Fwd: netX re-iteration to answer open questions

[Bruce Perens]

-------- Original Message --------
Subject: 	netX re-iteration to answer open questions
Date: 	Tue, 07 Dec 2010 22:17:12 -0800
From: 	Bruce Perens <bruce at perens.com>
Organization: 	Perens LLC
To: 	Russ Nelson <nelson at crynwr.com>, license-discuss at opensource.org, 
Christian Solmecke <solmecke at wbs-law.de>



>  On 12/05/2010 09:56 PM, Russ Nelson wrote:
  >  >  At the end of the cited thread, he asks two questions that nobody
  >  >  really answered.

OK. I guess I should apologize then. I tossed them off as irrelevant in 
the face of the larger issue with the license.

    1. Technology-Neutral
    As we have seen, in particular the special treatment of the kernel interface is partially
    considered as a violation of # 10 (License must be Technology-Neutral) of the Open Source
    Definition. This is why we would clarify, that the purpose of the License is not aim
    to create a special treatment of the kernel interface. Rather, the aim is the porting
    of the licensed software to other operating systems in order to guarantee the distribution
    of the software throughout the Community. Therefore the License asks to publish all
    changes concerning the porting to other operating systems for distributing purposes.
    All other changes or adaptations can, but do not necessarily need to be published. We
    would therefore ask you, what do you think is the right way to describe this issue in
    the License. We welcome any feedback and concrete proposals.
       

The recent adoption of the ZFS filesystem by the FSF's GRUB project, 
discussed at http://lwn.net/Articles/418869/
is a relevant example to this discussion. In that case, Sun/Oracle did 
not provide the ZFS filesystem driver to the GRUB project, but simply 
distributed the software in compliance with the GPL. The project picked 
up the distributed software and incorporated it, with the knowledge that 
Sun/Oracle's distribution of the binary and source code gave them 
sufficient rights to do so.

My experience, and it seems that of the other folks on this list, is 
that the upload provision is simply not necessary. You'll get the useful 
software without it. It's an unnecessary hardship upon the developer and 
needs to be removed from the proposed license.

Thus, my response to your question #1 is: The way to describe the issue 
is moot because the provision under discussion would be unacceptable in 
the license in any case.

    In addition we would ask you, if it makes any difference, if we prospectively would
    do not distinguish between the styles of interfaces. Accordingly, ANY changes of the
    program - regardless of whether the changes affect the Kernel or not - must be uploaded
    to a website. Although this would be a much stricter handling, in our opinion the requirement
    of the technological neutrality would be guaranteed. Do you think, this could be a way
    to resolve the problem concerning the technological neutrality?

    2. Mandatory Upload on a website
    Furthermore we ask ourselves, which provision of the OS-Definition is infringed by the
    mandatory upload of the changes of the Program
       

#5: No discrimination against persons or groups. The license would 
discriminate against developers who do not have access to the internet. 
Both those who have no access, and those who are out of access for a 
time during which they would like to distribute the software, for 
example if they are working in a disadvantaged area. In contrast, the 
OSI-approved licenses can be complied with without access to any 
particular network or site.

    , which is demanded by the NetX-Public
    License.  To our mind, the reference duties concerning any changes and modifications
    of the Program (in particular concerning the Kernel) according to § 4 clause 1 and 2
    netX Public License do not infringe the Open Source Definition.
       

Well, the real reason for disapproving of them is that they are an 
unnecessary burden. But it's trivial to make a case that the provision 
is discriminatory.

    According to # 4 Open
    Source Definition the License can request the documentation of changes and modifications
    to the Program in order to preserve the integrity of the author´s source code.
       

No, that's not what #4 says. It sets out one very narrow exception to 
the requirement that source code be directly modifiable, allowing 
instead the use of patch files, which is historical in nature and no 
longer necessary. It allows the license to request a change in name or 
version number to point out a modified version in the distributed copy. 
It requires that distribution of binaries built from modified source 
code be permitted. That is all that #4 does.

    Perhaps we can overcome your concerns, if the licensor engages himself to provide another
    suitable website, in case the websitewww.industrialNETworX  goes away.
       

In my opinion, it's the requirement for notification that is odious, not the site or number of sites.

    3. Renaming of the License
    Finally, we would like to follow the suggestion of Matthew Flaschen to rename the license
    in order to avoid confusion with other similar programs. Therefore we have the intention
    to rename the License to "industrialnetworx Public License"
       

Understood.

     Thanks

     Bruce