no, your solution is the shortest ;-) (I created an iterator to replace the "for each" loop). Someone reported another feature non compatible with the projection Lambert. I will show here my code changes proposal on this ML later, when I will have a solution with the other issue. Thanks in the mean time, Pieren
On Sun, Aug 3, 2008 at 1:32 AM, <[EMAIL PROTECTED]> wrote: > Hi Pieren, > > On Sat, Aug 02, 2008 at 12:19:56PM +0200, Pieren wrote: > > - a problem with the tool "align nodes in circle" which is calculating > the > > average position of the selected nodes. Unfortunately, this average is > done > > with a first value initialized like this: > > new Node(new LatLon(0,0)); > > but this is not workign with the Lambert projection (the east/north are > high > > negativ values). I made a fix avoiding this kind of initial node but > rather > > use the first "real" node as initial value. Is that the correct way to > > proceed (in which case I could submit the change) ? > > I wrote that code. > > That sounds fine - the only reason for the new LatLon(0,0) is to > get the initial node to have EastNorth coordinates of 0,0. This > was to allow an easy "for (Node n : nodes)" rather than doing a > loop that missed the first node. > > A quick guess is that the easiest way to fix is probably to add: > > avn.eastNorth = new EastNorth(0, 0); > > after that line, but maybe your code is shorter? > > Cheers, > > -- > Matthew > _______________________________________________ josm-dev mailing list [email protected] http://lists.openstreetmap.org/listinfo/josm-dev
