#dev 2022-04-05

2022-04-05 UTC
[snarfed]1, IWSlackGateway, [tantek]1, feoh, doosboox5, gerben, Zegnat, ^ilhalmer, stevestreza_, GWG-, bneil, joshproehl, darkkirb, jacky, johnnrs[d], gxt, angelo, keenan[d], mro, MarkJR84[d], tetov-irc, Weapzy and edburns[d] joined the channel
#
@GinaRaeMitchell Join me today for a feature on Ten Thousand I Love Yous by @LisaSlabach. Check out the massive #Giveaway, Author Guest Post & Interview, #BookTrailer. #WomensFiction #FindYourself #Divorce #WritingCommunity #BookTwitter @iReadBookTours #Romance #IndieAuth https://ginaraemitchell.com/ten-thousand-i-love-yous/ (twitter.com/_/status/1511313291611627528)
nertzy, mro, grantcodes[d] and jeremycherfas joined the channel
#
chewing_bever hi everyone! i set my site up about a week ago & i just wrote a post describing my current workflow, if possible i'd love to get some input on my workflow (or my site, always happy to hear), thanks :) https://rustybever.be/workflow/
#
chewing_bever i'm posting this here cuz it does contain some techy talk, so it might not be suited for the non-tech savvy user
jacky and jamietanna joined the channel
#
jamietanna Good stuff chewing_bever, that sounds good :) I'm at the point now with my site where I'm needing to migrate away from static sites / Hugo as my site takes several minutes to build 😬
mro joined the channel
#
chewing_bever thanks for the input! now i'm wondering exactly how big your site is that the builds times are that extreme 😅
[schmarty] joined the channel
#
jamietanna https://gitlab.com/jamietanna/jvt.me/ has 30,000 files in it, of which 18,000 are content and ~12,000 are contextual info. Doesn't help that the CI nodes I can run on generally have poor IO so the build is slow. On my laptop, with NVMe, it's only ~60 seconds
#
chewing_bever Oh wow, yeah that is really big
#
chewing_bever I've seen you site before actually ;p small world
#
jacky clever approach for caching cites
#
jacky I've been embedding them into the referenced pages
#
jacky but splitting them out probably makes it easy to refresh
#
jacky could go all of the way and split out the author info lol
#
aaronpk i used to store external posts separately but i switched back to embedding it in the post itself. just seems easier to keep everything self contained
#
aaronpk so yeah if i reply to the same post twice i end up with two copies of it, but meh
#
jacky the one case I have for it is progressive enhancement of my reply context
#
jacky like if I store it stubbed initially and have an async job to 'hydrate'/munge info for it, that'd be cool
#
aaronpk yeah that's basically how mine works. my reply post is created with just the URL of the thing i'm replying to, then later it goes and fetches it and expands it and replaces the URL with the full post contents
#
aaronpk if that fetch fails for some reason then i'm left with just the URL and that's why you see some of my replies show just a URL instead of the full reply context
#
jacky wrote out how I do it at https://indieweb.org/Koype#Asynchronous_reply_context_hydration
#
jacky figured it'd be helpful to document that part
#
jacky might finish the filesystem part
#
jacky mainly want to explore how to capture my entries
#
jacky like I have a `mf2.json` and `index.html` in a folder per entry
#
jacky but I have to stitch in `index.html` into mf2.json and do some expanding (syndication results, etc)
#
jacky might see if I can use some sort of composite form of this (Rust has TOML)
#
[schmarty] jacky++ very cool! i am curious what the phrase "authority of a URL" refers to?
#
Loqi jacky has 26 karma in this channel over the last year (67 in all channels)
#
aaronpk i'm assuming "authority" as in https://tantek.com/2011/238/b1/many-ways-slice-url-name-pieces
#
[schmarty] hahaha oh i had forgotten that post 🙈
#
Loqi awesome
#
Loqi [Tantek Çelik] How many ways can you slice a URL and name the pieces?
#
aaronpk wow that post is 10 years old, i wonder how much has changed
#
[schmarty] aha, you assumed correctly, aaronpk! incredible wiki updating speed jacky! 😂
#
jacky in this case (domain/host name + port) so like https://twitter.com/i/timeline would be 'twitter.com:443'
#
jacky lol I figured I'd update there before answering here :)
#
aaronpk jacky: i'm curious if you've found any cases where the port number matters in practice
#
jacky none yet - I figured it'd be more specific that just the domain
#
jacky nor have I come across pages not served from 80 or 443 yet
#
jacky I know I had a use-case but I forget
#
@toyinariyo66 ↩️ I guess the main part I find confusing is setting it up with GitHub Pages as there are loads of plugins like the WebMentions plugin and Syndication Links plugin there's only a certain amount of themes. How do you set up Bridgy to crosspost Twitter and GitHub Pages? (twitter.com/_/status/1511365828125544450)
#
@toyinariyo66 ↩️ Thank you, this is really helpful. I'm definitely going to join the chat when I'm free so I can try and set up WebMentions/Bridgy on my WordPress site. (twitter.com/_/status/1511374313089384448)
#
[manton] I wonder why people think they should migrate from Twitter all the way to GitHub Pages. That’s essentially going from least complicated to most complicated.
#
aaronpk for real
#
aaronpk honestly i wish github pages and static site generators didn't have so much mindshare. they are not simple and often not the best option for anything besides infrequently changing websites
#
[manton] Maybe it’s something about owning your content = “own the files on GitHub”? It’s weird to me that people would jump straight to that.
#
[manton] Yeah, GitHub-based static sites definitely have their benefits but integrating Micropub and Webmention will be a stumbling block for most people.
barnaby joined the channel
#
aaronpk basically the challenge comes in trying to use a static site as a dynamic site
#
[manton] Exactly. So if someone is going to hit that roadblock, might as well focus on easy posting and custom domain name first, rather than get derailed right away.
#
jacky aaronpk: lol
#
jacky their reasoning was to have a free backup of sorts
#
jacky which, is valid, but now makes me still wanting to make backups easy
#
jacky well
#
aaronpk it takes a very long time before a sql dump to text files becomes impractical
tbbrown joined the channel
#
jacky indeed
#
jacky this does inspire me to write up a story for backing things up
#
[aciccarello] Yeah, I think jacky is right about free being a key feature that makes people jump to GH pages
#
aaronpk free in dollars, but not free in time
#
[aciccarello] 💯
#
jacky time is the true currency
#
[aciccarello] Definitely takes a change in mindset too to understand static sites which is not very beginner friendly.
#
[aciccarello] Conceptually static is simple. But not in terms of what people currently expect from a website.
barnaby joined the channel
#
jacky this is what the page for showing webmentions looks like https://imgur.com/a/6gEXhfA
#
jacky for /Lighthouse
#
jacky an implementation of https://github.com/indieweb/webmention-ecosystem/issues/2 and https://github.com/indieweb/webmention-ecosystem/issues/1
#
Loqi [jalcine] #2 Defining the Retrieval Source for Webmentions
#
jacky need to find better copy for this page
#
[aciccarello] Nice, I'm starting to think about writing a webmention plugin for indiekit which might end up similar.
mossymaker[d] joined the channel
#
jacky oh you use indiekit?
#
jacky I've been using about what that looks like?
#
jacky *like
#
[aciccarello] I'm starting to get into it
#
[aciccarello] I'm working on getting a micropub endpoint set up.
#
[aciccarello] It looks really fresh design wise.
#
[aciccarello] I'm hoping to be able to contribute to the project a little too.
#
jacky gotcha!
#
sknebel what is indiekit?
#
Loqi Indiekit (GitHub repo) is “a small but powerful server that acts as the go-between your website and the wider independent web” https://indieweb.org/Indiekit
#
sknebel does that page seem up-to-date?
#
jacky it looks like it tbh
#
[aciccarello] Yeah, I'd say so
#
sknebel nice
mro joined the channel
#
jacky I really appreciate https://getindiekit.com/decisions/
mro joined the channel
#
@IreadTours Join me today for a feature on Ten Thousand I Love Yous by @LisaSlabach. Check out the massive #Giveaway, Author Guest Post, Interview, & #BookTrailer.#WomensFiction #FindYourself #Divorce #WritingCommunity #BookTwitter @iReadBookTours #Romance #IndieAuth https://ginaraemitchell.com/ten-thousand-i-love-yous/?utm_source=twitter&utm_medium=social&utm_campaign=social-pug (twitter.com/_/status/1511384152641355777)
#
jacky how is that being picked up?
#
[schmarty] hashtag IndieAuth
#
jacky oh the #IndieAuth
#
jacky lol
#
jacky I guess that's like independent author
#
jacky collison--
#
Loqi collison has -1 karma over the last year
#
[schmarty] yep. that was way more common a few years ago, tbh.
#
aaronpk i think i added a keyword to ignore they were often using which seemed to keep most of it out
#
[schmarty] that sounds right aaronpk++
#
Loqi aaronpk has 35 karma in this channel over the last year (142 in all channels)
adstew joined the channel
#
adstew regarding static sites: yeah i'm finding out that presents its own set of challenges over something like wordpress lol
tbbrown joined the channel
#
adstew but that's the fun part for me since my goal with setting up my site was to learn how to implement things myself
mro, mro_, Mr_Imposter[d] and [chrisaldrich] joined the channel
#
[KevinMarks] Static works ok for articles, less so for notes.
klez and mro joined the channel
#
[tantek]1 static files make people feel safe (about their storage, content etc.) and are widely understandable (every computer has files), whereas there is nothing "safe" feeling about setting up/using/migrating a database
#
[tantek]1 so when people leave Twitter because they have a loss of trust & safety, they are motivated to solve *those* problems/feelings: trust & safety. static files feel VERY trustworthy & safe
#
adstew obviously they should use sqlite so that they can have a file AND a database lol
#
jacky lol
#
[KevinMarks] https://twitter.com/bascule/status/1511094316994809863?s=20&t=f3P1qHmvKgeKZiuxVjWJPg
#
@bascule Xanadu, Ted Nelson's pre-WWW hypertext system, had a lot of interesting ideas. But the design is somewhat inscrutable, with jargon so thick it may as well be "take the dingle bop and push it through the grumbo, where the fleeb is rubbed against it" https://sentido-labs.com/en/library/201904240732/Xanadu%20Hypertext%20Documents.html https://pbs.twimg.com/media/FPh70EUX0AYrqx9.jpg (twitter.com/_/status/1511094316994809863)
Seirdy, yequari, sp1ff`, tetov-irc and barnaby joined the channel
#
[schmarty] tantek++ a need for underlying simplicity was a big motivator as I built my heap of a website.
#
Loqi tantek has 25 karma in this channel over the last year (77 in all channels)
#
[schmarty] Building interactivity around a static site generator feels like a heavy lift at times but it pretty much always fails well. Maybe I break something by and some update or new post doesn't show up on my site. Even in the worst case my site is still up and available!
#
[tantek]1 People underestimate the feeling of safety that comes with the forgiveness of “even in the worst case my site is still up and available”
#
[KevinMarks] Really deep dive into how the M1 CPUs based on measurement, testing and patents by Maynard Handley (who I know from Apple ~ 20 years ago when he was doing deep optimisation work and suggesting new CPU instructions to help it) https://github.com/name99-org/AArch64-Explore