No context conference slide.
Reminder for later... Paths are predicable enough in my setup that I can improve the netlify config to apply headers only to HTML by batching routes with wildcards.
Morning mist over the Sussex Countryside.
Next thing to do: add handing for h-reviews so I can rate mince pies.
Work in progress.
Testing out syndication of a note from my blog.
Between Lewes and Newhaven by train.
An interesting local mural.
Another nightmarish thing my partner bought.
Versioning edge case discovered on the Node.js website.
I'm not a huge webpack fan, but I'm wondering if I might change my own mind my porting my static site generator to it. As it stands my generator tries to be as efficient as it can be, but only for a single shot compilation. It doesn't cache artefacts. It looks like webpack, with a custom loader and plugins, might provide the necessary mechanics to make incremental builds while writing a blog post much faster.
Originally I was going to do a sort of dark-battenberg theme for dark mode on my blog, but now I'm rethinking. Since more and more devices are beginning to use OLED screens, I want dark mode to mean "mostly completely black" to save some energy on those devices.
I wrote a while back about how promises and event emitters in Node.js can play badly together. Since then, Node and JS have changed enough that I should write a follow up post.
I have a real completionist streak to me, which is a lot of fun as it interacts with IndieWeb stuff like microformats. The next thing on my radar is WebSub. I'm keen to implement both the hub and publisher in one as a glitch.
Next up, I’m considering micropub for posting full article entries. Hosting an editor and managing storage for drafts is a bigger task, but I can increment on that. Posts would be in markdown though. I wonder if that’s considered bad...
Idea for later: I need to markup each publication in the publications page as an h-cite. http://microformats.org/wiki/h-cite
I like the idea of a links page for things I've read and think others might enjoy. I'll probably build one soon.
I think I'll work on putting some of the sharing stuff at the bottom of each post into a <details> element at the next homebrew website club. That might mean splitting the webmention stuff up from sharing links. I'm not sure yet...
I think it's time I started writing some tests for the static site generator.
It’s possible using sed, jq, and date to read timestamps from scheduled posts and git mv and push them. This would be superior to the current Contents API approach because the content of the post will be properly attributed. It’s also lighter and agnostic with respect to git hosting. The command to get a time stamp will be something like: `sed -n ‘/^---$/,/^---$/p’ < post.md | sed ‘1d;$d’ | jq ‘.datetime’ | date -f -`
Now I’ve got a handle on the GitHub contents API, it’s tempting to build an editor into my blog. Could be a rabbit hole though...
Idea: Set up a web hook for webmention.io which posts to a glitch which in turn opens an issue on the GitHub repository for this blog.
Another possible set of articles I could write is on tips for writing kubernetes configuration files. I've accumulated a number of them over the last few years.
I might make the repository for this blog open soon. It's embarrassing in places, but I'd still like to show what I've done and compare notes...
Homebrew Website Club is this Thursday, so I need to either think of a blog post I want to write, or think of an enhancement to my blog. I’d quite like to integrate an editor, but that’s a big task. Something smaller, or an enhancement to an existing feature might have more impact...
Tickets booked for the next trip to Japan. Our first with the little one. I can't wait for his grandparents to meet him!
I can think of no better tribute (albeit late) than to wish the World Wide Web a happy 30th birthday on my own weird little personal site.
I have a glitch which tweets out new blog entries. Right now it just looks at changes in the sitemap. Reminder to improve it to include tags. The selector to use is `document.querySelectorAll('[rel="tag"]')`. https://tweet-new-blog-posts.glitch.me/
Reminder to myself to write a post about my recent contribution to Node.js.
All micropub client implementations appear to be a bit lacking (none seem to support a media endpoint very well). I'm considering making my own client and embedding it in my blog...
Testing alternative micropub server implementation.
I've come to refer to my preferrred way of creating microservices as "calving". Start with a monolith, and move bits into microservices where the coupling is weak (like a big iceberg calving off a smaller iceberg).
One of the cool things about micropub is that I can use a third party client really easily. This post was created with micropublish.net
Now I need to reconsider my landing page. It'd be nice to have the notes there, and some information about me.
I implemented #indieweb notes on my blog using a Netlify function and the GitHub content API. I'm using Keith J Grant's omnibear browser extension (in FireFox) to post. Not much to see yet, and it's a very basic implementation, but not bad for a couple of hours of work.
Notes are live! 👍
Testing again, 456.