I have now been at GoCardless for a month.
Which is kind of surprising really, it doesn’t feel that long, but yes, my calendar confirms that the 8th of March is a month ago. I think I’m settling in fairly well. There’s obviously still a lot about the domain that I don’t know, but that cloud of known unknowns is shrinking little by little. I’ve been spending my spare time1 looking through the codebase to figure out how everything fits together, and I spent most of last week refactoring one of the components we own.
This reminds me, I should check that “your first 30 / 60 / 90 days” thing I was sent about things to learn and do and see how I’m doing with that.
This week I read:
Gateway by Frederick Pohl
This is about a depressing future-Earth, where food is a slime grown on oil and coal because the world is so overpopulated conventional farming can’t keep up. It’s about an asteroid, full of alien spacecraft which can travel around the galaxy, and the hope that one of them will bring back some knowledge that can salvage the world. But mostly it’s about one man, who gambles everything he has on becoming a space explorer, and becomes traumatised by his experiences.
The narrative is split into two parts. The first part is following the main character, Robinette Broadhead, as he moves to Gateway and starts his career there. The second part, which is interleaved with the first part, is Robinette’s discussions about these experiences with his robotic psychiatrist many years later. It’s a great little story where both sides of the narrative build up to the climax together: in the one where we the readers see what happened on Robinette’s final mission that made him so rich, and in the other where he himself uncovers that repressed memory and confronts the root of his trauma.
The Book of Gaub from Lost Pages
This is an RPG sourcebook about weird magic and a mysterious entity called Gaub. The number seven is thematically important, with the Seven Fingers of the Hand of Gaub each having seven associated spells, a cabal of Gaub cultists having seven members, and various monsters, magical effects, and strange catastrophes incorporating seven somewhere: seven fingers, seven days, seven options, seven people…
It assumes an OSR-like system, but doesn’t have much in the way of concrete mechanics, so it looks like it’d be pretty easy to adapt to any other fantasy system.
This is exactly the sort of OSR content I prefer. A lot of OSR stuff, I feel, goes way to far towards the weird, in an attempt to stand out. But if everything is bizarre and trope-subverting, then it loses its impact. The Book of Gaub is close enough to your standard fantasy to fit into most games, but is also just slightly weird, where introducing it could significantly alter the tone of the game, and suggest that there is something deeper and darker than what is usually apparent.
And also the book itself is really nice looking, which always helps.
I made a few changes to my self-hosted set-up this week: I’ve stopped hosting git, started hosting a wiki, and made a couple of discoveries about systemd units.
Let’s start with the systemd units. I learned about the
BindPathslets you bind-mount something inside the service’s filesystem namespace, so it’s a mount only available to the service, and which won’t show up in the host’s mount table (eg, with
This is pretty neat, as some services are annoyingly inflexible and need files or directories to be in certain locations. In particular, for my erase-your-darlings NixOS setup, I had a whole unit just to bind-mount the Prometheus state directory onto my persistent volume. Now I don’t have to.
DynamicUserlets you allocate a random high-valued UID and GID for a unit on start, so it’s not running as any user which already exists. I used to use the
nobodyuser for that, but systemd started warning against that at some point, because it’s not as secure as it could be: there are files which can legitimately be owned by
nobody, and you probably don’t want random services to be able to write to them.
These options are both in the documentation, but there’s so much stuff in the systemd documentation that it’s kind of imposing to just read through. These two haven’t been the first options I’ve discovered after years of use, and surely won’t be the last.
systemd is a great piece of software. I was sceptical when it first began to rise to prominence, but those doubts are all long gone.
Next, git. I have stopped hosting gitea. It’s a perfectly fine piece of software, but it was not ideal for my purposes. It was annoyingly stateful, with some configuration done through environment variables and some through a configuration file inside a docker container. It was just slightly annoying having my private git repositories in one place (gitea) and my public git repositories in another place (github). And one time I managed to nuke its state and discovered my backup of its database was incomplete, so I had to start from scratch… and nothing of value was lost.
So I pushed all my private repositories to github, updated my backup script so it could handle private github repos, and dropped gitea.
Finally, a wiki. I’m now running wiki.js. I’ve resisted this for a long time, trying to believe that everything should be open by default, on this memo site. But I’ve come to realise that mixing personal-audience and public-audience writing doesn’t really work for me. Since this site is now almost entirely public-audience stuff, I’m just going to migrate private-audience stuff to the wiki. I’ve already done that with my recipes and found that it’s actually motivated me to add some more.
wiki.js itself is quite nice to use and run. And it appears to store everything, even media, in its database, so backing it up is a breeze too.
I like cooking and eating nice food, but I’m also very lazy and so left to my own devices will just make easy meals, snack, or order takeaways all the time.
So to combat that I started planning my meals some years ago. I’ve got a little 5-week calendar on my whiteboard and, at the start of the month, I’d work out what I wanted to make and eat. So I might decide to cook a big chilli in my slow cooker, which would make 10 portions, and I’d allocate those to days across the month. Then I’d think “well, I don’t really want two minced beef based meals back to back, so maybe I’ll make a curry for this, this, and this day…” and so on. I got quite a bit of variety in my meals, and also specifically allocated days where I’d try cooking something totally new. This had another benefit too: shopping trips became trivial to plan. I could just look at the calendar, see what was coming up for the next week or so, and buy the ingredients I was out of.
It worked very well!
But at some point over the past two years, I stopped doing the big monthly planning session. I began to plan only most of a month ahead, then only a couple of weeks, then just a week… it was so gradual I didn’t notice. But now that I have, the difference is stark.
So what, right? Most people don’t plan a month’s food ahead of time, and they do just fine.
Well, recall that I am very lazy.
I used to have a small repertoire of 5 or 6 meals I’d cycle through. Not a huge amount of variety, but enough to keep things interesting, especially if I did try a new meal a couple of times a month. But for the past couple of years I’ve switched more and more to just making the easiest one or two meals, and very rarely trying new things. If a meal needs some ingredient which is harder to get (for example, I’d need to go to an asian supermarket and there isn’t one nearby), I tend to put it off. And I’ve got back into the habit of ordering one or two takeaways a month, because the food I do make isn’t providing enough variety.
This needs to change. Fortunately, it should be pretty easy to fix now that I’ve recognised the problem: I just need to reinstate the monthly planning sessions. Then I know what I need to buy, I know what I have to look forward to, and I can specifically pick out a few days a month to try something new.
So I’ve put a recurring event in my calendar, for the first of the month, to plan meals. Let’s see how it goes.
- About Damn Times: A Campaign Event Table
- The Merchant Campaign
- Small Gods and Stone Soup: Deities made for Dungeon Crawling
Spare time at work, not spare time out of work.↩︎