I’ve been waiting to get here for a while.
Floatpoint 1.0.0 is finally rolling out to every panel, and this time I’m doing it right. Everything is automated through the RMM instead of me SSH’ing into every single box.
How I’m doing it
All the new code is already on GitHub. The RMM just connects to each web server, stops the panel process, pulls the latest code, builds it, and brings it all back up again.
It’s simple, but it means every panel gets updated at basically the same time, without me having to sit there and babysit it.
Each step is logged and verified, so if something fails, it rolls back automatically and reports it back to me.
That’s the kind of control I want. No surprises, no guessing what node missed an update.
Why this one matters
This release is Floatpoint’s first real stable version. Before now, everything was technically pre-release. It worked, but it was always changing.
Now that the base code is solid, I can focus on infrastructure and reliability instead of UI polish or feature churn.
Panel Analytics V2
This update also launches Panel Analytics V2.
Version 1 was honestly more successful than I expected. It just tracked hardware events, node restarts, and usage patterns, but people loved it. It was simple and gave real visibility into what was happening behind the scenes.
V2 builds on that. It still collects the same hardware data, but now it also connects reasoning data from the Observe beta. So instead of just saying something happened, it can tell you why and how it happened.
It’s using the same kind of logic engine we’ve been building for ObserveAMS, just in a lighter form for the panel side.
How data is handled
Data collection works mostly the same way it did before.
Events get sent to Redis first for real-time access, then moved into a persistent database for 30 days. If you’re running Observe, that retention jumps to 90 days.
That balance keeps performance high without losing historical value. It’s all still anonymized and event-based, no user data, just system behavior.
Security and updates
The RMM rollout isn’t just about speed. It’s about consistency and verification.
Every panel pulls directly from GitHub, checks the commit hash before building, and won’t restart until the signature matches. After that, it verifies uptime and load before marking the update complete.
We also added a built-in status and update system in the panel itself.
It shows uptime, patch status, and scheduled maintenance straight from the dashboard. That ties into the analytics side too, so I can see the health of web nodes, databases, and cache layers without exposing internal metrics.
What’s next
After this rollout, most of the work will shift toward infrastructure, redundancy, database distribution, and better caching between regions. Basically making Floatpoint more stable under load and easier to maintain at scale.
New laptop day
Also, tomorrow I’m picking up a new laptop.
This version of Floatpoint feels like the one that sticks. Everything before it was just getting ready for this point.