In reply to https://jan.boddez.net/notes/8970670852.
A daily cron job on my severely underpowered media server does the rest. Also, I’m currently using a hardcoded “bearer token” to access the not-so-public list, but there’s really no reason to not expose an actual (h-)feed for these posts.
Bookmarked https://jamesg.blog/2020/09/15/privacy-social-media-and-this-website.html.
Feels ironic to be bookmarking this, on my site. That said, there’s a reason I’ve created separate feeds for separate post types, and I do try to stay off the mainstream social networks.
I implemented support for bookmarks and likes and replies because these are commonplace on the IndieWeb. I understand why. Many people still use social media and use their website as a method of syndication. This is not me. I left social media [earlier this year].
This way of sharing (properly curated) bookmarks probably is the way to go. ’Til I have time, though, I’m going to have to do it the quick (and dirty) way.
I like [Ana Rodriguez’] approach to bookmarks. She writes one post every month with all of the articles she has bookmarked.
Quickly created this WordPress plugin that basically introduces a short-form CPT—well, two, but you don’t necessarily have to use ‘em both—and maps Micropub entries to it. That’s it. How’s this relate to Post Kinds? It doesn’t. I like how it keeps WP Admin (and permalink URLs) really neat and all, but there’s no meta boxes, and no markup gets generated for you. (Want to add, e.g., u-in-reply-to in there? Gonna have to do that by hand.) https://github.com/janboddez/indieweb-custom-post-types
Anyway, if you’ve got both a h-feed and an RSS feed—you’ve probably got more than one, which is great—for pretty much the same items, and they’re both reasonably error-free and understandable, I’m gonna follow the h-feed. (There’s occasional errors in my microformatted HTML, too, which I’ll then rectify, and so on. It’s sort of unavoidable and no biggie.)
Bookmarked https://davidyat.es/2020/09/05/fraidycat/.
I usually prefer to read articles on their original websites – partly so that they don’t all blend together like Medium sludge, and partly because syndicated content has an uncomfortably high probability of losing formatting and even chunks of text.
Why I start most—but not all—replies with “In reply to,” you ask? So that the few folks following me via RSS would know what the heck I’m going on about.
Just set up a #micropub media endpoint, and figured out how to get Indigenous to always use it. Up next: set up the back end so that a single photo uploaded thru a Micropub post also becomes that post’s Featured Image.
Bookmarked https://invidio.us/watch?v=X3SrZuH00GQ.
Nothing in this world is certain, except the death of social media sites and taxes.
In reply to https://jan.boddez.net/notes/9947c5da17.
Here’s something I hadn’t seen in a while: a published blog post of which the (RSS) date is a couple months into the future.
Looks like Yarns is getting syndicated URLs wrong, and shows all u-syndication URLs on an h-feed page rather than just the one(s) for a specific h-entry.
Building a social reader that’s sort of prepared to take on all sorts of edge cases and plain faulty feeds … is hard. (Kind of limited by what X-Ray, which handles the actual parsing, spits out, too. No need to reinvent that wheel. Just looking for a pleasant reading experience.)
So, if you’re adding a p-name class to your microformatted note’s e-content, feed readers, it seems, are going to consider it an article.
Suddenly seeing more than one broken image named “Array” in my “h-feed” timeline. Looks like a WordPress (plugin) thingy.
In reply to https://jan.boddez.net/notes/561d98de2a.
Now syncing channel_source pivot column values across channels, which isn’t perfect but works. I’d prefer a source_user pivot table instead of having everything tied to channels first (and use channels as taxonomies, really). But that would be a major change, and baby steps and all.
In reply to https://jan.boddez.net/notes/d66f3eab22.
Also, while I know the idea is to manage all of this from within a client app, I just spent a Saturday afternoon making it darn easy to copy and move sources from and to different Aperture channels, because that’s were I actually handle these things. (Because my client of choice—which I was thinking of somehow integrating into the service instead—just isn’t there, yet.)