[okfn-discuss] Is the Open Directory Project open?

Thu Aug 30 09:36:23 UTC 2007

Luis Villa wrote:
> On 8/29/07, Jonathan <jonathan.gray at okfn.org> wrote:
>> Thanks for your reply Mike!
>>
>> I had also been wondering about Section 4. which seems to be the one
>> that the FSF took issue with:
>>
>>   "4. Errors and Changes.  From time to time Netscape may elect to post
>> on the page(s) at the URL http://dmoz.org/become_an_editor certain
>> specific changes to the Open Directory and/or above attribution
>> statements, which changes may be to correct errors and/or remove content
>> alleged to be improperly in the Open Directory.  So long as you are
>> exercising the license to Open Directory hereunder, you agree to use
>> commercially reasonable efforts to check the page(s) at the URL
>> http://dmoz.org/become_an_editor from time to time, and to use
>> commercially reasonable efforts to make the changes/corrections/deletion
>> of content from the Open Directory and/or attribution statements as may
>> be indicated at such URL.  Any changes to the Open Directory content
>> posted at the page(s) at the URL http://dmoz.org/become_an_editor are
>> part of Open Directory."
>>
>> I suppose the issue hinges on the phrase "commercially reasonable".
>>
>> I'm still not entirely clear on whether or not this is compatible with
>> the OKD (http://www.opendefinition.org/1.0) - specifically, if it is
>> not, which section of the definition doesn't it comply with? If it looks
>> to be compatible, should it be?
> 
> If it conflicted with the OKD at all, it would presumably conflict
> with (3) Reuse, which requires that the license allow for
> modifications. I realize that I have in the past assumed that both OKD
> and OSI definitions require that the license must allow for *any type*
> of modification, but neither definition actually has language that
> would require that.

I think it is not just for reuse but for redistribution (item 2) that 
there are problems. An OKD-compatibile license should allow me to do 
what I like with the material once I obtain it (including continuing to 
distribute it/reuse it etc). That is, revocation is not allowed. 
Netscape's intent seems to be to allow them to remove 'bad' material 
(e.g. copyright violating, defamatory etc). However one worries that is 
also allows Netscape were to suddenly remove all content from the Open 
Directory -- would you be required to follow this update and cease 
distribution of your existing copies? Furthermore, and this seems to be 
the FSF's beef, it also imposes a rather non-trivial requirement on 
those who wish to reuse and redistribute the material (and plays merry 
hell with any packaging efforts since the material will never remain 
'stable').

>> Presumably the fact that the ODP service runs on proprietary software
>> doesn't in itself mean that the data can't be open, though it does,
>> according to the provisional open service definition, mean the service
>> isn't open.
> 
> Right.

I think Luis is exactly right here.

>> Also if the data is not fully RDF compliant [1] - does this constitute a
>> technological restriction to access?
> 
> If perfect standards compliance is a requirement for OKD openness
> we're in trouble :)

Provision 4 of the OKD, "Absence of Technological Restriction" currently 
states:

"The work must be provided in such a form that there are no 
technological obstacles to the performance of the above activities. This 
can be achieved by the provision of the work in an open data format, 
i.e. one whose specification is publicly and freely available and which 
places no restrictions monetary or otherwise upon its use."

This has always been one of the most controversial and most difficult 
provisions (not least because of lack of clarity as to exactly what a 
technological restriction would be). Its basic intent was to prevent 
endruns around the spirit of the definition by the use of some 
technological means (e.g. via TPMs or other encryption mechanisms). It 
was also intended to (strongly) encourage the use of 'open' formats.[^1] 
However it does not mandate them (hence the "*can* be achieved") and in 
a case like this where the material is provided as slightly broken 
RDF/XML while not perfect, IMO, would not render it non-compliant with 
this provision.

~rufus

[^1]: a companion project, which has now been inactive for rather a 
while, was the information accessibility initiative http://okfn.org/iai/ 
which detailed what an open format was and provided example listings).