[josm-dev] Handling large files?

[Frederik Ramm]

Hi,

Michael Speth wrote:
> So I would like to use DB4O (An object oriented DB that can be embedded).

You're asking for trouble. Of course you can do that. You can also use
Hibernate to serialize OSM objects if you so desire. It's just that
you'll not get anything out of it that can even remotely be called
"performance". If your main use is doing queries like "on what road is
the following point", then you absolutely have to have a spatial index.
I read that there is a DB4O spatial initiative but they haven't produced
results yet. You can emulate a spatial index by calculating "tiles" or
"buckets" like it is done in our current MySQL API but it'll be
considerably more manual work than just writing "db.set(something)".

> The DataSet dataSet that was returned from the parser can simply be added to
> DB4O by the following call assuming that the db object is already
> initialized.
> 
> db.set(dataSet);
> db.commit();
> 
> Easy huh?

Gives you a warm fuzzy feeling but won't give you speedy results.

> So are you suggesting that JOSM cannot do what I require?

JOSM is optimized for interactively working on objects (if anything) -
not for mass querying.

> If not, then the libosm would have parsers that are similar to JOSM?

Similar but leaner, with less memory consumption. But even with these
you'll probably run into trouble. Tell us more about what exactly you
want to do and we can tell you if it will work.

Bye
Frederik

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