Re: #1198 Proposed resolution - Implementation and derivative work

At 4:13 PM +0100 2/22/06, Simon Josefsson wrote:
>Harald wrote:
> >>> This is a WG Last Call on issue #1198, with a timeout of one week, and
>>>> I'll close it on February 28.
>>>>
>>>> SUGGESTED RESOLUTION:
>>>> ----------------------------------------------------------------
>>>> Implementations of IETF standards that incorporate pieces of IETF
>>>> documents (such as MIBs or data tables) need to have the right to
>>>> produce derivative works based on those pieces.
>>>
>>> FWIW, I agree with this solution.
>>>
>>> I note that some may consider 'text' to be something that you would
>>> have to incorporate in an implementation in order to implement it.  My
>>> point is that in the above, it may have been implied that only 'code'
>>> is applicable for this rule.  I'm not sure if this was intentional.
>>
>> It was very much intentional that I wrote "pieces" and not "code" -
>> that allows us to close this issue without having to solve the "what
>> is code" issue first....
>
>Ok.
>
>I suggest to add that the derivative works should be possible to
>release under licenses which are frequently used for IETF standards.
>For example:
>
> Implementations of IETF standards that incorporate pieces of IETF
> documents (such as MIBs or data tables) need to have the right to
> produce derivative works based on those pieces, and those rights
> need to be compatible with a wide range of licenses commonly
> used by implementers.

I strongly prefer Harald's formulation.  It says what the implementations
need and commits us to making sure that the IETF delivers it.

Simon's reformulation shifts from what the implementations need
to what their licenses need.  That is at least a different issue, and it
should be tracked and resolved separately.

>Without this modification, it is possible design a IETF policy that
>satisfy the above resolution text but still fail my intended point
>here: that free implementations may need to incorporate parts of a
>standard, and to release the work under a license that permits

Please, Simon, recognize that  you have two intended points:

1)implementations may need to incorporate parts of a standard

2) some implementors have chosen licensing terms that permit
modifications or mandate that successor code be licensed such
that modifications are permitted.

Harald has suggested we separate this and resolve the first issue,
around which there seems to be little controversy.  That doesn't
mean that the second won't be tackled, but it means that we
get the item for which there is consensus nailed down now.

			regards,
				Ted Hardie