This is the mail archive of the ecos-discuss@sources.redhat.com mailing list for the eCos project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Licensing of OpenSource code and eCos


On Wed, Sep 25, 2002 at 09:43:51PM +0200, Iztok Zupet wrote:
> OK, let's start a new thread (who's maintaining CVS).

I am not sure there is a lot of enthousiasm for that, but anyway.

> The fact is, that eCos is a self contained and self evolved product,
> thus it should be published under the eCos license (as the whole!), to
> whatever that license evolves (and whoever is the holder).

I do not clearly understand what you try to explain.

> GPL: can't be used here without the exception, because it prohibits
> linking of the eCos with the end user application.

This is correct if the end user application is not GPL. In most
commercial projects you don't want the end user app to be GPL.

> It's hard to get an
> approval from the FSF to publish some GPL-ed code under the eCos GPL
> with the exception. Only the RMS at FSF can give such an approval.

The philosophy of FSF is different and very precise on this matter. You 
can read the motivation of the FSF on this page:
  http://www.gnu.org/philosophy/why-not-lgpl.html
This discussion on LGPL defines exactly why FSF strongly prefers to use 
GPL as much as possible and avoid an LGPL (and for that matter ECOS-style)
license when possible. This is a deliberate, political choice for "Freedom" 
as defined by FSF.

> There can be a workaround if all the original authors of the GPL-ed
> code agree to publish it under the eCos GPL. As far as I know they have
> been granted that right when they assigned the copyright to FSF.

It is very impractical to trace back all authors and ask them to republish
the work under another license. This why the holder of centralized 
Copyright of GPL-style licenses (including the ECOS license) has a large 
advantage over the spread-out group of individuals that Assigned Copyright 
to a central Copyright holder. We could argue if that advantage for the 
central Copyright holder is a fair advantage or not. It may be a fair 
advantage to compensate for the development effort that the central 
Copyright holder has invested in the project to publish Free Software, 
but then again, this advantage can only be valorized into money, by 
making a version of the program that is (at least temporarily) 
proprietary or non-Copyleft. While the money that is recouped by 
selling/developing a non-Free version helps the centralized Copyright 
holder to further develop the Free Software code base, the specific 
code that was written for the customer or by the customer itself is 
not available for the Free Software community and this is a pity.

A little off-topic, but relevant in this light: In a discussion yesterday 
on modified BSD, we came to the interesting conclusion that in modified 
BSD, Copyright Assignment does not show this problem. In modified BSD, 
every single person, also a small contributor that signed a Copyright 
Assigned to e.g.  the Apache Foundation, still has the same right as 
the Apache Foundation itself, namely to derive a proprietary work from 
the complete public code base. 

> BSD: It seems OK to publish the BSD code under the eCos public license,
> as long as the original BSD license is also in-there.

This is what I believe personally. Am I correct that there could be two
possible implementations ?

- what was originally mod-BSD is "upgraded" to the ECOS license
  (this is a one way path, because the ECOS license is Copyleft 
   and mod-BSD is not)
  
- mod-BSD code is included in the project as separate files, retaining 
  the mod-BSD license
  (I believe this is allowed since the ECOS exactly limits the scope of 
   the Copyleft to the work itself and excludes other works that together
   form the "whole").

Jonathan, I did not understand exactly under which of these option the
inclusion of the lwIP as established mod-BSD licensed work could occur
in the eCos tree.

> PS.: Am I wrong? According to the copyright law, if the original code is
> modified by more than 20%, the one who modified it that way can claim
> copyright.

I am not aware of that.

Peter

> Regards
> iz

-- 
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]