North-Pole-Node seems to destroy the Map in JOSM

[Aleks-Berlin]

What steps will reproduce the problem?

1. load a continents.osm (all place=continent),
2. add node=269538499 (N-pole at 90.0,0.0),
3. try to zoom in the map (as usual.

What is the expected result?

typically zooming via mouse-wheel works fine,

What happens instead?

only black map is seen, no change in scale (constant >20k km).

Please provide any additional information below. Attach a screenshot if possible.

URL:http://josm.openstreetmap.de/svn/trunk
Repository:UUID: 0c6e7542-c601-0410-84e7-c038aed88b3b
Last:Changed Date: 2015-11-24 00:04:12 +0100 (Tue, 24 Nov 2015)
Build-Date:2015-11-23 23:14:21
Revision:9060
Relative:URL: ^/trunk

Identification: JOSM/1.5 (9060 de) Windows 7 64-Bit
Memory Usage: 123 MB / 989 MB (34 MB allocated, but free)
Java version: 1.8.0_66, Oracle Corporation, Java HotSpot(TM) Client VM
Program arguments: [D:\Sourcen\osmextract\planet\final\huge_continent_ocean.osm]
Dataset consistency test: No problems found

Plugins:
- turnrestrictions (31772)

Last errors/warnings:
- E: NaN in greatCircleDistance
- E: NaN in greatCircleDistance
- E: NaN in greatCircleDistance
- E: NaN in greatCircleDistance
- E: NaN in greatCircleDistance

[Aleks-Berlin]

93 KB OSM (all continents + oceans as node + AQ-relation

[Don-vip]

We have following errors:

ERROR: NaN in greatCircleDistance
ERROR: NaN in greatCircleDistance
ERROR: NaN in greatCircleDistance
ERROR: NaN in greatCircleDistance
INFO: bbox: [ x: NaN -> NaN, y: NaN -> NaN ] is out of the world
INFO: bbox: [ x: NaN -> NaN, y: NaN -> NaN ] is out of the world
INFO: bbox: [ x: NaN -> NaN, y: NaN -> NaN ] is out of the world
INFO: bbox: [ x: NaN -> NaN, y: NaN -> NaN ] is out of the world

Any idea how to solve this?

[stoecker]

Any idea how to solve this?

Switch to Discworld. No singular points there :-)

There probably are a lot of such issues in JOSM depending on the projection and the data. While the 180th meridian is more an issue and my eyes and we had only a few complaints in the years, the north pole probably is not. The normal maps we have are cutoff at 85° if I remember right.

[anonymous]

If there is a any problem with pole-coordinates, it should be easy to handle it internally as x > 89 deg ==> x = 89 deg.
Just give a warning that "2 nodes had to be adapted to avoid further problems."
But crashing or Map beeing unusable is no good option.

[stoecker]

A lot of the projections either get numerical instable or produce wrong results when you go out of the defined area of usage. JOSM currently assumes that users don't do this and has a set of more or less successful mechanisms to prevent unwanted usage. Getting this really right is a lot of work as it means a lot of testing. I myself e.g. spent probably 2 weeks at zooming and numerical issue for low and high zoom as well as display outside of the projection center area and stopped with some know issues still there.

As such issues affect only a very small percentage of users and need a lot of work usually nobody does fix them. Like always patches are welcome.

The NAN can probably be fixed relatively easy. But the next bug then will be proper display and that's more complicated as said. 90° (and values near 90) cannot be displayed in Mercator, as the coordinate will be ∞ and that cannot be displayed. Even 89 is VERY distorted. For other projections that's no problem, but they may have other singular points or parts which can't be displayed.

See e.g. for a small overview: ​http://gmt.soest.hawaii.edu/doc/5.2.1/GMT_Docs.html#gmt-map-projections

[anonymous]

The NAN fixes would be great! Perhaps it also allows zoom-in after the fix again.
I do not need to see the poles on right position! I just would like to have them inside my data.
I promise to spend much time with OSM and help the project (as I did before), but I'm no Java programmer to fix this.
Many Thanks for supporting this tool! Alex

[bastiK]

In 9101/josm:

fixed #12155 - North-Pole-Node seems to destroy the Map in JOSM