[OSM-talk] Frederik declares war on data imports...
szocske at gmail.com
Wed Aug 11 10:24:55 BST 2010
I'm one of those people who just wish this whole licensing thing would
be over without too much damage to the community. Right now it feels
like airport security theater, where the terrorists have already won
by making us inflict all that inconvenience on each other. Freedom is
important, corporations claiming ownership of knowledge and culture
must be fought, but how much hurt is this specific license debate
It's not like the arguing parties here have completely opposing moral
and ideological standards: the arguments I have seen so far seem to
stem from good intentions and practical considerations.
It should be possible to bridge the relatively small legal difference
between two licenses that are one-way compatible using technology.
Then the users could have a less free view of all the data, and a more
free view of all but the imported data.
Please note I never said it would be easy or immediate. Finding,
maintaining and fine-tuning a system to host data under different
licenses probably requires more patience, trust and cooperation than
we can see on the list lately. Taking deep breaths, suppressing egos
and pride helps a lot, calling people names with adjectives does not.
There was a great summary here yesterday, completely lost in the noise already:
A more permissive licence is more permissive to data users, but more
restrictive on data importers.
In Object Oriented Design called this is called covariance and contravariance.
Mainstream OO programming languages simply ignored contravariance for
decades, claiming simplicity over type safety.
Only academic languages experimented with it.
Today, it's in plain Java.
We'll get there.
Hoping for the best:
More information about the talk