vilhalmerI wonder how feasible it would be to write an algorithm for creating the urls deterministically from the original so you don't even need the mapping stored on the site?
[tantek]I'm now leaning towards why implementing (and using) an /original-of endpoint is better than doing discovery by asking the silo. Added a bit to the See Also / Why
[tantek]In general, the more we can enable IndieWeb site peer to peer IndieWeb site algorithms & protocols that do not require contacting a silo (e.g. surveillance, or silo death), the better
[tantek]I should consider brainstorming on the OPD algorithm accordingly, to only make it apply to if you already have the contents of a POSSE copy, and want to do discovery on those contents. Whereas if you only have a POSSE permalink, I think it's better to do inference on the username -> IndieWeb site (via nicknames cache), and then ask the IndieWeb site to lookup the /original-of
[tantek]OPD << Brainstorming: update the OPD algorithm: only do content based discovery if you already have the contents of a POSSE copy. If you only have a POSSE permalink, better to do inference on the username -> IndieWeb site (via [[nicknames cache]]), and then ask the IndieWeb site to lookup the /original-of. Consider also caching that result (POSSE permalink -> original) so you can re-use it in the future (e.g. network transience), and perhaps
Loqiok, I added "Brainstorming: update the OPD algorithm: only do content based discovery if you already have the contents of a POSSE copy. If you only have a POSSE permalink, better to do inference on the username -> IndieWeb site (via [[nicknames cache]]), and then ask the IndieWeb site to lookup the /original-of. Consider also caching that result (POSSE permalink -> original) so you can re-use it in the future (e.g. network transience), and perhaps" to the "See Also" section of /original-post-discoveryhttps://indieweb.org/wiki/index.php?diff=69967&oldid=69966
aaronpkwith websub it at least moves the burden onto the hub, which is good because the hub is a much simplier piece of technology compared to an entire blog engine, and also is generic/agnostic of the thing that is actually generating the content, so one hub can serve many different CMSs
[tantek]Right this is the flip side. It’s easy enough to say “this is why WebSub” but if I’m practice no one is consuming subs from hubs then it makes a weaker case
aaronpkah, if someone who follows me boosted a post of mine, and someone from that host was following that person, and then they favorited my post, it would send that to me
aaronpki've "blocked" the host on my end, which just means i throw out the messages before I do any work. i'm not sure if there is some way I can actually tell the host to "reverse block" me or something
[snarfed]aaronpk yeah i got those on bridgy fed too. i originally returned 501 Not Implemented, but they kept retrying and retrying, so i finally changed it to return 200
jackywebsub seems like more adjacent in terms of implementation / use to webmentions than microsub (though it does and can power real-time work for microsub)
[LewisCowles]Is pixelfed Open Source? I'd love to 👀 at what collection of things is capable of making a federated drive / images. It might help me do more and pontificate less for the known attachments / uploads as URL's
swentelthe idea is also to extend the app with more account types, e.g. AP connections with pixelfed/mastodon etc (but that part is still in brainstorming phase)
aaronpkevery time i go into the code that bothers me too but i'm always trying to solve something very small and specific and i never want to tackle that
swentelaaronpk, if you have code, feel free to put in a branch. Not sure about the timeline yet, I'm busy setting up projects and milestones now, and this one is def on my mind.
sebselit is a very nice app, have been writing in it for years. it lead me to Kirby which in turn lead me to IndieWeb. Nice that the connection to the IndieWeb is now so direct
LoqiiA Writer is a markdown-based writing application that supports the Micropub standard for publishing to numerous services, and specifically to WordPress servers, Medium, and Ghost https://indieweb.org/iawriter
[tantek]saw this go by and figured both that some devs here may be intersted in it, *and*, I figured some of you have webdev blogs that ought to be added (file a GH issue!) to this list: https://github.com/impressivewebs/frontend-feeds
[tantek]^^^ perhaps also a good reason to have a "webdev" or similar tag (with its own feed), as the blog post about the list does refer to excluding blogs with "Too many personal posts and other things that have nothing to do with web development, with no way to filter"
[KevinMarks], [aimee] and miklb_ joined the channel
ZegnatLol. You just made it click for me jacky! The only free honey I could think of is the one from all the YouTube sponsorships “Honey is the free browser extension that …”. I was having a horrible time figuring out how this was linked to the browser extension
[tantek]this is unintentionally prophetic: "Yet most web sites are architected in a way that require you to download the css for the entire site when you try to view one page. This is a broken model."
[tantek]Most web sites are architected in a way that require you to download the JS for the entire site when you try to view one page. This is a broken model.
[LewisCowles][tantek] they could write the cascading bits in SaSS/SCSS if they really cared about the problem as mixins. Then they get rid of dead code, and can namespace their nonsense.
[LewisCowles]I Feel like a lot of the deferred rendering / JS only sites basically take a quick-fix and architect around it for... What's it been 8-10 years? Before React it was jQuery.ajax or DHTML only menus
[LewisCowles]I spent over 2 hours trying to help someone pivot to a mildly less awful iteration of their code, only to panic at the end realising they were not using source control
[tantek]yikes. There are many aspects to maintaining CSS over time that I can relate to in this article, even "just" for one person maintaining their own CSS.
[tantek]one of the nice aspects of my old static blog was how literally every month I got to rethink/rewrite the CSS knowing it would only affect that month.
sknebelI think I've seen checks for basic rules like "check important elements stay visible/within the viewport", "check text doesn't get hidden by other elements"
[Katherine][tantek] ehhhhhh i took a look at that. but i just discovered that there's an integration for jekyll to post via email.. that might be easier to set up https://begemot.grin.io/about
[tantek]from my understanding, once you have Known setup, it's got LOTS of really good IndieWeb support, theme etc. without having to install anything else
[tantek]yeah I'd go ahead and just say hi in the #knownchat channel that you're setting it up on Dreamhost and does anyone else there have experience with that or tips?