Repository navigation
Rescan a controller after a programming change, from the Point Selection page, with logging carrying on while it reads.
Already on 0.1.0? git pull (or unzip this version over the old one), then restart the server. No new dependencies; your configuration and database are not touched. Upgrading has the details.
New? Download Source code (zip) below and double-click install.bat. Getting started takes it from there.
Added
- ⟲ Rescan controller on Point Selection. Pick a controller (the one holding the unit on screen is chosen for you) and it reads its object list again on the server's own BACnet connection. It takes seconds for a panel, a minute or two for a supervisory controller with thousands of objects, and shows its progress.
- New points go on their unit's list, unticked, placed by the equipment map.
- Renamed points keep their logging and their history.
- Points that already have a unit are never moved, so a split you made by hand stays as it is, and nothing is deleted.
- Points the controller no longer has are listed for you to untick, and only after its whole object list was read: a short read is never taken for deletions.
- A controller whose address changed is found again by Who-Is.
- Reads only, like everything else here, and not offered to other computers on the network.
- How it works.
Fixed
- ⟳ Rebuild model said "Rebuild failed" after every successful rebuild while the statistical baselines are off, which is the default.
- A side-stream separator is no longer taken for a condenser pump. Its pump's run status was classified as a condenser pump status, so a filter pump that runs all the time counted as proof the condenser pumps were on; its DP shared the loop DP's role. It now has roles of its own. The first rescan of a real plant panel found this.
Also
- The test suite now checks that every page's script parses, and runs 458 tests on Windows and Linux, Python 3.11 to 3.14.
Full list: CHANGELOG.md.