Subject: Re: HTTP/1.1 RFC copyright statement
From: Alex Rousskov <rousskov@measurement-factory.com>
Date: Fri, 17 Sep 2004 11:42:00 -0600 (MDT)

On Fri, 17 Sep 2004, David Van Horn wrote:

>> Technically, if you extract a piece of software from RFC 2616, then 
>> your ability to create derivative works seems to be limited to 
>> "works that comment on or otherwise explain it or assist in its 
>> implementation". The latter is pretty broad, but can be viewed as a 
>> "field of endeavor" restriction by OSI.
>
> My understanding is that this is not a restriction, but this is at 
> the heart of my question.

If there was no restriction intended, instead of saying
 	"derivative works that ..."
the text would say
 	"any derivative works"
or
 	"any derivative works, including works that ..."

Thus, IMHO, there is a restriction. I have no idea why Internet 
Society lawyers(?) put it in there.

Technically, if I extract an embedded reference implementation of 
protocol P and distribute its modification that has nothing to do with 
protocol P (but uses some useful and large piece of modified reference 
code), I would be in violation of a copyright because my derivative 
work does not "comment on or otherwise explain it or assist in its 
implementation".

While I feel relatively safe because I have enough trust in Internet 
Society intent, your situation is worse because your SRFIs are owned 
by individual authors.

> Our *intent* has been that these documents, as are the pieces of 
> software contained within them, are free, although this has been a 
> subject of contention.  We'd be happy to use a license more widely 
> agreed upon as being a free license, but there are close to 60 
> documents carrying the current RFC-like statement, and contacting 
> the authors in order to change the statement is infeasible.

Well, you can hope that OSI opinion would differ from mine :-/.
You may have to formally submit the license for approval to
get a formal verdict. That would be a little weird because
it is a documentation license though.

If you do not get OSI approval, you can at least change the license 
for future SRFIs to one of the existing OSI-certified licenses. It may 
be a good idea to give _both_ "free documentation" and "open software" 
licenses.

The situation is somewhat better for IETF because Internet Society 
owns the copyright, not RFC authors. While I doubt it will happen, I 
suspect that Internet Society can change their copyright statement to 
make it broader and apply the new broader statement to old RFCs.

TINLA,

Alex.