I’m in the progress of creating my own OSM server. I have followed the guides found on the net and I now have:
Postgresql database with entire planet PostGis and Mapnik
Apache with Mod_tile serving tiles and renderd - All levels from 0 til 12 prerendered.
Now I want a editor for this database, like Merkaator, but for that I need a API/0.6/ URL on my apache. I have followed the guides to install the “rails port” and do have the ruby server on port 3000 now…
But, I’m not sure how to do the following:
Connecting the rails to my “gis” database - The rails installs three databases, but I just want to use the one I have.
Putting the rail stuff into my Apache (passenger ?)
Just a quick answer to one of your points, which is the use of multiple databases. Unfortunately you can’t combine the API database with the rendering database, as they use different db schema. What you will need to do, is use osmosis to extract minutely diffs from the editing API and replay them back into your gis/rendering db with osm2pgsql.
So, would it be much easier to do the whole thing “offline” - If I extract, with osmosis, the data from my rendering database and import that into the editor - and the import the changes back into the rendering database - is that possible ?
Most of my map changes are going to be rather static, a “one time deal” - slow changes anyway - Its most information that need to go into the ocean (I’m creating software for shipping) ?