BSD-like licenses and the OSI approval process

[Chris Travers]

On 10/16/07, Lawrence Rosen <lrosen at rosenlaw.com> wrote:
> I will restate my point yet one more time so Chris can better understand his
> options.
>
> I don't understand why Chris' business is hurt because he can't claim that
> each of the components of *his* software are under OSI-approved BSD-like
> licenses. Who cares about that? Chris is already free to:
>
> 1) Combine all those BSD-licensed components into *his* software.

There are both technical and logistical reasons why this will not
work.  For example, conflicts regarding package management, and the
unwillingness to ask people to download many MB extra of software in
the initial download.

>
> 2) Release *his* software under an existing OSI-approved license, and not
> worry about the licenses on the components. *His* software will be open
> source as long as he really does provide all that source code!


My concern is that there seems to be pressure (Mr Tiemann's blog for
example) on vendors to either call it something other than "open
source" or use OSI-approved licenses.  If I want to say, "depends only
on open source software" I am worried about pressure in the future.

If we all agree that there is no "open source" brand that the OSI is
"protecting" then the problem goes away.  But unless that happens,
there isn't a lot of room to maneuver.

In short, I think that OSI needs to choose either to abandon the
question of "protecting" the "open source" brand, abandon the worry of
license proliferation, or look for a structure which will allow for
some attempt to balance these.  Otherwise, vendors like me get pushed
into a corner regarding public statements by board members.

>
> As long as Chris complies with the terms of all those other BSD-like
> licenses on the components he uses, he shouldn't care one whit whether
> PostgreSQL or those other licenses actually get OSI approval.

My issue is that OSI approval is being positioned as being a
requirement for calling a project "open source."  If that goes away,
then I have no interest in getting the license approved.  If it does
not, then I have every interest in either doing so or publicly and
loudly challenging the right of the OSI to make that connection.  If
you don't mind, I would rather work with you and avoid the
confrontation altogether (or at least limit it to this list).
>
> Not only that, Chris, but I already suggested that you use AFL 3.0 for
> *your* business software and stop tying up this list with approval requests
> for all those other old and inadequate BSD-like licenses!

Our business software is licensed under the GPL v2 or later because of
inherited code.  We have no plans to change this.  We briefly
considered the OSL, but I opposed that because I felt that it would
make it harder to make our official download sites *the* primary and
authoritative point of download (I also generally do not like the
external deployment clause for other reasons).  However, we depend on
software licensed under other licenses (all permissive licenses).
Some of this software is widely distributed separately, so we don;'t
want to duplicate the effort.

Unlike you, I do not think that the BSD License gives me the right to
take a verbatim copy of the software and simply change the license, so
a separate distribution of such dependencies would be also out of the
question (at the moment PostgreSQL is the only component which poses
this problem).  The PostgreSQL core team also objected to such an
idea.

I am sorry, but your license change solution does not meet my needs.

Best WIshes,
Chris Travers