Hi Elrond, hello list, ... I did *not* yet test the patch, but, while biking around with a Garmin this morning, I realised that getting the rounding correct is an improvement, because your position is more precise. This could help in cases where a subtle rounding error will just put you on another road. You won't notice the difference as an end user, because there's many places where the choice of two parallel roads is difficult and any GPS unit will make some mistakes. My guess is that you could only see the difference when you would try to gather data on how many times you are on the right track and how many times you're 0.5 off... So if you ask me, "better rounding is better" and we should use anything that rounds better :) V. Clinton Gladstone schreef:
Elrond<elrond+openstreetmap.org@samba-tng.org> wrote:
"incorrect" is probably too hard (except for the very special case, where I needed the better rounding). "suboptimal" would be a better term. I tested your patch, and could see differences in the map. However, I could not really determine which would be more 'optimal'.