This is so cool. I love this so much. I read the article on my RSS, and then immediately hopped over to HN just so I could upvote.
> Fortunately though, over the course of Bauble’s development, I had produced a comprehensive suite of test scripts with reference images that demonstrated all of the edge cases and problems that I had faced and already fixed and… No, of course not. I can’t even type that with straight fingers. There were no tests.
Intensely relatable.
What RSS reader do you use, may I ask? Mine (readrops for Android) doesn't render the images. It is probably quicker to just review the XML but I'm committed to this comment now.
Nor did mine, interestingly. I'm not entirely sure why, the HTML for the first image seems to render fine on its own. It's a very long <picture> tag, seemingly optimised to return different sizes at different resolutions. It looks like something generated by a web framework. I still hand-write my <img> tags like it's 1999, so I'll leave it to the frontend wizards to explain what the problem here is.
Re my choice of reader, I host FreshRSS[0] on a home server, using the official Docker image.[1] It comes with pretty good in-built webpage change tracking too, for websites that refuse to offer RSS. I don't feel confident enough to expose it to the Internet, though I imagine you could use something like Tailscale to tunnel home securely for it.
[0] https://freshrss.org/; https://github.com/FreshRSS/FreshRSS/
Funny, I'm using FreshRSS as my feed aggregator as well, but then reading it using the android app. One of us should file a bug.
That is amusing! I'd think that your Android app would be using a different rendering engine to my desktop browser though? There must be something non-standard about those images.