Re: [libreplanet-discuss] Potential of the Sleepycat License

[Michael Pagan]

On 17 Apr 2017 at 14:22, Nicolás A. Ortega wrote:
>> On Sun, Apr 16, 2017 at 11:32:58PM -0400, Michael Pagan wrote:
>> <SNIP>
>> This is the direct strategy employed by the GNU Project.  I believe RMS
>> got it right when he decided that the license that should be chosen for
>> a library should be based on whether it provides an advantage or not for
>> the free software community.  If the library does not provide an
>> advantage (i.e. the library shares the same features as non-free
>> libraries), then a "weak" license should be used; if the library *does*
>> provide an advantage (i.e. pioneering a new significant feature), then a
>> copyleft license should be used.
>> 
>> Learn more about this strategy here:
>> http://www.gnu.org/licenses/why-not-lgpl.html
>> 
>
> The issue with this kind of method is that many people are put off by
> the idea of using a GPLed library (mostly because that brings up a ton
> of other legal concerns about the other libraries that you may be
> using), so it makes things extremely messy. LGPL keeps the library code
> Free, but does not ensure (legally) that the library is only used by
> Free Software.

I think I see where you're going here.  Could it be that a copyleft
library can't be used with non-copyleft libraries in parallel?  Do you
mean that if someone "copylefts" a library that it will no longer work
with a non-copyleft library in a legal sense?  I am excluding the LGPL
From these questions when I mention /copyleft library/.

The key question here is: Does it really matter?  The copyleft library
can still be used, and if it does not work with other libraries, then
those libraries can either be: Substituted with copyleft ones, if they
exist; refactored into a single copyleft library; or maybe even forked
(as a last resort).

The above may sound like a duplication of effort, but if any of those
non-copyleft libraries ever become non-free (not saying that they will),
then the TRUE duplication of effort will begin (i.e. reverse-engineering
the proprietary features, and adding them back into our free library).

I'd rather have a guarantee of freedom (i.e. copyleft), then simply
freedom in the moment (i.e. free, but non-copyleft), no matter what the
cost; even if legal hassles and certain sacrifices (no non-copyleft
libraries, which may be a stretch for many) must be made.

Perhaps you are seeing something else of significance that I do not?
Maybe you have envisioned a scenario where using a copyleft library in
specific circumstances makes certain types of computing more difficult
due to some computational sacrifices.

> Nope, goes a little further than that. To my understanding (however it
> seems that I may have been wrong) Sleepycat is copyleft upon derivatives
> and forces source disclosure on software that uses the Sleepycat
> licensed software.

I have read the license.  _The Sleepycat license is non-copyleft_.  It
is most definitely a lax, permissive "weak" free software license.

Sleepycat behaves just like the non-copyleft Modified BSD License; thus,
it is also non-copyleft.  For further clarification, let's dissect the
license-- shall we?

Here it is in part.  I eliminated the copyright notice and duplicate
clauses (clause #3 is removed, since it deals with attribution, which is
not our main focus right now).  The first 2-clauses deal with the
distribution of a hypothetical program:

    * Redistribution and use in source and binary forms, with or without
    * modification, are permitted provided that the following conditions
    * are met:
    * 1. Redistributions of source code must retain the above copyright
    *    notice, this list of conditions and the following disclaimer.
    * 2. Redistributions in binary form must reproduce the above copyright
    *    notice, this list of conditions and the following disclaimer in the
    *    documentation and/or other materials provided with the distribution.

Because the conditions (clause #1 and clause #2) only deal with
"redistributions" (i.e. exact copies), "derivative works" or "modified
versions" are not affected by the license.  Redistribute means to
distribute again or anew.  The following represents a hypothetical
revision that would need to take place in order for the Sleepycat
license to be copyleft (the changes are in ALL CAPS):

    * 1. Redistributions of source code, AS WELL AS MODIFIED VERSIONS,
    *    must retain the above copyright notice, this list of conditions
    *    and the following disclaimer.
    * 2. Redistributions in binary form, AS WELL DISTRIBUTIONS OF
    *    BINARIES PRODUCED THROUGH THE COMPILATION OF A MODIFIED VERSION
    *    OF THE SOURCE CODE, must reproduce the above copyright notice,
    *    this list of conditions and the following disclaimer in the
    *    documentation and/or other materials provided with the
    *    distribution.

In parenthesis, I'm making it clear the obscure way that the Sleepycat
license and the Modified BSD license provides the 4 essential freedoms:

    * Redistribution (FREEDOM #2) and use in source (FREEDOM #1) and
    * binary forms (FREEDOM #0), with or without modification (FREEDOM
    * #3), are permitted provided that the following conditions are met:


Your thoughts?
-- 
Michael Pagan (pegzmasta) [71B46D72]
B942 AD12 82B1 6D14 27A0  86AE 119E 451E 71B4 6D72
Free Software Hacker, <https://savannah.nongnu.org/users/pegzmasta>

_Please consider the following before replying to me_:
Avoid *proprietary* attachments; see 
<https://www.gnu.org/philosophy/no-word-attachments.html>.
Consider using Email encryption; try <https://emailselfdefense.fsf.org/>.
Support software freedom today; join <https://my.fsf.org/join?referrer=835840>.

signature.asc
Description: PGP signature