Edit: My mistake -getting old. I need to have ANSVR set to J2000 like nova.astrometry.net
Same problem I had a few months ago. But I've been playing with ANSVR and changed the settings.
By the way: It is now possible to install astrometry.net under WSLand set up ANSVR to run that instead.
Running Kstars/Ekos 3.0.0 on Windows 10 x64.
Last night got a bad alignment. I was aligning on NGC2775 but after alignment I was off by about 15 arcmin in RA and NGC2775 was on the edge of the FOV.
I've had something similar before and it was due to Ansvr being set to J2000 instead of JNow. So I checked that it was set correctly.
I've attached the Kstars log and the Ansvr log. The problem is apparent at 21:22:20 (line 20718 of Kstars log and line 1662 of Ansvr)
Target Ra/Dec in JNOW is 9h 11m 21s RA, 6d 57m 33s Dec or in decimal degrees 137.8375, 6.95916666666667
Ansvr returns JNOW solver result of 137.585347251, 7.038789885
Near enough to dec 0 so I make that approx 16 arcmin
Ekos confirms the solver coords on line 20718 but on line 20719 reports them as RA 09h 11m 21s DEC 06d 57m 33s and determines that is within the acceptable range. It looks like Ekos is treating the Jnow coordinates as J2000
Is there a setting somewhere I've changed by mistake?