Minor issue with Digital Loggers swicth after upgrading to .767

Hello,

A minor issue with the Digital Loggers after upgrading SGP from .700 to the .767

I can access the switch and set the individual states manually using the ‘Control Panel’ but SGP generates an error ‘WebClient does not support concurrent I/O operations.’ (should be captured in the log).

Thank You,

Mike

Link to Logs

Useful Info

OS: Microsoft Windows 10 Pro
Ver: 4.1.0.767 (32-bit)
.NET: 4.8
ASCOM: 6.6.0.3444

Hmm. Im not really sure what this means. SGPro 4.1 has made significant changes to the way it interacts with switches. When I see an error like that it would mean to me that the driver has been asked to do 2 things at once and it hates that. The problem with that interpretation though is that this error is emitted when SGPro asks to change the state of the first switch. Not sure, in this case, what the “concurrent operation” is unless maybe its that the driver doesn’t want to be asked for the current state of a switch after SGPro sets it? That would seem awfully restrictive though.

Hi Ken,
I’ve seen that error pop up before, though very intermittent, and the way it was generated is when I first would start up SGP (.700) and quickly selected Control Panel → Switches → select any switch to manually set the state to on, I would see that same error sometimes. If I waited a few seconds (figuring maybe something in SGP still needed to complete) after seeing that error and attempted it a second time, it would go through successfully and SGP would not generate any further errors during the sequence session start or when SGP would power down the outlets at the sessions end.
I’ll do some trial runs and see if the solid failure is reproducible.

Mike

Hi Ken,

I’m pretty sure I found the root cause. In the Digital Loggers setup menu there is an option to set ‘On sequence delay’ and that was set to 4 seconds. I set it down to 1 second and the repeatable errors (using SGP .767) went away. I raised it up to 8 seconds delay and the errors appeared in every test run I did. I backed it down to a 2 second delay and every test cycle after that runs with no errors.
Hope this helps others for future reference.

Mike