[OSM-legal-talk] The big license debate

[Frederik Ramm]

Hi,

Robert (Jamie) Munro wrote:
> "OpenStreetMap is a project aimed squarely at creating and providing
> free geographic data such as street maps to anyone who wants them. The
> project was started because most maps you think of as free actually have
> legal or technical restrictions on their use, holding back people from
> using them in creative, productive or unexpected ways."

I think this sounds very good. At closer inspection, it promises 
nothing, nada, zilch. But neither do preambles of legal text, they just 
say something about the "spirit" in which it was created.

I'm all for free data to be used creatively.

> To start the debate: I believe that the aims are valid. I believe that
> we need a license that lets people download the data, make maps from it,
> and be able to do what they please with those maps. 

Yup. I'll try to outline my own hopes & expectations.

In the long run, I hope that OpenStreetMap will become OpenWorldMap, a 
(probably distributed) one-stop self-service system for geodata of any 
kind, where everybody can participate and everybody can use the data for 
any purpose. I firmly believe that data about our planet should be free 
and available to anybody; it isn't today, and I hope OSM will rectify that.

I also hope that users, especially big commercial users who have the 
resources to make major improvements, will contribute stuff back. I am 
however a bit skeptical about how this can be done. Especially, I expect 
that by that time the project will have grown a lot of administrative 
overhead, where, say, any change in the data model would break a hundred 
interfaces and thus, if ever, needs to be planned two years in advance 
and so on; so that an entity using our data for a specific purpose that 
is not properly covered by our data model will have to convert data into 
their systems, and not be able to contribute it back without first 
convincing a dozen administrative OSM bodies that changes are required, 
and so on.

Furthermore, I want this to happen freely, not forcedly, and I believe 
that the large community behind OSM will make it very attractive for 
users of our data to actually use our current data, instead of something 
they siphon off and from then on have to care for themselves.

But if someone really wants to take our data, make changes, and seal it 
off for restricted use, then so be it. I am willing to accept that; it 
is the price I'd pay for maximum openness, and for avoiding legal troubles.

I would like to avoid any anti-commerce sentiments by the OSM project. 
Commerce is what makes our world go round, and even people using our 
data commercially to make profits and not giving anything back will 
further our goals, if only by making us more popular. I see a possible 
parallel with Linux here; commercial entities adapting Linux were rare 
in the beginning, and those who used it were quiet about it, for fear of 
being laughed at. Meanwhile using Linux is en vogue. I could see
commercial entities using our data, at first experimentally, later 
creating products from it, but not yet mature enough to make a public 
commitment to OSM. However, after a few years, they might indeed come 
out and publicly involve themselves in OSM.

I would also like to avoid putting up tons of rules just to make people 
feel bad, or feel like they're jumping through loopholes or "getting 
away with something" which depends on a complicated interpretation of 
some legal text (e.g. the interpretation of words like "suitable", 
"applicable" or "practical"). It is bad enough to have that feeling when 
I fill out my tax forms every year. Whatever license we chose has to be
very clear to both users and contributors, with a well-defined process 
of deciding on grey area cases (as suggested by Laurence). Nobody should 
feel restricted in his creativity by a latent "don't know if this is 
really allowed" (and this is a big killer of creativity).

I would also like to avoid putting up rules that I cannot (or don't want 
to) enforce; that makes the whole license ring hollow.

I would also like to avoid being unfair in that I am more or less forced 
to let large companies get away with (at least moral) violations of the 
rules I have thought out - because they can afford a lot of lawyers and 
engineers to construct something that is legally and technically not in 
breach of my rules - while I put smaller companies or individuals at a 
disadvantage by providing anyone with the legal fodder to sue them for 
breach of my rules.

A "public domain" license would have everything I ask for.

An "attribution" component, if attribution to OSM suffices, would be 
acceptable, but I would really like to make this optional, a suggestion 
instead of a forced rule. It would create problems for a small number of 
users who, for political reasons, cannot do attribution, but I could 
live with that.

An "attribution" component requiring attribution to every contributor is 
very problematic, especially as contributors might not always want it. 
("This warhead sent to you using data mapped by Frederik Ramm!" Yes, our 
data can and will be used to kill people.)

Any kind of "share-alike" component, whether aimed at the data or at the 
finished product, has the huge problem of properly defining derived 
works in our context, or more colloquialy, the level of "infection". I 
can make up a hundred examples all in a grey area. If interpreted too 
broadly, share-alike can be a killer for commercial applications.

I could live with a share-alike component if it is very limited, i.e. 
requiring only the data taken from our database and changes made 
directly to that data to be "shared alike", but if the whole final 
product or whole new layers of information introduced to the data must 
be shared, then we are firmly on the path to become the "hobbyist's 
world map", or our license will be heavily circumvented by constructs 
where people only distribute "diffs" of our data, have their end-user 
download OSM data, merge it with the diffs, and use the combined 
dataset, or things like that.

Also, any kind of share-alike component should be controllable by the 
foundation or some other body to whom contributors transfer this right, 
so that legally binding, reliable decisions and exemptions can be made.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00.09' E008°23.33'