What are you doing this weekend?
18 points by caius
18 points by caius
Feel free to tell what you plan on doing this weekend and even ask for help or feedback.
Please keep in mind it’s more than OK to do nothing at all too!
Travel prep.
During a very long and fun architecture discussion between myself, the CTO, and one of our Brazil devs, I made an offhand comment to the CTO in a DM how much nicer it would be to just get us all in front of a whiteboard for a week.
2h later he was asking us to book flights to Mexico City so we could meet up at a hotel with whiteboards and figure everything out for a week.
Wild fuckin’ ride, guys.
I'm hoping to find the time to sit down and finally go through ziglings.
I've been feeling like many of the newer languages coming out either have meh communities or feel too much like work, so I've felt like I don't really have something fun to program in when I want to do stuff casually. I watched the JetBrains interview of Andrew Kelley recently and it influenced me to look closer at Zig 'cause it sounds quite healthy.
I've never done ziglings, but have heard great things. I picked up Zig two years ago by going through the cryptopals challenges. I ultimately didn't finish cryptopals, but I had a lot of fun, and felt quite productive in Zig after just a couple of weeks of coding in it for a few hours a day.
I am now getting back into Zig, also similarly inspired by Andrew Kelley's interview. I agree that there is something very fresh and inspiring about Zig and its community.
If you do find the time, I hope you have fun with it, too!
Pathfinder 2e campaign with the group, work on my Portland-specific Google Maps replacement, and hide from the heat (94 degrees!) while reading China Mieville's "The City & The City".
My morning brain read this as playing PF2e with a Portland map which sounds pretty baller haha.
Are you DM or player?
Honestly, that sounds like a fun one-shot! Drop the players in the city and turn all the Portland cliches/stereotypes up to 11 in the story. They all know the city super well, so they'd probably really enjoy that.
In this campaign I'm a player. I DM D&D 5e for a mid-week game though. You?
Currently playing in 3 different games weekly and occasionally DMing one-shots. Also working on my own TTRPG system while patiently waiting my turn to DM haha.
Games are 5e, 2e (For Gold & Glory), and Wildsea
Nice! If your system ever needs a play tester, feel free to hmu.
I’m still working on it but the system rules are freely available here:
https://ettes-eternal.3digit.dev
I’m still actively building the classes right now. I had them all defined up to level 7 but as I’ve been fleshing them out further I’ve been kinda doing a full revamp of them too. So the first 3 are DONE but the rest are WIP.
EDIT: the name is a placeholder. Names are hard.
ooo mind sharing more about your Portland-specific map tool? I'm based in Portland (OR) and have been vaguely thinking about the enshittification of Google Maps and what I actually want out of a maps app.
Trying to find a way to make some money on the side since I took a big pay cut this year. I'm an "in-the-guts" engineer, so I'm trying to develop more "product-level" thinking to be more valuable at $WORK. If I end up making or doing something on the side that only made a single extra dollar, that'd be a win and a confidence boost.
Weed cleanup outside and a new SGI Fuel power supply inside. Also trying to unfunk my Talos II after a Fedora upgrade fail.
Playing my first “solo” gig! Actually it’s 3/4 of my band, and a borrowed piano player, but my wife is normally the main singer and she’s not going to be there.
Seeing Massive Attack in a few hours! My last chance was 2015 in London, but I missed it as the company I worked on folded and I spent that time during an impromptu wake in a pub in Shoreditch with my old colleagues. Honestly regretting it a little bit as MA stopped having gigs for many years after that and I moved countries.
Tomorrow, recovering and coding/doodling something stupid, simple, hopefully fun, certainly and importantly - useless.
Volunteering at State of the Map US for OSM and working on understanding how to make Ansible less of a pain in the neck for my homelab.
I will say using an AI for this made a huge difference for me as I went the extra mile on everything I probably would have skipped if it were just me. Paid off last month when the machine died and I had to reprovision. Even worked out on a cheap Radxa board I bought in the interim to get by with a handful of adjustments.
Re-reading Zen and the Art of Motorcycle Maintenance and dreading the first day on the new team the upcoming Monday the 15th. The new team seems awesome, which activates my impostor syndrome.
Continue reading Designing Data-Intensive Applications (the new 2nd edition). I've been making an effort to improve my systems design knowledge as it becomes more and more important with AI handling a lot of the development.
Grinding away to get Compound Interest, our new game, ready for Steam Next Fest. The deadline is on Monday!
It's an absolute chore at this stage, but it's working out well, don't mind if I link: https://compoundinterestgame.com/
Last week-end before starting new $JOB. Not sure why but I am back to reading about OCaml. Looking back at the last six weeks off where I did mostly none of the planned goals. I guess I decompressed too hard after 5 years on my previous job.
Family time, laundries and picking some books for the 1h30 travel to the new office on Monday.
I'm making youtube videos mostly I think? I set out to work on that skillset for 30 days.
Fixing the auto connect installed for the mower deck on my lawn tractor...it makes tons of noise and the dealer clearly did not install it correctly. Fixing a couple issues on my Miata for auto inspection. Then, I'll be updating some Caddy and CoreDNS configs on the home lab to make a few things better and finishing a new Gitea deployment.
I hope to complete the rewhack of my van's house power system which will free up a fair amount of floor space. I'm new to camper van design and learned that components that are rarely accessed are best positioned vertically against walls so they're out of the way.
Diving deeper into the colour science! Trying to figure out a good enough gamut mapping technique that is not built on top of layered approximations (myriad of magic numbers all over the place). Lots of people were asking about C/C++ port of my colour space transforms written in Odin, so that's also on my tracker for the weekend.
Been building a declarative deployment platform, targeting internal tools https://github.com/openrundev/openrun. I added support over the last few months for service binding, that is configure a Postgres or MySQL service and then each new app can easily bind and get a unique schema/database within the main DB instance. Works on Mac, Linux, plan to boot up my Windows server this weekend to test it on Windows.
Techwise, figuring out how to get the Surface Pro 5 battery to show up under Void Linux and building a device-specific kernel for it. Otherwise, I'm going to be attending a 10 year reunion event of my high school class.
Trying to put together a talk for EuroBSD. If you know what I've been hacking on, you can probably guess the topic.
I developed an AI agent from scratch using DeepSeek V4 Flash to complete some simple tasks. I'll continue to improve it over the weekend.
Manually cleaning up some Claude-generated scripts that make it possible for me to do $DAYJOB on FreeBSD.
The aim was to see if it was feasible to run everything we normally run in Docker in Jails. This worked swimmingly, but the resultant code is a mess. Multi-line ternaries, half-page-long partially incorrect comment blocks, scary failure modes (deleting the mounted source directory from the host if unmount fails), etc. etc.
I was expecting that it'd be bad; I wasn't expecting that it would be this bad. Still I consider the experiment a success - and it means that (once manual cleanup is done) I can ditch my final "serious" workload from Linux.
Last-minute writing like it's college.
A few months ago I submitted my ionospheric modeling work as a poster to a scientific conference, with a thesis of "this approach is awesome because it allows you to incorporate multiple modalities of observations without having to write special code to merge them". The only problem is, after I submitted and was accepted, I ran my study and those weren't the results I got. If I ran the model with one kind of observations it returned perfectly great results, but if I gave it both kinds of observations the results didn't get better, they got worse (in fact, worse than the baseline I'm comparing to).
So all the time I should have spent writing my poster, I spent frantically tweaking and re-running things. And I actually managed to get acceptable results, but now I have like three days left to do the writing, get it reviewed, and get it sent to the printer. Without compromising my dayjob (which is something completely unrelated) or abandoning my family. Much.