I find the whole thing hugely exciting. Congratulations on just doing
it, without getting a committee together to study it. :)

It seems to me that asking people to help FRBRize things on a system
with no other purpose isn't going to get a lot of traction. People
FRBRize on LibraryThing because it improves the system for their
books, and when you do it you get immediate changes to the data you
were looking at—all of a sudden you can reviews for ALL the editions
of X. Will people spend time typing in ISBNs and fiddling with the
FRBRization around it? I just don't buy it.

It seems to me this functionality should be embedded in other
applications—LibraryThing, for example, or WorldCat, or really any
library catalog. How about popping OpenFRBR data up whenever you're on
a page in Amazon, a la (and perhaps integrated with) LibX, etc.?

What are the plans to use existing data sets? I'm interested in
working something out, but LibraryThing's xISBN service has limits,
including non-commerciality, which the MIT license doesn't have. The
same is true for OCLC, who are actually trying to sell the service
commercially, so they have more to lose.


On 11/1/06, William Denton <[log in to unmask]> wrote:
> On 2 November 2006, Alexander Johannesen wrote:
> > I can't get to that site (is it down?), but a few words on what you're
> > trying to do (is it a technical approach, model approach, philosophical
> > approach?), and how you want to do it would be great.
> It's back up now.  Sorry about that.  Some one-time server stuff was going
> on.
> OpenFRBR will be all of those, I think, because it'll be figuring out new
> things as it goes.  There is no full FRBR implementation so there are more
> questions than answers.  I don't know how the database will look, what
> Ajaxy stuff will help users manipulate information, how to let people do
> the four FRBR user tasks, how to handle complicated relationships between
> things, or any of that.  It'll be fun to figure it out.
> On the site, I say what it might look like in a few months: "A person
> grabs a book off the shelf and enters the ISBN into OpenFRBR. OpenFRBR
> checks all available sources and figures out what is known about the book,
> what work it is, what expression it is, what other entities are involved,
> and how they are related. The user will be able to confirm what is right,
> change what is wrong, and add what else is known. The resulting
> arrangement of information will be available in a standard format for
> other systems to use. Everything will be searchable."
> If it gets to that, I'll be happy.  I'm new to Rails, so I'll be figuring
> out a lot as I go.  I've never done anything with a shared source code
> repository, either.
> I do think FRBR is very useful and really needs an example, if not a
> reference, implementation, so that people can say, "Oh, this is why we
> should bother with it.  You know, I could use this for X, and for Y ..."
> Bill
> --
> William Denton : Toronto, Canada : :