From OpenStreetMap Wiki
Jump to: navigation, search

What does it mean if you get an error message saying that the spellcheck test could not be initialized? Do I have to get a file from somewhere? --Frederik Ramm 00:38, 12 May 2007 (BST)

[~/.josm/plugins/validator] # l
total 44
drwxrwxr-x  2 elwood  user    512 11 Mai 15:25 .
drwxrwxr-x  6 elwood  user    512  8 Mai 20:26 ..
-rw-rw-r--  1 elwood  user  13172 11 Mai 15:24 1178889896189-words.cfg
-rw-rw-r--  1 elwood  user  13172 11 Mai 15:24 1178889897248-words.cfg
-rw-rw-r--  1 elwood  user  11281  8 Mai 20:23

--Elwood 07:49, 12 May 2007 (BST)

I was getting an initialisation error also, and telling me it was unable to download the words.cfg file. Thanks to the ls output of Elwood, I figured it tries to download it to the ~/.josm/plugins/validator/ directory. This directory isn't there by default and isn't created automatically by the plugin. After creating this directory with 'mkdir validator', the error was not returning, and ls shows a 1182339140528-words.cfg file in my validator directory now.

--Remco 12:40, 20 June 2007 (BST)

Exception when installing Validator plugin


at startup time I am getting the following error if I try to activate the validator plugin (stack trace taken from console):

loading validator
   at org.openstreetmap.josm.plugins.validator.Severity.<clinit>(
   at org.openstreetmap.josm.plugins.validator.tests.TagChecker$CheckerData.getData(
   at org.openstreetmap.josm.plugins.validator.tests.TagChecker.initializeData(
   at org.openstreetmap.josm.plugins.validator.tests.TagChecker.initialize(
   at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
   at sun.reflect.NativeMethodAccessorImpl.invoke(
   at sun.reflect.DelegatingMethodAccessorImpl.invoke(
   at java.lang.reflect.Method.invoke(
   at org.openstreetmap.josm.plugins.validator.OSMValidatorPlugin.initializeTests(
   at org.openstreetmap.josm.plugins.validator.OSMValidatorPlugin.<init>(
   at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
   at sun.reflect.NativeConstructorAccessorImpl.newInstance(
   at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(
   at java.lang.reflect.Constructor.newInstance(
   at java.lang.Class.newInstance0(
   at java.lang.Class.newInstance(
   at org.openstreetmap.josm.plugins.PluginInformation.load(
   at org.openstreetmap.josm.Main.loadPlugins(
   at org.openstreetmap.josm.gui.MainApplication.main(

Some weeks before all worked fine. Then I installed Ubuntu 8.04.1 for 64bit architecture.

All other plugins but the validator seem to work fine. Any error with the plugin?

Edit: Fixed it by re-downloading josm-latest.jar - definitely a living project :-) --Marcus-K 13:31, 26 August 2008 (UTC)

Way end node near other highway

The "Way end node near other highway" warning seems at a first sight clear but looking at the warning I do not get why it is complaining there and not somewhere nearby. Looking in some sources I do see limits seems to come from node_way_distance and way_way_distance keys.

In my install of JOSM, I have not set validator.UnconnectedWays.node_way_distance and set validator.UnconnectedWays.way_way_distance and I am wondering what defaults are used in this case. On [1] I read that you can set these keys to 0.5m with the comment that then only real unconnected ways are reported. I this a good idea?

Emvee 12:34, 2 August 2010 (BST)

  • Thanks, this solved my problem with an avalanche of false positives Bulwersator (talk) 07:09, 28 September 2013 (UTC)

Overlapping areas/ways

1. Overlapping areas Why does Validator question adjacent landuse areas using shared nodes? For example landuse=farmland and landuse=residential have no land between them. Polygon lines have no width so it would seem logical that polygon lines could be one over the other without considering them to be overlapping. Or is it just bad OSM-mapping to create areas with the characteristic mentioned above?

2. Overlapping areas with ways (roads,tracks,etc) I personally do not create areas and ways using shared nodes because the ways do have a width. But other OSMers do and Validator questions them to. I find the editing of current OSM-data to be far more difficult if ways and areas have shared nodes. What is then the general standard that we mappers should be using.

NOTE: OSM-Gurus: Please let me hear from you. --Dcp 16:19, 25 September 2010 (BST)

Validator notifications are categorized as
  1. Errors: duplicate nodes and some others: these should probably always be corrected before upload.
  2. Warnings: for example "way end near other way"; things that might be errors - a road that just barely isn't connected to the next road, or it might just be something like a hedge that ends near a footway. Users are encouraged to check these spots before upload just to make sure they've drawn what they intended to.
  3. Information: Things that might be missing information normally entered, such as major roads without a reference number - but not all of them have a ref. Or using tag values that aren't in the JOSM presets - could be a typing error, or a deliberate new value. The overlapping areas test is apparently in this category.
The full list is here. I don't know if it's up to date, though.
As to your question #2, there's explanation of the current nonconsensus at Talk:Editing Standards and Conventions#Areas adjacent to ways: source? Alv 16:50, 25 September 2010 (BST)