[OSM-talk] Using maps produced from OSM data

Schuyler Erle schuyler at nocat.net
Mon Mar 20 23:08:02 GMT 2006


* On 20-Mar-2006 at  1:49PM PST, Richard Fairhurst said:
> 
> Hm... as some of you know, I've been wondering for a while about 
> exactly how far a 'derivative work' extends.
> 
> I noticed Imi posted an explanation to the Wiki FAQ page, which is good 
> - much better to be unambiguous. And Tom posted a lucid comment on my 
> blog where I was musing on the same issue. But I'm still unsure as to 
> where the boundary lies when it's not a simple JPEG-in-HTML case.
> 
> Here's four stages of the same thing:
> http://www.systemeD.net/stuff/sharealike/
> 
> Could I ask for some views/consensus on where the boundary lies?

The map itself obviously would also be licensed SA, since according to
the stated standards on the wiki, it is substantially "based on" the
BY-SA-licensed geodata. 

You seem to be asking "when does the label need to be licensed SA as
well?" The point of Share-Alike, I believe, is to guarantee the
preservation of the right to reuse. If your use of new original
content alongside BY-SA-licensed content does not interfere with
unlimited reuse by others of the prior BY-SA content, then, no, I
don't think your new original work needs to be BY-SA as well.

If we accept this premise then I believe (not a lawyer) that your #1
(Separate pages) and #2 (Same page) do not constitute a "Derived
Work", because someone could tear out the map, and use it separately.
By contrast, I think your #3 (Same page, with label) and #4 (Same
page, placed on map) would constitute a Derived Work, because, if they
were not also licensed BY-SA, then the unlimited right to redistribute
the prior work (i.e. the map, or the geodata via the map) would be
infringed.

Naturally, an actual lawyer, judge, jury, or creator of a prior work
might disagree.

The fact that you have identified such a razor thin line between what
might and might not constitute a Derived Work points to the likely
unsuitability of existing Creative Commons licenses for geodata, as
much as I applaud OSM for adopting *something* open and not waiting
around for better licenses to simply appear.

The problem is that geodata is not like art or music, because it has a
raw data form and potentially many finished visual or analytic forms.
By the same token, it's not really like software, because although it has
a dual form (akin to code's source versus executable), it doesn't
describe a process, it describes facts or opinions about the physical
world.

I'm starting to think that one responsible option to do might be to
seek an LGPL-like solution, where modifications to the original data
demand ShareAlike redistribution, while visualizations or analysis of
the data in derivative works might permit something weaker. Think
direct modification versus linking or inclusion... I'm sure this is
full of all kinds of potential loopholes, though.

On the other hand, we might say damn anyone who wants to keep their
use of the community's data to themselves! and stridently insist that
OSM's data be Free in any form, for ever and ever.

Tough choice. What's more useful to humanity?

> (Much more fun than all of this... I cycled 100 miles at the weekend 
> along National Cycle Network route 4, with GPS on handlebars. OSM is a 
> great excuse to discover interesting parts of the country!)

So that's where you get all the time to think about these fine
details. :-D

> yours sounding increasingly like a stuck record,
> Richard ;)

I sure hope that's a stuck record for which you have a license for
public performance of derivative works!

SDE




More information about the talk mailing list