********THIS IS NOT A GOOD PLACE TO MAKE BUG REPORTS************. Do that here: https://github.com/iandees/xapi-servlet/issues

This is not a good place to make bug reports. Do that here: https://github.com/iandees/xapi-servlet/issues

@Tilo: No, the jxapi is/was not down. It was getting inundated with duplicate requests so the process queue was very long. I restarted postgres to clean out the queue.

I compute the sum of the areas of the requested bounding box(es) and compare that to 10. If larger, 500. If less, let it through.

There are no plans to do areas larger than 10 square degrees. As you can see already the hardware has a hard enough time keeping up.

@Katzman:

1. It is possible to enter multiple tag predicates:

[amenity=foo|bar] := "amenity"="foo" OR "amenity"="bar"
[amenity|leisure=foo] := "amenity"="foo" OR "leisure"="foo"
[amenity=foo][area=bar] := "amenity"="foo" AND "area"="bar"

2. Not at this time.

Your map is still incomplete: the majority of power station/sub_stations are mapped as areas, not nodes. At least put a dot on the map for the first node in such a polygon. Also, as others have mentioned before there are plenty of power=line that don't have power=[tower/pole] marked on their nodes (since it's so tedious in all the editors).

MapDust is littered with so many bugs that are not reproducible I don't waste my time with it anymore. The last few times I visited I spent more time clicking through the process of closing a bug than actually fixing the data, so I gave up and continue to map my area as perfectly as I can.

I recommend making a post to the imports mailing list (to answer your projection question) and follow the import guidelines found on the wiki.

I know that I've mapped dozens of substations in the midwest with polygons. Are you accounting for those? How about full power stations?

@Geonick: yes, this really does follow the XAPI query syntax. You can indeed supply multiple tag predicates. It will eventually be possible to include multiple bounding box predicates, but at the moment I only allow one.

It bounced a few more times and is down hard now. There appear to be some ECC memory errors, but I put fresh memory in a while back. Looks to be a bad motherboard. Thanks for the stress test, everyone!

The machine restarted itself at 16:00UTC or so, which lines itself up with your request fairly well.

But yes, large bounding box queries are mildly slow -- I don't remember them being particularly fast on the other system, though.

Dude, learn how to modify the style files yourself and contribute or shut your trap. Yelling about old bugs is not a way to attract developer attention to your cause.

Hi PJ! Great to hear that Portland is using OSM data. I recommend that you sign up for the talk-us mailing list and point everyone to this post. Keep us informed about how you're coming along and if you need any help or have questions! There are lots of people in the US interested in how governments can use and contribute to OSM, and Portland seems to be partly leading the way.

Great work!
Ian (@iandees on Twitter)

@andrewpmk: ESRI didn't claim any size when it confirmed the 400M for its map count. Each OpenStreetMap tile is indeed a discrete map, so I counted it that way.

I wrote the new Add WMS dialog and will go ahead and allow the dialog to change size. It definitely doesn't need to be that tall!

Perhaps all of these usability things that you could go on and on about should be written as bugs against JOSM? I certainly would enjoy having an "afternoon project"-sized bug to pick off every once and a while.

This is a pretty large task! The issues mentioned here are common to all imports and have not been solved yet. I invite you to join the talk-us mailing list so we can include you on any discussion around this topic.

Can you post a screenshot of the result? Have you tried using the SVG symbolizer to allow for arbitrarily-sized symbols?

Data that has not been loaded into the OSM database yet should use negative ID numbers. There are no "safe" positive ID ranges as they will all be used eventually.

Have you seen Lars Ahlzen's TopOSM stuff? I'm working with him on covering the entire US with TopOSM right now. Looks like you guys might be duplicating effort! http://toposm.com/ and osm.wiki/TopOSM

Can you describe what you would be looking for in such a tool?