Updated with MRmP's experimental auto-calibration and bugfixes and stuff...
- - - - - Additional Content Posted / Auto Merge - - - - -
Fixed. I hope most of the fixes will be in the new version of RN.
Should all be present in 1.72. FPS, please retest
- - - - - Additional Content Posted / Auto Merge - - - - -
System.ArgumentOutOfRangeException: Length cannot be less than zero.
at RegulatedNoise.Form1.OutputEddnRawData(Object text) in d:\git\RN3\RegulatedNoise\RegulatedNoise\Form1.cs:line 2602
I couldn't replicate this, it seems like corrupted or unexpected data has been received from EDDN.
If it does it again with v1.72, it should generate a RegulatedNoiseDump.dmp file - this will be ENORMOUSLY useful in debugging the problem...
- - - - - Additional Content Posted / Auto Merge - - - - -
After moving from 1.6 to 1.71, the OCR got a whole lot less accurate. It can no longer tell the difference between D and O, and Y and V, among other things.
Also, when editing the ocred valued using the edit results button, i've noticed that everything is duplicated at least twice, often as many as 5 times.
I also get that Length cannot be less than zero. error when listening to EDDN
Hm. The OCR itself hasn't changed since v1.6, strangely (check GitHub if you don't believe me!

). You could try MRmP's new auto-calibrate and see if it helps. Or it could be that before, you were at systems and stations it recognised (and auto-corrected) but are now at stations it doesn't know about and can't auto-correct. Less likely though.
The duplication thing is odd, there is code to prevent that but it's not working I think. Fancy raising a GitHub issue for it please?
Again, with the "less than zero" error, please try to recreate it in v1.72 and pass the dump file across.