replicate-sequences
a small and simple tool to get the correct replicate sequence id of OpenStreetMap's replation-diffs from a timestamp. Currently deployed on a server near you: http://osm.personalwerk.de/replicate-sequences/
API
Besides the Web-Form the tool can handle the following API-Call:
http://osm.personalwerk.de/replicate-sequences/?2013-01-01T10:00:00Z
This can be used to always fetch the latest state.file in conjunction with some unix-foo:
curl "http://osm.personalwerk.de/replicate-sequences/?`date --utc "+%FT%TZ"`"
Setting up you own
Setting up your own copy of this service is quite easy. Check out into a PHP-Enabled docroot. Create a MySQL-Database and a User that has INSERT
and SELECT
permissions on it. Copy conf.php.tpl
to conf.php
and fill in your database name and user credentials. You also MUST FILL IN an E-Mail-Adress where admins of planet.osm.org can contact you, in case this script runs crazy on your server.
Now populate your database with schema.sql
to get the empty tables. You could now start running update.php
and it will download >750000 state.txt
-files to populate your database. This would take at least one complete day. Instead, i'd suggest to also import initial.sql.gz
, which already contains the first 750000 state-files and pre-loads your database with this information. You can then use update.php
to stay on track.
Contact
If you have any questions just ask at osm@mazdermind.de.
Peter