Mark:
Thanks for the new version 2.5.7. I got it loaded just in time to see what happened at midnight. The Dew Gauge went through the midnight rollover without a hitch. There was no change in the Windchill autoscale range between midnight and the first data upload of the new day. Thanks for fixing that. And thanks for updating the version history on the Cumulus Wiki.
However, I did notice the same errant autoscale behaviour on the main Temperature Gauge. I have not seen that before tonight. It did clear itself and go back to normal within 30 seconds.
See attached screenshot. Sorry, I was not quick enough to grab a copy of the realtimegaugesWC.txt file before it corrected itself. I assume this is the same error trapping that affected the Dew Gauge.