i just replied that
# edgeduchess[d] so I guess there's that 😛
# [schmarty] edgeduchess[d]++ haha, wonderful 😂
# edgeduchess[d] wait until you see my full reply ahah
# edgeduchess[d] but anyway that person's point goes beyond that, and I think has its merits
# @EssentialRandom ↩️ If I understand correctly, IndieAuth needs a server to work, but your own website can be that server. But it goes back to your point about "yeah just generate PGP keys" not being a solution. The protocols are there, but if there's no easy path to adoption it doesn't matter. (twitter.com/_/status/1478448236796911616)
# edgeduchess[d] they were saying in a follow up tweet that wallets don't have to mean crypto wallets
# edgeduchess[d] and unfortunately the strong point of web3 is that if you get a bunch of money to do so is much easier to streamline the experience for public adoption
# edgeduchess[d] it's not a win of the protocols, but it's going to be interpreted that way
# [schmarty] like many conversations about wallets (crypto or otherwise), there's a big chunk of conversation missing about the actual experience of managing and using wallets
# edgeduchess[d] oh 100%
# [schmarty] feels particularly telling that they contrasted "make a crypto wallet" favorably versus "make and post a GPG key"
# 
@kevinmarks ↩️ Multiple independent IndieAuth server implementions exist https://indieweb.org/IndieAuth#Server_Implementations (twitter.com/_/status/1478449321225986051)
# @kevinmarks ↩️ No it isn't dependent on that service, you can host your own Indieauth endpoint - http://indieauth.com is a utility that uses existing OAuth servers you link to to implement RelMeAuth so you don't have to (twitter.com/_/status/1478448812482076681)
# edgeduchess[d] because the experience is easier at this stage?
# edgeduchess[d] like it doesn't have to be easier
# [schmarty] edgeduchess[d] because wallets come close to doing exactly that under the hood.
# edgeduchess[d] but AFAICT "sign in with ethereum" doesn't require that many steps to work
# edgeduchess[d] yes
# edgeduchess[d] but that's the whoel point
# edgeduchess[d] they do that under the hood
# edgeduchess[d] it doesn't matter to end users what they do as long as it works and it's easy
# edgeduchess[d] should it? yes, does it? no
# [schmarty] i bet for every bit of user design sugar a wallet has added to make it "easy" there is an example of someone getting hacked or tricked into approving the wrong thing 😂
# 
stevestreza the "wallet" bit is really incidental, the key insight there is "digital identity lives on your phone for SSO", which will beat any RelMeAuth-type sign in if just for ease of setup and accessibility
# [schmarty] the general usability complaints about GPG/PGP-style key management are valid, but the crypto wallets have the same challenges under the hood and they are only pasted-over at best.
# [schmarty] stevestreza: there are indieauth implementations with mobile integration that works exactly as you described. i think aaronpk uses one? i know jacky had one working for himself at some point.
# edgeduchess[d] yeah but how good are they going to get vs how good are cryptowallets going to get?
# [schmarty] depends on the incentives of the people who are building each, haha
# edgeduchess[d] I think people repeat a lot "it's a UX problem, not a technology problem" which is true, and that's why Web3 has a huge advantage: they can pay money to solve their UX problems despite not having an innovative technology
# aaronpk like stevestreza says, the idea of proving you control a private key by signing something is the core of this idea, and there's plenty of experiments going on with that in various circles, with and without blockchains, but I strongly suspect we have yet to see the range of new issues created by that core idea
# edgeduchess[d] yeah agree with the "range of new issues" problem
# edgeduchess[d] ooh? i'd love to try it out
# edgeduchess[d] and also point crypto people to it next time
# edgeduchess[d] ah, yes
# edgeduchess[d] but then it goes back to not being actually practically useful
# edgeduchess[d] plus the other point: if the public doesn't know about it, cryptowallets are still going to end up as the adopted solution
# edgeduchess[d] that's an interesting point
# edgeduchess[d] I wonder what Web3 people think of that
# edgeduchess[d] I'm sure people have opinions
# edgeduchess[d] I think the thinking of a bunch of people is that it gives you money to address those problems
# edgeduchess[d] also incentives for adoption
# [schmarty] i think the word "address" is potentially doing a lot of work in "address the problems"
# edgeduchess[d] which you can absolutely "but money makes everything worse" which I agree with, but it's also why it pisses me off it's likely to eventually *be* """solved""" by Web3 technologies
# edgeduchess[d] (and yes, aaron, totally fair)
# edgeduchess[d] there's a bunch, i think
# edgeduchess[d] not all of them good
# stevestreza crypto wallets have the same fundamental problem as any other passwordless key login, either you are decentralized and there's a risk of users losing the key, or you're centralized and you can have some authority be the key master with some proof of identity
# edgeduchess[d] one very meta argument would be "people get money to build web3 companies => they add authentication because it's the web3 things to do => people get sensitized towards owning your identity stuff => people demand that that type of authentication be added"
# edgeduchess[d] also "people get wallets cause $$$ => you can suddenly rely on people having a wallet and managing the keys cause their $$$ is tied to it"
# edgeduchess[d] I don't know if sign in with ETH also gives the website access to your public transaction history, in which case a) lol b) that's still incentive
# edgeduchess[d] another argument would be if you sign in with ETH you can also import your NFTs which means websites that use NFTs to "personalize your experience" or for the other mechanisms enabled by NFTs would also have an incentive to add them
# edgeduchess[d] and personalization has generally been shown to positively correlate with adoption, so there's a growth argument for spending resources on adding stuff like that
# stevestreza there is plenty of examples of people losing banking passwords or stickying them to a monitor or reusing them with their neopets account to ever think that just because it's connected to someone's money, they'll suddenly be defensive and security-minded about keeping it safe :P
# edgeduchess[d] i mean, agree
# edgeduchess[d] doesn't really matter though, i think
# edgeduchess[d] cause then you can blame the user for being careless
# Zegnat What I think is interesting is that none of those incentives seem to have made it any more attractive for existing platforms to implement it. Which makes me wonder if we are not overvaluing those incentives. But I am also super happy to have time prove me wrong, as I would love to login with throw away identities all around the web :D
# edgeduchess[d] I think most existing platforms are biding their time
# edgeduchess[d] no point getting into this "war" at this stage
# edgeduchess[d] they can afford to let smaller fishes fight each other, and then acquire the winners
# edgeduchess[d] and the PR hit is definitely a consideration
# edgeduchess[d] but whether the scales will tip....
# edgeduchess[d] an interesting point
# edgeduchess[d] but XML doesn't seem as focused on end users, and I think that distinction matters
# petermolnar Indieauth used to work with client side certs at one point, no?
# stevestreza yeah I used to use PGP with it
# edgeduchess[d] [tantek]: I mean, Web3 is big, so any statement on Web3 is bound to have at least one counterexample
# edgeduchess[d] but I think saying Web3 isn't focused on end users in general isn't necessarily true
# petermolnar "web3 is big" no, it's not. It's noisy.
# edgeduchess[d] like, there's a bunch of... "circlejerking" going on, but there's also people that are looking into doing stuff for users
# [tw2113_Slack_] like too many things, it likely has a small audience shouting extremely loudly and constantly
# edgeduchess[d] is it smaller than ours?
# edgeduchess[d] but "big" is a relative term, is big in the sense that there's a lot of sides and facets to it
# edgeduchess[d] and a lot of people with different ideals
# petermolnar It might be. With that many bot accounts, who knows.
# edgeduchess[d] unsure what you mean tantek?
# petermolnar That nobody is using web3 for anything beyond hype.
# edgeduchess[d] https://mirror.xyz/
# 
[tantek] edgeduchess[d] this: https://twitter.com/t/status/1477799190382055425
# 
@t Posting notes on your site is the “Hello World” of the independent web. If you speak to the ideals of the #openWeb #decentralizedWeb #distributedWeb #federatedWeb #indieWeb or even #smallWeb, walk your talk: post your notes somewhere you control. ...
https://tantek.com/t5G_1 (twitter.com/_/status/1477799190382055425)
# edgeduchess[d] https://gitcoin.co/
# edgeduchess[d] even opensea, for all its troubles
# edgeduchess[d] saying "no one is doing anything" is unfair
# [schmarty] there definitely used to be an indieauth client side cert example! https://indieweb.org/indiecert.net
# edgeduchess[d] and likely to just help them blindside us in the future
# edgeduchess[d] (reading now, tantek)
# petermolnar [schmarty]++ for remembering
# edgeduchess[d] I think that's reductive, again
# edgeduchess[d] I think it makes people feel better to think that, but it's detrimental in the long term
# [schmarty] gitcoin looks like it's designed to tip github users in crypto
# [schmarty] (literally centralized around github sign-in)
# petermolnar proposal: #indieweb-crypto room for these discussions? So it doesn't taint the rest of indieweb.
# edgeduchess[d] I think that would be nice, it's an important topic to keep an eye on
# [schmarty] mirror is pitching itself as a co-operative publishing platform but their site doesn't actually feature any ... published pieces? 😅
# [schmarty] oh oops i assumed we were in chat 😄
# edgeduchess[d] i can get you some published pieces
# petermolnar Chat is for community, not for hate :P
# edgeduchess[d] if you don't keep an eye on it, you're going to miss the hello world
# [schmarty] hahaha my entire body clenched looking at this URL 😂
# edgeduchess[d] especially if the hello world doesn't look like you'd expect it to look like
# edgeduchess[d] i'm not a fan of web3, but I do keep an eye on them cause not doing so is not going to make me more effective
# edgeduchess[d] anyway there's also this article on mirror.xyz and how it works, if people are curious about their theses on "tokenomics": https://medium.com/digital-marketing-lab/mirror-xyz-review-186e0960bac2
# edgeduchess[d] eh, it's hard to judge both crypto and anti-crypto essays cause they're all full of imprecisions
# edgeduchess[d] web3 is full of idiots, you need to *really* dig to find the people that have answers
# edgeduchess[d] I dont' like my use of idiots
# [schmarty] => #indieweb-chat
# zerojames[d] I joined Twitter (again) this week and all of the web 3 stuff is already stressing me out.
# zerojames[d] But this is more chat.
# 
@capjamesg ↩️ I could technically stop using my endpoint tomorrow, replace a HTML tag on my site, and start using the IndieAuth hosted service, with no lock in. (twitter.com/_/status/1478478150812196867)
# @capjamesg ↩️ I second @kevinmarks point. I implemented an IndieAuth server for personal use and there are many existing, interoperable, open implementations. (twitter.com/_/status/1478477670379933708)
# Loqi mute is the ability to hide posts in your reader that have specific words, or from particular publishers, sometimes only a specific type or format, optionally with an automatic expiration, after which newer posts are visible again https://indieweb.org/mute
# [KevinMarks] blockparty.app muted zerojames for me on twitter. I have no idea why
# [snarfed]1 [KevinMarks] odd, maybe unrelated, but Twitter is soft hiding some of your tweets for me. eg your reply to https://twitter.com/cupidhack/status/1478483755513757700 is behind a "Show additional replies, including those that may contain offensive content" click-through warning
# @cupidhack @schnarfed @capjamesg @kevinmarks @EssentialRandom yeah I don't think it's a "debate" necessarily, just optimizing for different things it's not clear to me how portable identity works under the traditional paradigm, for example (twitter.com/_/status/1478483755513757700)
# [KevinMarks] that may have been because james got auto-muted so my threading went awry
# zerojames[d] I am capjamesg on Twitter 🙂
# zerojames[d] Confusing but that was the username I could get.
# zerojames[d] I guess I’m too late to the party to get my desired username haha.
# zerojames[d] I wonder what I did to get blocked. Odd.
# [KevinMarks] so goodness knows
kimberlyhirsh[d] and [fluffy] joined the channel
# [KevinMarks] ah, it has your join date as jan 2022, and it mutes brand new accounts by default
tetov-irc joined the channel