"20m AFTER SUNSET" schedule is not working

[brunetton]

Device

No response

Version

1.18.0-gita518080a+github240830

Question

Hi

Sorry to bother again :s but I'm trying to make a schedule with 20m AFTER SUNSET but this didn't done anything (I waited hours after sunset)

• I compiled 1.18.0-gita518080a+github240830 with SCHEDULER_SUN_SUPPORT
• set my long / lat for correct sunrise / sunset calculations
• flashed my device
• tested that the device is responding correctly when using switch buttons in the "Status" page
• checked again that long/lat coords are okay in UI (same float format)
• added two schedules:
  • "1h AFTER SUNRISE" / "relay 0 on" / "Relative" / Restore on boot: no
  • "20m AFTER SUNSET" / "relay 1 on" / "Relative" / Restore on boot: no

But nothing moved today. Nor the relative sunrise or sunset seems to work in my case :(

Some keys command values that I think might be useful:

> schAction0 => "relay 0 on"
> schAction1 => "relay 1 on"
> schEnabled0 => "1"
> schEnabled1 => "1"
> schHour0 => "9"
> schHour1 => "9"
> schMinute0 => "15"
> schMinute1 => "16"
> schRstrDays => "0"
> schSwitch0 => "0"
> schSwitch1 => "0"
> schTime0 => "1h AFTER SUNRISE"
> schTime1 => "20m AFTER SUNSET"
> schType0 => "3"
> schType1 => "3"
> schUTC0 => "0"
> schUTC1 => "0"
> schWDs0 => "1,2,3,4,5,6,7"
> schWDs1 => "1,2,3,4,5,6,7"

I'll test with bare SUNRISE and SUNSET to see.
Any tip is more than welcome !

Thank you

[brunetton]

I'll test with bare SUNRISE and SUNSET to see.

Unfortunately, it didn't worked neither with SUNRISE only. Did I missed something during compilation time maybe ? I thought it would be suffisent to enable SCHEDULER_SUN_SUPPORT but I maybe missed something 🤔

[mcspr]

Have you checked 'dev' branch instead?
Whats terminal output of get cfg?
Whats the output of schedule 0 & schedule 1?
Whats the output of event?

[mcspr]

...although...
version above has a bug related to .last & .next swap. AFTER uses .last, but it would update only after max(sunrise, sunset) timestamp happens to pass
'dev' has incorrect assumption about when the update should happen as well, but it is sometimes bound to midnight of the next day (which is also obviously bad)

let me have a look at how to correctly order those 3 together, then. debug info I asked about is probably not relevant