[OSM-dev] {To All} Re: OSM Data Question from FireGirl

Brett Henderson brett at bretth.com
Wed Feb 13 04:08:05 GMT 2008

Fire Girl wrote:
> Meep Mepp, information overload! LOL! I see so many paths and almost 
> stressed out thinking about them all! Haha. I like the pre-staged data 
> Jeremy described, and that looks useful, and think Jason's concepts 
> are very good. I was thinking of setting up my own Database to that 
> end.... but I digress, I am not sure with path I need to go in yet! 
> Maybe someone can help me make a decision? Brett and others mention 
> this idea of extracting Boxes and Polygons. What is sure fire way to 
> get that working right? I mean, if I want just a city, let's say, 
> Makaha, Hawaii -- What would be proper steps (in any platform), to get 
> to a Makaha Hawaii .OSM file? Is that even possible to get that level 
> of detail into a OSM?*** I don't want to go further in my requests, I 
> feel I should be able to do something with that sub-set of data.... 
> but still, don't want to go thru unecessary lifespans with this nearly 
> 80 gigabyte file LOL not really knowing what I am doing. But, actually 
> though, I am willing to set it up into a database if that you all 
> think is best way to go... but I am still not sure how to get to very 
> local data for any worldwide spot. So I mean, this Osmosis is the 
> sure-fire way to get to this?
Options are good :-)

Firstly you need to figure out what you want to do with the data :-)  
That will help limit your options.  You'll need to determine things like:
1. How many different areas do you need?
2. How large are the areas that you need?
3. How up-to-date does the data need to be?

If you're pulling out many different areas at regular intervals then 
something like osmosis will give you a lot of flexibility.  If you need 
a specific area that isn't *too* large, osmxapi will be the simplest.  
If up-to-date data isn't critical then a pre-generated extract may be 
more appropriate.

In general, large amounts of data are best served by osmosis.  With 
osmosis you can slice and dice to your hearts content without consuming 
server resources (other than the initial planet download).  Very small 
areas may be served by the standard osm api but try to limit your usage 
of this.  Somewhere in between lies osmxapi which will handle larger 
areas than the main API, but probably not ideal for downloading whole 

The main API will provide the most up-to-date data, osmxapi will be 
within 10 minutes, and osmosis just works with whatever files (ie. 
planet file) and changesets (ie. daily/hourly/minute) you choose to 
download but can also be within 10 minutes with a bit of effort.

If you're just trying to get used to OSM and don't know what you need 
yet, I'd suggest downloading one of the extracts suggested by Milenko 
(http://planet.king-nerd.com/usa/) and playing with it.  When you get 
something working using that data then you can figure out how to get 
your updates to that data.

More information about the dev mailing list