Re: #1400 Opinion poll - question draft

[Harald Tveit Alvestrand <harald at alvestrand.no>]

John,

Summary of a long note:

If you want a third alternative on the table, send text.

Longer note:

I believe that this document *has* to state things in a way that the people 
who are trying to agree upon it understand.

And I don't believe that an interpretation of "translation" to mean the 
same as "extracted, modified, and used by anyone in any way desired" is 
useful in that context.

To recap: The use cases we've been tossing about include, among others:

- Modifications to make code compile with a specific compiler suite
- Modifications to fix bugs in the code
- Modifications to make the code run faster
- Modifications to adapt the code to specific runtime contexts

These are rights that can be seen as "required to implement the standards".
THe other groups of modifications include:

- Modifications to experiment with other ways of doing things (such as 
figuring out possible candidates for a later version of the standard)

(some have argued that such activity should only be done "within the IETF"; 
others have argued that such a distinction is not a good one to make.)

And, based on the discussion of distributing code under various open-source 
licenses:

- Modification to reuse the same code for different purposes, with or 
without release of those modifications under specified conditions.

Note that reimplementation in another programming language is *not* on my 
list; I've always thought (but without benefit of legal advice) that such a 
reimplementation would be seen as a new work with new copyrights - based on 
the older code, but not a derivative work from it. You can't do things 
exactly the same way in Perl and C. (Jorge, if you read this - please 
advise).

Details...

--On tirsdag, desember 05, 2006 22:10:13 -0500 John C Klensin 
<john-ietf at jck.com> wrote:

> --On Tuesday, 05 December, 2006 08:54 +0100 Harald Tveit
> Alvestrand <harald at alvestrand.no> wrote:
>
> > This issue is now #1400 in the tracker.
> >
> > And I think it's time for one of my (in)famous opinion polls.
> >
> > First - let's see if I can get the question right. If it's OK,
> > we'll go on to state opinions. So please do NOT spend
> > bandwidth on stating your opinion now - this is about whether
> > the question is right.
> > ...
>
> Harald, I don't think it is quite right because I think the text
> you are using as examples (both versions) does not properly
> capture one aspect of the discussions.   Without proposing
> specific text (and further complicating the discussion), my
> understanding of where we are is that we have agreed to
>
> 	(i) unlimited copying, [re]distribution, and translation
> 	of full text, with attribution.
> 	
> 	(ii) nearly the same permissions for excerpts, also on
> 	condition of attribution.
>
> and the debate is about rights in addition to those.  If that is
> not a correct formulation so far, then I am seriously confused
> and the rest of this note can probably be safely ignored.

That fits my interpretation of what issue #1400 is about, yes.

> Now, with regard to those other rights, you've made a relatively
> big deal (IMO) about adapting code for use in specific contexts,
> e.g., mapping from one programming language or data structure
> representation to another.   I think those are non-issues and
> should be dealt with by (explicitly if necessary) simply making
> them special cases of "translation" above.  You should check
> with Jorge, but I believe that there is some case law supporting
> interpreting those types of modifications as "translation", even
> without very explicit definitions.
>
> I assert, without taking a position as to whether it is the
> right thing to do, that, with a proper definition of
> "translation", all of the rights needed to strictly implement a
> standard are covered by "copy, extract, distribute, and
> translate" or words to that effect.
>
> An assertion that a standard will not be implemented unless
> additional rights are available appears to me to require a
> separate question, noting that similar assertions have been made
> about IPR options and restrictions (such as RAND on underlying
> patents) that we explicitly permit.
>
> If that separation were made, then the question you would be
> asking would be about precisely the two issues where the
> controversy lies, i.e.,
>
> (i) whether unlimited modification of "code", for any purpose,
> is to be permitted, and
>
> (ii) if so, whether that is best accomplished by the IETF
> obtaining and granting "outbound rights" or whether it is better
> accomplished by the type of parallel licensing mechanism that
> has been discussed on and off since San Diego.

I believe that the question on the table is on whether unlimited 
modification of "code" is the desire of the IETF.

The poll text as written is intended to figure out whether the IETF prefers 
that formulation, or a slightly more restrictive formulation suggested by 
Frank. If you want a third alternative on the list, please send text.

Once we have consensus on that (your (i)), we can discuss mechanisms for 
accomplishing it (your (ii)), but I don't believe that needs to be 
considered as part of #1400.

                      Harald


_______________________________________________
Ipr-wg mailing list
Ipr-wg at ietf.org
https://www1.ietf.org/mailman/listinfo/ipr-wg