[OSM-legal-talk] Exception in Open Data License/Community Guidelines for temporary file

[James Livingston]

On 29/06/2011, at 4:25 PM, Frederik Ramm wrote:
> On 06/29/11 05:21, James Livingston wrote:
>> I don't think it would be treated differently, because I believe that an
>> in-memory data structure would still be a database (in the ODbL and
>> database right sense of "database"). I don't see how the storage
>> mechanism makes a difference.
> 
> Would you therefore say that before I can use proprietary software to process an ODbL data set, I would have to request from the software provider a legal statement about whether or not it does create a database internally?

That's a slightly different point, what I was trying to say was that if you have something in memory that doesn't qualify as a database, then it won't magically become a database if it happens to be swapped to disk. Conversely, if you have something that is a database, it doesn't stop being a database because you load it into memory. Being a structured representation of data makes it a database, not where the bits happen to be.


To your point, what happens when someone loads a .osm file into JOSM? First, I'd claim that a .osm file is a database. Obviously not a relational database that gets handled by SQL-using software, but still a database. I'd also claim that the in-memory data structures of JOSM form a database too.

The ODbL saud "Database - A collection of material (the Contents) arranged in a systematic or methodical way and individually accessible by electronic or other means offered under the terms of this License". I think the data structures JOSM uses to view and edit certainly qualifies. If you immediately saved it back to a file after loading, it would go Database (on disk) -> Database (in JOSM memory) -> Database (on disk), not Database -> Non-database -> Database. The data stored in the in-memory database is equivalent to the on-disk one, but it's still there.


I would think that the vast majority of software that does useful work will need to create a temporary database to do so. I've never looked how Mapnik is implemented, but I assume it creates a bunch of data as it goes about what POIs and labels it is rendering where, so the map doesn't get too crowded. Is that a derived (temporary) database?


> If I use software that builds an in-memory data structure which you believe to be a database in order to make a produced work, how would you suggest that I fulfil my obligation to make such derived database available on request?

I have absolutely no idea. It's one of the many things I don't know about how the produced works part of the ODbL will work in practice.

-- 
James