[okfn-discuss] Fwd: [Autonomo.us] Dmoz

Rufus Pollock rufus.pollock at okfn.org
Mon Nov 24 18:30:41 UTC 2008


My bad, I ended up answering (affirmatively) the question: is this
data open? not the question: is this an open software service?

If, as it appears, the software is not free/open then of course it is
not an Open Software Service as defined by the OSSD
(http://www.opendefinition.org/ossd/).

More comments below.

On Mon, Nov 24, 2008 at 4:13 PM, Mike Linksvayer <ml at creativecommons.org> wrote:
> On Sun, Nov 23, 2008 at 7:52 AM, Rufus Pollock <rufus.pollock at okfn.org> wrote:
>> In my view http://dmoz.org/ does indeed meet the OSSD:
>
> What about the http://en.wikipedia.org/wiki/Dmoz#Software ???
>
>> a) Data is available (via web and in bulk)
>> b) Has a license [1] which allows for free use, reuse and
>> redistribution though, as you point out, subject to a relatively
>> 'arduous' attribution requirement. However I don't think the
>> attribution requirement is such as to render it non-open (I think the
>> test here would be: does this hinder use -- or reuse -- in any
>> significant way and here, IMO, the clear answer is no).
>>
>> [1]: http://www.dmoz.org/license.html
>
> Clause 4, Errors and Changes, makes the data non-free
> http://en.wikipedia.org/wiki/Dmoz#License_and_requirements

Yes, I saw that but given vagueness of 'commercially reasonable'
efforts I figured this didn't buy much more than an exhortation.
However, this very vagueness is a problem and one might well consider
this an overly onerous condition. I'm therefore uncertain on the 'Open
Data' requirement.

> Dmoz was a very important early open content pioneer, but I'm afraid
> it does not run on free software, does not make its non-free source
> available, and its data license has problems.  So it is non-OSSD
> compliant all around.

In this case it is definitely non-OSSD compliant. I also take your
point about the data license having problems though as I'm not really
clear what clause 4 requires I'm unsure whether it falls outside the
definition of open data. One option would be to task them to
clarify/remove -- do you think it is worth writing to them on that
front (it seems a pretty useful RDF dump).

Rufus




More information about the okfn-discuss mailing list