[OSM-talk] Older planet files (from 2007)?

Ciprian Talaba cipriantalaba at gmail.com
Wed Oct 21 13:45:49 BST 2009

Hi everybody,

On Fri, Oct 16, 2009 at 9:57 AM, Frederik Ramm <frederik at remote.org> wrote:

> Yes, it's called planet.openstreetmap.org!
> Note, however, that if you go any earlier than October 2007 you will
> encounter version 0.4 planet files which cannot be processed using today's
> tools. You will have to use SVN to retrieve an old version of
> /applications/utils/osm-extract/planetosm-excerpt-area.pl
> from SVN, and then use
> /applications/utils/conv05/04to05.pl
> to produce a version 0.5 extract which can be processed with today's tools.
> You can also get a primitive history animation (only nodes shown as dots)
> from labs.geofabrik.de/history, or if you give me the bbox then I could
> make an animated GIF like the ones on
> http://www.geofabrik.de/gallery/history/. But I must admit that the ITO
> animations look best so if you get them to do one that would be preferable.
> Bye
> Frederik

Thanks for all the info. With your help I was able to obtain older extracts
for Romania and to create some evolution movies. They are not great compared
with what ITO have done in the past but they covered our needs (we have also
contacted ITO, actually Peter Miller, but we had limited time to do this,
and we didn't get an answer in this short period).

I had only one small issue with older planet files: after converting from
API 0.4 to API 0.5 some of the ways had members without ref (something like
 <way id='23312491' timestamp='2008-03-16T11:15:13+00:00' user='Ciprian'
    <nd ref='' />
    <nd ref='' />
</way>). I have removed these ways and compared the results using 2 versions
of JOSM (one for 0.4, another for 0.5) and the results were similar, so I
move forward. Is this something that is documented somewhere?

If you want to see the results here are some of them:

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20091021/597e1ce4/attachment.html>

More information about the talk mailing list