The first route had distance errors due to bugs in Ride With GPS. You corrected those for the second route. Then finally corrected for controls being in slightly different places than on the route sheet.
Thanks for reminding me freeflow, that was indeed the case.
FWIW, for those who are into mappery and such like, it appears to be a bug in RideWithGPS in that if you leave a route in Edit mode overnight (which I do from time to time, as I face-plant my keyboard around pumpkin time
) then it magically gains distance by the morning. RWGPS are aware of this. It makes not one single iota of a difference to the route itself, except that the total distance appears to be over by a couple of kms. There's a standard fix, which is to waggle all the control points, forcing a recalculation on each segment, which is both time-consuming, dull, and means all the cuepoints have to be recreated; RWGPS themselves have a big red button to do this for them. I saved the previous version of Hereward to see if I could script up a GPX-checker to see whether a specific route suffered from this particular issue before I published it — realising the problem is there in a specific route is half the battle. The current routes now don't suffer from this overdistance — 309km is what I expect we will ride on the road, as it's close enough to what we rode last year, although Fidgetbuzz and I did add on a 6km detour due to chatting too much and ignoring the GPS in the dark
(my GPS recorded 316km including the detour)