[OSM-talk] Why I don't use JOSM (was Re: Non-map-based OSM editor)

[Frederik Ramm]

Hi,

Elizabeth Dodd wrote:
>> (And, if my memory
>> serves, you get slapped on the wrist for trying to download too much.)
> 
> based on a formula which is related to the number of square kilometres
> involved, and which has no relation to the amount of data involved.
> It is presented as a restriction from the API, whether or not that is
> true.

The API restricts downloads to an area of 0.25 degrees squared (i.e. you 
can download an area 0.5 degrees high by 0.5 degrees wide, or 0.1 
degrees high by 2.5 degrees high). Note that degrees != square 
kilometres; the nearer you are to the Equator, the more square 
kilometres you are allowed to download. JOSM informs you of this 
restriction in the download box, but allows you to try your luck 
nonetheless (it just tells you that your query is likely to be 
rejected). In addition, the API will reject requests for an area that 
contains more than 50,000 nodes, even if it is smaller than 0.25 degrees 
squared; sadly, JOSM has no chance of informing the user beforehand that 
they might hit this limit.

> In places where I have been and tend to go I can get the slap on the
> wrist described above when in fact there is nil data at all to
> download :(

Then I guess you have to map more ;)

No, seriously, it would be great if someone found a way to modify the 
API (more precisely, the cgimap program) so that it accepts requests for 
larger bounding boxes in sparsely mapped areas. It is probably not easy 
to do this in a performant manner though, as by the time you have found 
out that the area that was requested contains a million nodes, you have 
already spent a lot of time slowing down the database - maybe one would 
have to have a special mapping density index for various regions of the 
planet.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00'09" E008°23'33"