[Incubator] Mapbender Incubator Process

[Frank Warmerdam]

Rich Steele wrote:
> Yes, I am referring to that copyright and permission notice.  The
> requirement that the notices "be included in all copies or substantial
> copies of the software" means just that.  We certainly do this with
> MIT-licensed code, as do others (see e.g.,
> http://www.paradigmsim.com/support/sc_view/Vega_Prime_Release_Notes_2.0.
> 2.pdf for another corporate user (and attributor) of Proj.4).

Rich,

I am gob-smacked.  For years I have been explaining to people that
the license means they can't strip the notice out of the source code,
but that it doesn't place any requirements on their distribution of
binary software.

This is horrible.

> I think if you are not reproducing the notices, then strictly speaking
> you are not complying with the license.  Since this is really the only
> requirement of the MIT license, it isn't a lot to ask for free software.
> My friend Heather wrote a good overview of the Notices conundrum:
> http://www.linuxinsider.com/story/48494.html

Well, this is a very interesting and spot-on discussion.  I feel that
Heather did not fully point out that many developers (such as myself)
never *intended* to require the credit in binary software, only in
the source.

In the case of GDAL there are many copyright holders, and presumably
complete compliance would require listing all of them, with complete
copies of the license terms for each, is that technically correct?

What approach should we take to this?

I could strip out the "notice" portion of the license for all files that
I am the copyright holder of, but I can't alter the files held by other
parties, and getting them all to agree would be an arduous process.
Would removing it from as many as possible be a worthwhile thing?

The further one goes "down the stack" the more of an issue this is.  So
for instance, fixing it in GDAL, PROJ.4, libtiff, libgeotiff, and libjpeg
might offer value to many projects.

Let's imagine I was interested in relicensing code under a new license
without the notice requirement of MIT/X, but also none of the "viral"
aspects of LGPL or GPL, what would you suggest?   Is there a "best practices"
license that would be appropriate to apply?  Hopefully one that isn't
onerously long and complex?

I'm basically looking for a license that is roughly public domain.  In
fact I would be happy enough to employ a license that would allow users
of my code to re-license it any way they please, as long as it doesn't
impact my and other licensee's rights to do with the software as we please.

/me needs to process this situation a bit, and check on supper.

Best regards,
-- 
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | President OSGF, http://osgeo.org