No earth shattering new features addressed this weekend. Amanda and I spent most of the weekend fixing various "niggles" some of which were new and some old. But a lot of progress overall culminating with a long session of schedule controlled runs.
- A contact indicator on the route through the junction to Eastleigh Depot was misplaced, and has always been. The new Traincontroller V10.0B2 release was not happy, and wouldn't track a train into Eastleigh because of it. Move the indicator, and problem solved.
- Steam engines were derailing coming into Portsmouth platform 4. When watched closley, you could see the front bogie lift on one side. It turned out there was a thin strip of araldite on the inside of the rails, near the frog (a leftover from gapping the frog rails and inserting araldite into the gap). Cut away with the scalpel, and fully fixed.
- Some schedules announce their completion with a spoken message. The correct path to the "speech" script is e:\users\laure\OneDrive\documents\south downs railway\rr&co\speech.vbs
- The Virgin class 47 and Dapol Voyager have been declared dead. The first never really ran after a decoder install; the latter wouldn't run on a straight track without derailing, let along go round a curve or through a junction.
- The class 57 loco had taken a tumble, and wouldn't run. While investigating, I shorted something on the sound decoder and trashed it. It will need replacement.
- Several points were sticking. Contact cleaner seems to lubricate the point motor armature but the one into the crossing to Clanfield still has the electrical contact in slightly the wrong position.
- All the automatic joins and separates do work.
- 3 locos were speed profile. The little green central trains DMU looks like it may need reprofiling - its threshold speed is high.
- The Dapol grange steam loco has lost a driveshaft.
- We had Hans Tanner's approach to using MQTT working: messages are getting to a broker on a raspberry pi, from there to NodeRed and from a UI on there to a cell phone browser. Much more to do here, including decoding more messages (the <SE> messages aren't decoded, and sensor message don't report "board number, sensor number").