On Mon, Aug 25, 2008 at 18:05, Karen Coyle <lists_at_kcoyle.net> wrote:
> Actually, we have met a few times with the NISO AVIAC
> group, the only library vendor group that I know of. It turns out that an ad
> hoc subgroup was coming to the same conclusions: that we need identifiers
> for our elements and vocabularies, and we need to be able to create
> application profiles.
I'm glad of this development, albeit about 10 years later than the
rest of the world. *giggle* I know I shouldn't giggle, though, but
SemWeb and Topic Mappers and anyone who knows anything about digital
identity control knows that global identificators are a must for a
*first* step into the world of serious collaboration, and have said so
far too often. I think the library world was for too long hanging on
to the LC / OCLC record numbers, and the authority records (which
really ticked me off as it led to the wrong focus alltogether).
> Now, that does not mean the the library vendors, who
> do not move quickly (for a variety of reasons, many beyond their control),
> will jump all over this in the near term. I actually have more hope for OSS
> activities.
Yes, indeed. In fact, this is a golden opportunity for the (F)OSS
movement to lead ahead and possibly save the day. (Evergreen, Koha,
are you listening?!). OSS would have no trouble with integrating with
already existing OSS SemWeb technologies, and I know Talis certainly
work in this field a lot lately. The W3C push of SemWeb, flawed as I
think it is in turns of their own agenda, is an excellent and open set
of standards for doing these things.
> An interesting note on the RDA "product." RDA (the rules themselves) is
> being developed into an online product that will allow libraries to create
> their own cataloging profile -- selecting the sections that they use, adding
> their own instructions, etc. In other words, something very like application
> profiles. If nothing else, I think this will help people get the "AP idea."
Certainly, and it makes us able to make smaller sets for people who
will be flustered and scared senseless (I was going to use another
word here, but I've been told that the gentle people on this list will
ignore me completely if I use anything approved by Carlin [RIP], so
I've toned it down for you ... :). The first time I looked at LCSH I
was floored, just like the first, secod and third time I was wading
through what was available for RDA. Needless to say, I was floored by
the shere amount of knowledge and information one needed to have in
order to make one single good MARC record. (And this - as opposed to
the format itself - is why people hate MARC. hate the culture, folks,
not the format)
>> Briefly looked through them, and some of them look good, but you need
>> better structured and clearer instructions.
>
> Ah, this is a part that needs a better explanation. The registered elements
> are intended to be as neutral as possible -- defined semantically, but
> without input instructions. The input instructions will be part of the
> application profile.
Hmm, I might have lost it there. When I go to the task-groups homepage
(http://dublincore.org/dcmirdataskgroup/) there is very little
information there which I'd considered easy to follow or use. I'll
allow myself a little tangent here ;
Anything which isn't simple, easily explainable and implemented in a
few lines of hacker code will *not* save us from doom. This is not
about the quality of the information, nor the meaning, or the
abilities of the proposed solution. This is all about the fact that
we're a large industry with so much diversity that needs to come
together as one, so that anything more complicated than a few lines on
the idea and how your lovable hacker can try it out (including links
to software that they can try it out with) is going to fall into the
"too hard" category and quickly forgotten. The reason isn't that there
ain't smart people who's doing these things or will be using it (or
try to), but because time is of the essence. We need to play with this
stuff *now*; not soon, or nowish, but right *now*. Make a toolkit, lik
to software, tell them how they can get this stuff up and running
quickly. If you must, tell them about URIs as persistent identifiers
and the single genius idea about using that instead of numbers as a
means of being resolvable and ripe for further exploring. Use the web.
If you must, explain the web and why it gives us the power to do these
things.
This is a bit hard to explain, and when I say "URIs as persistent
identifiers" a lot of folks thinks they know what I'm talking about,
thinking "hey, we're not stupid, we've been using URIs for a long
time." Have you used them as *persistent* identifiers for things? One
thing is to have a URI in an ILS as
http://mylib.org/in/ils/page.cgi?pageid=348348&selection=23 with no
guarantee that the vendor will keep those URIs forever. Another is to
design a set of URIs (such as your repository have an URI for every
"thing", which is fantastic!) as http://id.mylib.org/isbn/29384758695.
Resolve it to get more info. Browse your resources as if it was the
web itself (because, you know, it is :)
Anyway, it's easy enough to explain the power of these URIs than
coming up with prose about how great they are. Examples. Prototypes.
And it needs to be easy, so what is more easy that clicking on a link
in your browser? It needs to be that easy. Maybe a bit of embedded
JavaScript. A touch of namespaces. A pinch of OSS on the side. And
more examples with working code.
Then you might have a chance. If it isn't that easy, if people need to
think hard about this solution, they'll choose what someone else is
doing instead. And that might be ... uh, turd.
> I'm not 100% sure this works, but it comes out of the Dublin Core Singapore
> Framework, and I think it's worth a try. I do think it means that we'll have
> to get more specific in our data elements (e.g. distinguishing title from
> title page v. title from cover) to create compatibility between communities.
> I think it's going to take a while to get where we want to be. I wish it
> could happen more quickly, because I feel the 'moment of opportunity'
> slipping away from us sometimes. *sigh*
This is exactly what I was struggling with when I was working in the
library world; the time was ripe, yet due to lack of funds, resources,
passionate people, people who got it, or people who could bring it
further, things dwindle off into the sphere of prototypes and legacy
systems that will be replaced with something else in a couple of
years.
I think most normal librarians and catalogers wait for the vendors to
create the tools, while the vendors are waiting for directions from
the librarians, and the custom development we actually do is "in the
meantime" stuff. We need to make the stuff you're doing into the prime
thing we do. Why librarians aren't involving themselves more in the
model discussions is a bit strange, given its metaphysical abstract
fluffy nature, the perfect fit for any librarian, but maybe they've
been so used to committees solving problems for them they can't spot
the Wiki and the blogs in the corner.
Dunno. But if you need help with anything, just holler. Saving the
library world is my quest, even if I'm not in it at the moment.
Alex
--
---------------------------------------------------------------------------
Project Wrangler, SOA, Information Alchemist, UX, RESTafarian, Topic Maps
------------------------------------------ http://shelter.nu/blog/ --------
Received on Mon Aug 25 2008 - 13:17:09 EDT