[ntpwg] Re: Reworded Algorithm Draft

["William Kasch" <[e-mail]>]

Yaakov,

In light of your observations, some help rewording the draft from the entire
group is requested.  Furthermore, based on your recommendations, I submit
that we follow the following process for "rewording":

1.  Completely remove all descriptive text that could be related back to the
book (essentially most of the verbiage in the document)
2.  Provide state machine and variables list only
3.  Reference the book heavily for more descriptive algorithm
documentation.

There isn't any way around this if what you're saying is true.  I'd be happy
to take other suggestions if anyone has any, as well as integrate any other
rewording of the document that anyone else would like to provide.  If not,
I'm proceeding this way for the 02 revision.

Thanks,
Bill Kasch


On 4/20/06, Yaakov Stein <[e-mail]> wrote:
>
>  Bill,
>
> I am cc'ing the list as I believe that my remarks may be of general
> interest to the entire NTP WG.
>
> I read through the updated version of the algorithms document that you
> sent
> and compared with http://www.ietf.org/internet-drafts/draft-ietf-ntp-ntpv4-algorithms-01.txt
> .
>
> In the following I am assuming that the latter is essentially the same as
> the material
> in the NTP book, and that the copyright of the book is held by the
> publisher.
> If either of these assumptions is incorrect, then the following analysis
> may be as well.
>
> My conclusion is that the edits, while definitely improving the
> readability of the text,
>  (although I have technical questions about a few changes)
> do not change the potentially infringing status of the draft.
>
> Let me explain why.
>
> When considering whether a given work infringes the copyrights of a
> protected work,
> there are three main criteria:
> 1) a SUBSTANTIAL PART of the protected work must be copied
> 2) there must be OBJECTIVE SIMILARITY between the work and the protected
> work
> 3) a CAUSAL CONNECTION must be proved.
>
> The first criterion means that the work in question must copy all
> or a large part of the protected work. That is why book reviews can
> reveal the contents and other authors can quote a passage or two
> without infringing. In the present case the copying is clearly of a
> substantial portion of the protected work.
>
> It is important to realize that the second criterion does not
> mean word-per-word identity - the criterion is qualitative not
> quantitative.
> In fact derivative works, such as translations or plays based on novels,
> do require the copyright owner's permission (they are specifically
> designated by the Berne convention as part of the bundle of exclusive
> rights).
> It would be up to a court to decide in a particular case,
> but in the present case it is clear that there IS objective similarity
> (in fact, it is so similar in purpose, structure, and content
> that I wouldn't even classify it as a derivative work).
>
> Causal connection means that the author of the work in question
> usually needs to have seen the protected work and to have based his work
> on it. There are loop holes in both directions here, in some jurisdictions
> making an architectural plan from a building would be infringing the
> copyrights of the original plan, even if the draftsman never saw the
> original plan. In the other direction there is a defense of subconscious copying
>
> if the author saw the protected work in the past but it was not before him
> while writing the new work, and he claims that he did not realize that
> he was copying. In the present case we have a clear email trail
> establishing
> that the draft author had a copy of the material before him and based
> his draft on it.
>
> Now there is a strong defense that I need to mention.
> If it is claimed that C is a copy of protected work B,
> C's author can show that in fact both are copies of an earlier work A
> (assuming that A is in the public domain or C's author
> has received permission to use A).
> In the present case I had hoped that the new draft and the book
> were derivative works of the NTP C-code (which I am assuming
> is in the public domain, or alternatively that the author would
> grant us permission to make a derivative work based on the code).
> Unfortunately, this does not seem to be the case here.
>
> So I am afraid that the proposed draft version is no better than
> the previous one in this regard. You should take into account
> that copyright protection extends to expressions but not to ideas
> or mathematical concepts, and thus it is perfectly possible to write a
> draft
> detailing the algorithms WITHOUT infringing the rights of the publisher.
> But such a work must start from public domain sources or
> from sources for which rights have been granted to the IETF.
>
> Y(J)S
>
>
>
>
>
>
>