The second post in my event sourcing series is about how to enable better query capabilities to find the needed event streams while still holding back on read models. Of course, with the reasoning why this is a good idea in the right context.
www.planetgeek.ch/2026/04/21/e...
F# Weekly #16, 2026 - .NET 11 Preview 3 & SwaggerProvider 4.0 beta #fsharp sergeytihon.com/2026/04/19/f...
Sorting the events.
Two taughts:
- comparing uuids is a bit more expensive
- we sometimes change the “application” (our version number) to fix data. That wouldn’t be possible with uuids
Otherwise, I think it would work.
cool stuff. I'll play a bit and give you feedback.
How long does it take for the agent to react to a new issue? Or do I have to trigger it somehow?
The start of my new blog post series about Event Sourcing. The first post starts with a simple projection-only approach: www.planetgeek.ch/2026/04/14/e...
You'll see why we don't use event numbers for ordering, but timestamps, and why this is our default approach.
#EventSourcing #dotnet #fsharp
Obviously, I spent some time on my #eMTB in the last 12 months (and the 12 before as well) 🙈
The Mega Drive game I believe has aged the worst.
I love that FSharp.Data is being worked on again #fsharp github.com/fsprojects/F...
I‘ll let it run as everyday and see whether I spot wrong colours. Thx
And there are also changes where the automated tests already cover everything, like computation changes in business logic with no different behaviour in the UI (except for numbers that change).
We have a different approach to limiting exploratory testing needs:
Our system is heavily modularised. More or less, every view and every action is its own slice. So when making changes to one slice, most or all others stay the same and don't need (re)testing.
Interesting. Do you know what the agent tests?
I still prefer a human to do exploratory testing because only a human tester sees the app as a human user would (contrast, overlap, etc.).
I would never ship a feature without first running an exploratory testing session. So many things can go wrong that we can't specify explicitly; only exploratory testing finds these.
And I want a dry run on difficult data migrations, too.
2/2
Lately, I see a lot of posts saying we need a robust suite of automated tests to ensure AI agents produce high-quality results.
Please keep in mind that one cannot test quality into a software product. Automated tests are a necessity, but they are by far not sufficient as a sole approach.
1/2
@schlenkr.bsky.social regarding @PXLclock: is the currently running Random Colour Wheel still the F# version, or the C# one?
I'm asking because there are some colour errors from time to time.
Grab the #pxlclock VS Code extension ("PXL Clock), code your own pixogram in #csharp, and ping me here: I'll record it on the real hardware and post the video.
Who makes the coolest one?
VS Code: marketplace.visualstudio.com/items?itemName=pxlclock.pxl-clock
#pxlclock #csharp #fsharp
F# Weekly #14, 2026 - Serde.FS Brings Compile-Time RPC to F# #fsharp sergeytihon.com/2026/04/05/f...
We started using event sourcing over ten years ago. One of the hardest lessons was that there is a kind of events that is not obvious at first but have a big impact on your system design. I call this kind secondary domain events.
www.planetgeek.ch/2026/04/01/t...
#EventSourcing
F# Weekly #13, 2026 - Deedle 4.0 is out! #fsharp sergeytihon.com/2026/03/29/f...
I would favour more efficient solutions, but none are in sight to me. Maybe once, when we have to pay for the real costs of training and running these agent farms, a correction will happen.
The picture Gen-AI created is, however, quite funny. 😄
3/3
From single LLM via single agent (repeat until succeeded) to multi-agent setups (giving themselves feedback to iterate on) - burning tokens like there is no tomorrow (or no token limit, if you care more about your money than the planet 😅 ).
2/3
I'm not a fan of "let's throw more resources at the problem" solutions. What currently happens in the AI-assisted software development community is, however, exactly this:
1/3
From sourcecode to step-by-step tutorial: I wanted to advertise the #pxlclock using programming tutorials. So I built this:
- AI breaks the #csharp code into steps; thanks @ursenzler.bsky.social
- Renders animations
- Writes tutorial
- Remotion makes the reel
Want a video about how I did that?
The 5th and final part of my post series about "to test, or not to test" takes a look at why the shown concepts help with LLMs and concludes the series.
www.planetgeek.ch/2026/03/24/t...
Do you like these longer topics, or do you prefer short single posts?
🎉 Deedle 4.0.1 is out! The F# data frame library just got a massive update - new packages, new operations, better stats, performance wins, and many long-standing bugs squashed. A huge amount of community-originated requests in one release. Full notes 👇
github.com/fslaborg/Dee...
#fsharp #dotnet
Yes, about 10 years 😀
Still, having everything in a single process (and synchronised) makes debugging much easier.
