[pd-discuss] Canada PD calculation code

Tue Jun 9 16:12:43 UTC 2009

2009/6/5 MARGONI THOMAS <thomas.margoni at unitn.it>:
> Hi all,
>
> I know that this would turn me quite unpopular, but I'd like to repeat what we
> (or at least I) have been arguing in London.

Not at all unpopular and it is good to have a critique!

> The problem to me, which makes me feel quite uncomfortable with, is that you
> talk about something that is not defined. One "calculates" the Public Domain,
> but for what reason? I guess, the answer is to give information about the type
> of uses one can do with a specific work. In fact, if we ascertain that it is in
> the PD then people can do... what?

I think we should distinguish (at least) 2 purposes:

1. Determining PD status for a particular item (this book, this
recording) so that you know what you can do with it (in particular can
we post it freely online). In doing this "automatically" one may only
be able to give approximate answers, e.g. saying: "we are 99.9% sure
this is in the PD" but if you want 100% certainty you'll need to ask a
lawyer.

2. Doing statistical analyses where we interested in working out the
general "size" of the PD. Here we may be willing to tolerate
imperfections and ignore edge cases if we are right more than 99% (or
even 95%) of the time.

> Depends. In fact if a work is in the PD in France may be used in different ways
> than if it were in the Netherlands, or Belgium or Canada, or US. The fact that
> 70 (or 50) years p.m.a. there are still moral rights prohibiting to perform a
> given number of uses is something that cannot be forgotten. And the amount of
> time during which they last (somewhere they don't exist, or expire along with
> the economic, or after N years, or never, being perpetual) is also an aspect
> that cannot be left a part.

Right, but, for example, in almost any jurisdiction simple posting
online for others to download will be permissible for PD material
(moral rights won't affect that). Moreover I don't see why we can't
have a warning for e.g. French jurisdictions saying: "Be careful about
moral rights". And again I don't think we should let the impossibility
of perfection prevent from us doing something pretty good.

> Regarding Crown copyright it is pretty much the same. Crown copyright is typical
> of all Commonwealth countries, and it tells us a characteristic that regards the
> authorship/ownership aspect. It tells us that a specific category of works
> created by government's employees of UK, Australia, New Zealand, Canada, etc
> belongs to the Queen of England. But it doesn't tell us anything about the type
> of uses one can do with it. In fact, a Crown copyright work has different type
> of protection in, say, UK or Canada. If we don't clarify that, and a potential
> user from Canada is told by the calculator that a work is Crown copyright, s/he
> probably will act as the category allows him to do in Canada (where enactments
> and maybe reasons for judgements are freely reproducible), but the work could be
> under UK (where the terms are different), or New Zealand (enactments and reasons
> for judgement are Public Domain and not Crown copyright), or Australia
> (see http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1297263 for a more clear
> explanation of what I want to say...).

Right! But again this seems like an edge case: if necessary let's just
ignore all Crown Copyright material for the time being.

> Don't misunderstand me if I sound too critic, my intention is just to address an
> issue that I have pointed out in more than one occasion, with respect to a
> project that I find particularly interesting. And I believe that you see this
> point as well, as witnessed by the article of Torremans on your webpage, which I
> would put as the very first and not last link.

Good point. You know it is a wiki and anyone can edit the page!

> On a higher level of abstraction, I think the problem is that the project
> pretends to calculate, thus to measure, something that a priori refuses to
> define. The connected danger, to me, is that the project might facilitate

Do you mean we don't define the "public domain"? I think we can have a
go at that along the line of Torremans summary for the "Economic and
Social Impact of the Public Domain" project.

> information that at best is not accurate. And you all know what is the common
> outcome when the wrong informations have legal relevance...

I completely agree with you that we have got to flag uncertainty as
clearly as we can. That said I must reiterate that I don't see our
inability to obtain perfection as a reason not to produce algorithms
that are 99% accurate.

Regards,

Rufus