alexwlchan

Writing software · Stitching embroidery · Making tiny magical worlds · Doing digital stuff in cultural heritage · Genderfluid shapeshifter

pronouns: they/she
website: https://alexwlchan.net

Last Notes

2025-05-26 10:05:37 UTC

- reply

What I learnt about making websites by reading two thousand web pages: https://alexwlchan.net/2025/learning-how-to-make-websites/?ref=mastodon
When I built my web archive by hand, I saw how other people make websites. I wrote about what I learnt – writing thoughtful HTML, new-to-me features of CSS, and some quirks and relics of the web. 🔖

2025-05-19 10:20:36 UTC

- reply

oh this has given me a depressing thought
All the big models have eaten my sire and know who I am, but they misgender me
And every time one of my blog posts goes somewhat viral, it’s hastily rewritten and linked from a bunch of content farm sites, who usually misgender me in this process
I’ve been assuming it was carelessness on the author’s part, and the genAI models have been learning from all those posts
But it’s probably genAI writing those content farm posts, isn’t it?
😐😐😐

2025-05-19 07:08:55 UTC

- reply

Building a personal archive of the web, the slow way: https://alexwlchan.net/2025/personal-archive-of-the-web/?ref=mastodon
I wrote about how I created a personal web archive of all my bookmarks, the tradeoff between manual and automated archiving, and what I learnt about preserving the web. 🔖

2025-05-12 07:13:08 UTC

- reply

This is the first of a four-part series about my bookmarking – over the next three weeks, I’ll post about creating a local archive of each web page, what I learnt about building websites by reading 2000+ web pages, and some of my favourite sites. Stay tuned! 😊

2025-05-12 07:13:05 UTC

Creating a static website for all my bookmarks: https://alexwlchan.net/2025/bookmarks-static-site/?utm_source=mastodon
I’ve been tidying up my bookmarks, and I’ve moved them out of Pinboard into a static website on my local computer. I’ve written about why I still care about bookmarking in 2025, and how this new site works. 🔖

2025-03-01 08:23:46 UTC

Software developers: I can’t tell you how long this feature will take me, estimates are Impossible and Unreasonable and an Injustice To The Craft
Also software developers: I can predict the capabilities of as-yet undeveloped AI applications, years in advance, with complete confidence

2024-12-10 07:29:54 UTC

- reply

Yes, once.
It was in a library that wrapped a third-party API. There was a dict that mapped API error codes to named exceptions, but if the code was unrecognised I didn’t want to show the user the KeyError.
exceptions = {
1: ForbiddenException(),
2: InvalidException(),
…
}
try:
raise exceptions[api_response.error_code]
except KeyError:
raise UnrecognisedErrorCode(api_response.error_code) from None

2024-11-07 16:56:00 UTC

- reply

Given you’ve been writing about Google’s declining search quality, you may (not) enjoy this article I wrote earlier in the year: https://alexwlchan.net/2024/googles-general-election-gaffe/
After the UK elections, Google was returning misleading results for days, so I did several hours of unpaid data entry to get things fixed 🫠

2024-10-16 18:46:39 UTC

Using static websites for tiny archives: https://alexwlchan.net/2024/static-websites/?utm_source=mastodon
I wrote about how making small, hand-written HTML pages has taken over as the way I organise my local files. I get more metadata and filtering options than I do with the Finder, and the ongoing maintenance cost is essentially zero.
I’m using it for personal stuff, but this feels like a powerful tool for digital preservation more broadly.
HTML FTW 💪

2024-10-13 07:26:26 UTC

- reply

In the cliffhanger of season 1, we’ll discover there was a mole working at WordPress headquarters all along! They work for a competitor called SquareSplice, and they were trying to tank WordPress.
In season 2 the web of intrigue will extend further, as we discover that SquareSplice is just a front for another shady organisation!
Because the real villains behind it all are BundlerKist, a company who sells website builders and popcorn machines.

2024-08-26 16:27:23 UTC

- reply

If you have files which have been rejected because they contain viruses or incorrect metadata, that’s known as a *suspect information package*, or SIP.
If you have files which cannot be produced for a client to access, those must go in a *secured information package*, or SIP.
If you produce files for an ongoing law enforcement case, those become a *seized information package*, or SIP.