Re: Clientwide Webmentions

💬 Clientside Webmentions

I searched Webmention implementations on GitHub and found Max Stoiber’s impl. It fetches a simple count of webmentions, and then paginated full text responses. I figured I should try adapting that.

Ah, I hadn’t seen this particular clientside implementation before. Cool to see someone building something with React. Also that gives me some ideas for things I could add to webmention.js, namely doing a paginated series of async queries instead of just setting the page size to be really large.

(Not that I’ve ever seen any blog post ever have enough mentions for this to make a difference, but we’ll get there someday!)

Re: Proposal for #IndieWeb website swap day/week/event

💬 Re: Proposal for #IndieWeb website swap day/week/event

IndieWeb! What if we suddenly… swap sites?

No, not actual sites with actual content, but site runtimes. Set up a subdomain on our server and then set up someone else’s software on that subdomain and try to blog on it for a certain timeframe, like a week maybe?

This will allow us to test each other’s sites, including their onboarding procedures and UX. Also it may be fun!

I love this idea, mostly because I’m all in favor of getting more eyes on Publ (especially improving its documentation!), and I’d also love to see how other people are solving the various problems with web publishing.

Re: Intelligent webmention sending

💬 Re: Intelligent webmention sending

FWIW, Pushl also handles incremental webmention updates, including support for removals of targets and deletions of articles. Right now the only thing I feel like it’s missing for most IndieWeb folks is h-feed support, which should be fairly easy to add (I just need to get around to it).

It also could do with some better retry logic (namely saving failures in a queue for next time), which is also On The List.

Re: Hello, Indieweb!

💬 Re: Hello, Indieweb!

Hey there Matt, welcome aboard.

I also manually send webmentions for each link in an article, but it would be nice to make a script to do that hard work for me. Additionally, I should make accounts on more sites (like Twitter) and join a service like to sync my content between them. I would also like to experiment with owning my issues and pull requests, but that’s a thought for another day.

In addition to webmention.js which Jamie already, uh, mentioned, I’ve also written a tool called Pushl which automates the sending of webmentions. I have it set to run as part of my post-publish git hook on my site’s repository, as well as in a cron job that runs every 15 minutes. It requires an RSS or Atom feed at present (adding h-feed support is on my TODO list) but it only uses the feed for post discovery, rather than for target discovery, so it doesn’t need to be a full-content feed or whatever. It also maintains state so it can handle deletes and edits and so on.

There’s a few other automated Webmention sending things out there, like a lot of IndieWeb folks use which can be used in combination with IFTTT, although I can’t speak to the reliability of that solution.

Re: Auto-XYZ

💬 Re: Auto-XYZ

It looks like there are some percent-20 characters I need to clean up and I should try to show your posts in chronological order—so this has already been great for catching problems.

Yep, I’d noticed you weren’t handling spaces or dt-published. Glad you caught that too. :)

One thing to keep in mind is that your posts will really only show up under the first tag in the list.

Makes sense. What happens if I change what the primary tag is, and then re-ping?

Anyway, great work on Publ. It’s cool to see what you’re doing with logins—love the idea of IndieAuth on

Just to clarify, Publ can’t run on as far as I know (it needs the ability to run WSGI apps); my mention of was specifically about people having their IndieWeb profile live on there but delegating IndieAuth to an external endpoint (e.g. IndieLogin) – in theory a user could delegate to an external endpoint which spoofs a different identity (such as AnyAuth), which is an issue with the URL validation step of the IndieAuth spec.

Authl’s URL validation already adopts my proposed change, where the validated URL must be more specific than the requested one (for example, can become or, but it can’t become or

Re: Magic auth catch-22

💬 Re: Magic auth catch-22

Ahh—ok. Makes sense. I’m not doing this in my reader—I don’t want to risk rel=“self” being wrong. Is there a compelling reason to do this? I mean if I’m able to fetch the feed, why risk it?

Because sometimes sites migrate to new platforms or new domains and RSS feeds change location, and this helps to avoid linkrot or unpleasant surprises with realizing that you’ve missed a few years of updates because the feed quietly went missing.

Incidentally, I have implemented AutoAuth on this site. Gonna make a more formal announcement shortly.