There has been a major furore in recent days on social media about the rise in AI citations that have been hallucinated. These citations appear in manuscripts sent out for peer review, and even in the published literature.

This year we had a familiar argument in my household about when the Christmas tree should go up. My other half insists that it should be a late arrival, just a few days before Christmas ideally, whereas I would prefer to have the month of December as a lead up and infused with festive spirit, etc. In the end, because my mother was coming to visit, we decided that it would go up around the 14th or so of December.

As is my custom, I am writing, for my own historical logging, to show what I read and wrote this year. This year I have been working at Knowledge Commons, which has been a great environment with lovely people and a worthwhile project. It has meant that much of my time has been consumed with technical matters as is proper, but I did still manage to get up early in the morning and write.

Today I had a weird problem. Our BuddyPress activity stream comments were just not working. When you clicked the reply icon, then typed a response and pressed “post”, it simply grayed out and never succeeded. I could see, in the logs, that there was a 403 Forbidden error being thrown.

Another annoying error that you can get, during a docker build, that basically does not explain what’s going on is something like:

I spent the morning bashing my head against a brick wall, trying to sort out a problem with my Lando install. This worked on Friday, but by Monday was misbehaving. I hadn’t touched the codebase, but every time I hit the primary URL, I got: “404 page not found”.

I have a script (a custom static site generator) that produces the output at https://ticitaci.com – a page for the record label on which I have released music (and that I really, really love).

I have been playing around with [Pangolin](https://github.com/fosrl/pangolin), a really nice management system for exposing internal services over HTTPS. However, I found that its internal wireguard networking does not play nicely if you already have another wireguard system, like tailscale, on the box. The solution was actually simple, but has a potential tripwire.

I was having a pretty good week last week, until we got to the closing minutes of play. At that point, I learned that Amsterdam University Press (AUP) had been acquired by the for-profit corporate publishing behemoth Taylor & Francis. This is not really a surprise in some ways. AUP had been transformed into a private, for-profit enterprise in 2019.

Today, I have been battling a frustrating bug. In the latest versions of Chrome and Edge, users cannot highlight text in Full Site Editor or Post/Page Editor in WordPress (at Knowledge Commons. This turned out to be a complete nightmare to fix.

This morning, having been re-reading and thinking extensively about Moore, Samuel, ‘A Genealogy of Open Access: Negotiations between Openness and Access to Research’, Revue Française Des Sciences de l’information et de La Communication, no. 11 (2017), https://doi.org/10.4000/rfsic.3220 but also the awful news in Tim Sherratt, ‘Update on Trove Data Access and My Suspended API Keys’, Tim Sherratt – Sharing Recent Updates and Work-in-Progress, 2025