[Insight-developers] Patented code

[Jay Udupa]

Terry:

I would like for us to consider again the codes whose algorithms are 
patented to be treated differently. I don't quite remember as to when 
and how we agreed upon abolishing that separate treatment. I can't make 
an independent decision on FC (although I wished I could) implementation 
being made freely available. FC is the property of Penn. I will not be 
able to join the t-con on this Friday because of prior commitments. I 
can join the discussion if this is held next week. Please let me know.

Jay
_____

Terry Yoo wrote:

> Stephen R. Aylward wrote:
>
>> Also, to clarify - all code in ITK is considered open-source. You can 
>> read the terms at the top of any of the header files. It grants the 
>> distribution, modification, and use of the code. That has been the 
>> case from the start for ITK. We have never agreed on what to do with 
>> patented code. There is no provision for patent code in ITK.
>>
>> The terms of distribution given at the top of every file in ITK are
>> "Redistribution and use in source and binary forms, with or without
>> modification, are permitted provided that the following conditions 
>> are met:"
>>
>> The conditions do not limit use in any way. If a method is patented, 
>> the terms agreed to by submitting the code to ITK grants the world 
>> the right to use the method in any form, without paying a fee, 
>> including use for commercial purposes.
>>
>> Stephen
>>
>> Jay Udupa wrote:
>>
>>>
>>> I thought there is a mechanism that has already been incorporated to 
>>> handle implementations of patented methods. Fuzzy connenctedness 
>>> definitely belongs to that category since I hold a patent on that 
>>> method.
>>>
>>> Jay Udupa
>>> ____
>>>
>
> Stephen,
>
> I believe that Jay is essentially correct in his impression.
> We *did* consider holding a patented directory which included
> those elements that were encumbered by patents. We discussed
> the need to keep those elements separate and to build no
> dependencies within the toolkit upon anything that resided
> in that directory. Essentially, it requires two builds for
> every test... "with patented" and "without patented." This
> was essentially modeled after a version of VTK which has/had
> this provision mostly to provide Marching Cubes, which *is*
> patented. Bill tells me that he intends to hold a party the
> day the Marching Cubes patent expires.
>
> After a while, since no one requested that anything be placed
> in that directory, we concluded at one of our quarterly
> meetings to abolish the patented directory. Jay, that was
> when the understanding changed. I believe you were present
> when this happened. At no time has anyone requested that
> their modules be set aside in the patented directory, so
> the unused provision was essentially dropped from consideration.
> It appears that we must reconsider this notion.
>
> The consortium is strongly directing its focus toward open
> sources copyrights and licensing without restrictions. I
> strongly endorse this idea, but realistically I know that
> we must deal with patented methods.
>
> Luis points out that there are as many as four methods that
> we must consider under this discussion. I suggest that
> we take up this issue on the t-con and direct some energy
> toward resolving this issue. Luis has said that there was
> considerable interest in an open-source active shape
> implementation, which was in fact the exact point of
> making the A2D2 awards.
>
> This is definitely not the phun part.
>
> Talk to you on Friday.
>
> Terry
>
>