btremDoes anyone do self-webmentions? I am trying to keep track of my ferments, with one post for when I make a ferment, and a second post for when I open it, as an update/review. So the update/review would be an h-review of the first post, an h-recipe.
btremBut linking them is tedious. I originally had to figure out the url I used. Seems easy, but I often make the same (or almost the same) ferment. So I have lots of urls with a slug of "radishes" or "pickles".
btremI have taken to adding a database-like id number for each post which must be unique. It does not appear in the url, but I can put the number on the jar of pickles or what have you, so I can then use that number for my review post.
btremBut even this can be quite tedious. I dunno. I think maybe a static site generator is not the way to go, and I need to resort to a database. But if anyone does lots of self-webmentions -- maybe post updates of some sort -- I'd be curious how you handle it.
btremLike when I want to create a new post, I have to run a dev copy of the site to see what my last post id is, so as to avoid creating a duplicate id. Then I put that number on the jar label. Then I need to remember that id when I leave my kitchen to create the post.
aaronpki use quill to post, and i have a bookmarklet in my browser that opens up quill with the in-reply-to field populated with whatever page i was on
btremIronically, taking photos of my ferment is part of what can be tedious, because I have to determine what the next number is for a post. To do that, I have to start up my dev server to see that most recent post. Then I write that number on an adhesive label along with a title, attach it to the jar, and take a picture of the jar.
btremI have so many recipes that they were clogging up my site's home page. So I'm in the process of creating a new site that will end up in a new domain (maybe a subdomain of my personal site). And I'm wondering if an SSG is the right tool. I could instead make a recipes database, in sqlite perhaps. And then build the recipe site using data from queries.
[schmarty]I have a little web service set up that accepts webhooks from my git hosting to trigger a rebuild. It pulls down the latest, runs Hugo, puts the updated files up, sends webmentions for new or updated pages. https://git.schmarty.net/schmarty/ssg-hook-publisher-php
aaronpkYou don't need a database to manage a static site with micropub, you need to make the micropub endpoint automate all the steps you currently do manually
[aciccarello]btrem, I use Indiekit as my micropub server for my static site. It's a little node application that can create files in GitHub or other repos.
LoqiIndiekit (GitHub repo) is “the little Node.js server with all the pieces needed to own your content and share it with the wider independent web” https://indieweb.org/Indiekit
gRegorLove_, gRegorLove__ and CRISPR joined the channel
btremOk, some links to look at, and some thinking to do. It's not just micropub. It's making the right connection between recipe and rating. But that might not be any easier with a db, since (part of) the problem is that I have many similar posts. So I have lots of pickles posts, and lots of radishes posts, etc.
Loqi[preview] [Tim Bray] A few annoying problems with the #Fediverse could be solved if there were a new URI scheme with the right semantics. I notice that Bluesky has already provisionally-registered the "at:" URI scheme for similar purposes. So, I sketched out a design for...
Loqi[preview] [Johannes Ernst] @timbray Would be nice to enumerate the problems and related problems to see whether this is the right "slice" of them to solve together.
AramZS, barnaby, CRISPR, [jeremycherfas] and geoffo joined the channel
[manton][KevinMarks] [tantek] The fedi:// proposal reminds me of feed:// for RSS, which never seemed to take off. Basically solving exactly the same problem.
[manton]Yeah. Also it’s hard to control on iOS in the domain handling too… If I want Ivory on my phone to stop intercepting Mastodon URLs, there’s no obvious way to do that.
[manton][snarfed] Just got a reply from someone on Bluesky and it showed up in http://Micro.blog via Bridgy with their username as their Bluesky hostname. So cool! Great work on that.
[tantek][manton] [snarfed] looks like some Meta-employee Threads accounts are supporting following now, e.g. this shows a Follow button! (replace the domain with your instance domain otherwise it redirects to his Threads profile) https://indieweb.social/@mosseri@threads.net
[manton]I think the Threads support is hardcoded to specific Mastodon servers somehow. I still can’t get it to work more generally from curl or http://Micro.blog.
[tantek]So, confirmed, Threads federation outbound is turned on for at least one Threads user who is a Meta employee, and is supported/allowed by at least *some* Mastodon servers (w3c.social, indieweb.social tested so far) and BridgyFed
capjamesgThe predictions are ordered by probabilities; the higher the prediction, the more likely it is I would type that word given the letters you provide.
[tantek]Oooh this is great. One annoying bug so far, if I click earlier in the text to edit it, the cursor gets repositioned immediately to the end of the text 😂
[tantek]oh correction, I can use arrow keys to reposition but not the mouse pointer?!? (i.e. clicking with the mouse pointer into a prior word causes the insertion cursor to get slammed to the end of the text)
gRegor[snarfed], I see some "doesn't look like a domain" and "is not a domain or usable home page URL" in BF logs when I federate, for a few followers. Should I file an issue?
gRegorDisregard that part, it does look like it worked. One instance did show "has no delivery target" though, and that one didn't receive the post, based on the public search at least.
[snarfed]angelo no, BF already does authorized fetch (ie includes HTTP sigs) with all AP object fetches, since a significant chunk of the native fediverse already requires them
[manton]To expand on the earlier Threads compatibility discussion, I realized that Threads requires HTTP signatures but it doesn’t return a 401 if the request isn’t signed. It just returns HTML. My code wasn’t signing unless I got a 401 back first.
btrem Ironically, taking photos of my ferment is part of what can be tedious, because I have to determine what the next number is for a post. To do that, I have to start up my dev server to see that most recent post. Then I write that number on an adhesive label along with a title, attach it to the jar, and take a picture of the jar.
gRegor https://indieweb.org/User:Martymcgui.re might give some ideas
[schmarty] I have a little web service set up that accepts webhooks from my git hosting to trigger a rebuild. It pulls down the latest, runs Hugo, puts the updated files up, sends webmentions for new or updated pages. https://git.schmarty.net/schmarty/ssg-hook-publisher-php
[aciccarello] btrem, I use Indiekit as my micropub server for my static site. It's a little node application that can create files in GitHub or other repos.
Loqi Indiekit (GitHub repo) is “the little Node.js server with all the pieces needed to own your content and share it with the wider independent web” https://indieweb.org/Indiekit
[KevinMarks] Missing the point in a big way? https://cosocial.ca/@timbray/111575034761844339
capjamesg [tantek] AutoWrite: https://jamesg.blog/assets/autowrite.mov
[tantek] [KevinMarks] indeed. That URL scheme proposal is a good example of what NOT to do. johannes's simple request was ignored: https://social.coop/@J12t/111575260878262958 j12t++
[manton] [snarfed] Just got a reply from someone on Bluesky and it showed up in http://Micro.blog via Bridgy with their username as their Bluesky hostname. So cool! Great work on that.
rubenwardy how do I type "bob" without it autocompleting to "bobbing"
angelo inside that i find https://www.threads.net/ap/users/0xjessel/ which i then request via HTTP Sig in the same fashion as i do mastodon