[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [odin-devel] [LONG] Broken (lib)vista copyright



Hallo everyone,

I am currently in the process of checking out how to proceed.
I now have the official permission by the Max-Planck-Institute
to release our code and software.

Specifically, we have developed two packages:
1. a package called 'lipsia' which we have developed
for the analysis of functional magnetic resonance data,
and 2. a package called 'via' for 3D image analysis.
'lipsia' is linked to 'via', and both 'via' and 'lipsia' use the
'vista' data format. We have reduced the 'vista'-library to
its essentials and have added a new data type called
'volumes' for 3D image analysis. The vista-code
is a very small fraction of the entire code.
Over the years, we have made some additions
and changes to 'vista' independent of what Gert did.

I do not want to risk any problems with 'lipsia' or 'via'
since both packages are currently used by many researchers
in our institute.
Our software has been in use for many years (since 1999)
and is very well tested. It was used and cited in far more than
100 international publications so that maintaining its integrity
is a primary concern.
In order to avoid having to test 'lipsia' and 'via' all over again,
I think that changes to our code should be avoided if at all possible.

I would be glad if Od1n would also work with our version of
the vista-library. Perhaps the best way to proceed is to
check that out first ?

Best wishes
Gaby





Andreas Tille wrote:
> On Fri, 1 Feb 2008, Gert Wollny wrote:
> 
>> About the status of my fork (in part repeated from my last post):
> 
> 
> Thanks for the detailed explanation.
> 
>> If we don't wait what comes out with Gabi's code, I'd suggest, to check
>> out whether Od1n can be build with the vistaio library (I'd do it
>> within the next week). If this is possible without touching real code, I
>> would then backport my additions from the mia2-included version, add the
>> changes Thies did in his version, remove the debian-dir, and send the
>> needed Od1n patches to him (Hopefully only a bunch of changed includes).
> 
> 
> Removing the debian dir is always welcome because it just leads to
> confusion with the debian dir that is maintained by the Debian developer.
> 
>> I'd call the library vistaio to make clear that it is stripped down and
>> use CMake instead of autotools, because, (A) I need to build a MS
>> Windows version outside cygwin/mingw, and (B) current cmake-cvs which
>> will eventually become cmake 2.6 will have a deb-package generator:
>> http://www.cmake.org/Wiki/CMake:CPackPackageGenerators#DEB_.28UNIX_only.29
>>
> 
> 
> If cmake is able to build static and dynamic library it is fine.  If this
> is not possible we Debian people have finally to deal with it - if this is
> the price to merge two forks we might pay it.  It might be a good idea to
> keep the automake stuff stored anywhere which would allow us to add it as
> a patch for the Debian package afterwards and leaves us free to ignore
> cmake.
> 
>> Opinions?
> 
> 
> I'd say go for it.  Once this is accomplished we could turn back to
> Gabi's version and see what she thinks.
> 
> Kind regards
> 
>            Andreas.
> 
> PS: Does anybody know Gabi's e-mail address and can keep her informed?


-- 
-------------------------------------------------------------
PD Dr. Gabriele Lohmann
Max Planck Institute for Human Cognitive and Brain Sciences
Stephanstr. 1a, 04103 Leipzig, Germany
Phone: +49-341-35521738, Fax: +49-341-35521740
Email: lohmann@cbs.mpg.de
-------------------------------------------------------------


Reply to: