[Freebase-discuss] Guidelines for adding data?

[Gordon Mackenzie]

Just a quick note, though Jamie will respond more completely.

We are aware of some community confusion engendered over namespaces,
which are a result of not of any overt strict policy of usage or by
fiat decreased levels of restriction by staff but by the lack of a
declared policy for the most part. Only some namespaces have been
locked down, primarily because they were created out of internal
projects and the quality of data would not and is not reasonably be
exceeded by the broader community. Some were locked down because the
number of keys liked to the namespace was finite, fixed and
considerred considered complete. A few were restricted because of
vandalism and again their importance, at least in the past if not
currently, with usually an internal project. Unfortunately a few were
set to be restricted due to a staff member(s) mistaken or incorrect
assumptions. Hopefully the latter were changed to be open to the
community at large for adoption and assertion of values.

We will not change permissions on a user-created namespaces, unless we
are requested to do so by that user (and maybe after community
consultation?).

Web ID editing will hopefully be made easier (ie in the client) on an
instance level sometime in a future release of the client topic page
(no ETA at this time I'm afraid, but I am always actively pushing for
it!), lists of keys for a namespace are best asserted through a
vetting process like creating a set of triples (
http://wiki.freebase.com/wiki/Triple_Loader ) and/or QC'ed on sandbox
like with Refine+Refinery ( http://wiki.freebase.com/wiki/Refinery ).

A recent example of mistaken impression about ISBNs being restricted
for book editions was incorrect, there is a known bug with client in
trying to reassert an existing soft ISBN value but ISBNs by themselves
are not restricted to entry by only a few administrators.

We apologize for any confusion in the past, and if there are
namespaces currently restricted on Freebase and the community has
questions about access please contact us, ideally through this list.

On Tue, Feb 1, 2011 at 8:42 AM, Tom Morris <tfmorris at gmail.com> wrote:
> p.s.  It's implicit in my comment about aliases/soft links, but just
> to make it explicit, there's really nothing magical about names from a
> technical point of view -- either from the point of view of using them
> or from discovering them in Freebase.
>
> Here's a namespace that I created to hold the identifiers for all U.S.
> Congresspeople:
> http://www.freebase.com/inspect/base/uspolitician/thomas_id
>
> I called it /base/uspolitician/thomas_id (putting it in a "base" I'd
> created and had write privileges for rather than my personal
> namespace), and that name/id still works, but it's also known as:
>
> /authority/us/gov/loc/thomas_id
> /authority/sunlight/bioguide_id
> /authority/us/gov/congress/bioguide
> /user/jamie/sunlight/bioguide_id
>
> Any of these will work and will give you the same results.
>
> I haven't written to the namespace recently, but by inspection it
> still has the same permission node I gave it
> (http://www.freebase.com/inspect/m/05kngtl), so I suspect I've still
> got write privileges.  I could lose these at any time, but I could
> also lose write privileges for /user/tfmorris at any time.  It's a
> right that Google has granted me that it's free to revoke at any time.
>  That's kind of the nature of free services.  Of course, I get to keep
> my data no matter what, but I'd have to find my own hosting
> infrastructure.
>
> Tom
> _______________________________________________
> You are receiving this message because you are subscribed to the Freebase-discuss mailing list.
> To post a message to the list: Freebase-discuss at freebase.com
> To unsubscribe, view archives, etc: http://lists.freebase.com/mailman/listinfo/freebase-discuss
>