Platform status

The platform has been running for over a year now, and at the point when I slowed down on building it I considered it good enough, but it was pretty far from it, ofc. And in the recent months I haven’t done much with it at all.

Work and other things have taken focus, time and energy. After the initial little rush of actually accomplishing what I set out to do, most of the motivation faded. And without the pressure of actual users, some of the momentum is always lost. That’s how these things go, at least in my experience.

However. Lately I have been coming back to the project for a couple of reasons.

First, I have started using agentic software development. Say what you will about it, but these days you can get pretty good results giving an agent a specific issue and enough context to work with. Especially for a project like this, where the foundation is already there. It has been pretty mind-boggling. Again, I am very much aware that “good results” and the actual long-term value of agentic software development in general are very debatable.

Second, I may have a test user, which of course would up the stakes quite a bit. It is one thing to run my personal apps and break things left, right and center with no real damage done. It is something else if someone is actually depending on this thing to work reliably over time.

These things have given me both the motivation and the ability to take care of some of the things that were easy to postpone when I was the only user: private networking, better container isolation, proper backups, metrics, monitoring etc. etc. Big boy pants. Some of these things are actually now in place, but there are a few more in the pipeline.

So that’s where the platform is at these days. It works, but there is quite a gap between “it works” and having someone else use the thing with confidence.

It would be a fun challenge to figure out how to close that gap. We’ll see.