So I've been watching all this with some bemusement and much enjoyment
(really liked Ed's Five Laws of Repository Science).
Because from where I'm sitting, having run IRs for coming up on six
years, a good repository is both simple and complex. Simple, because
my definition is "a repository whose sponsors have clearly articulated
what they want it to accomplish (presumably having consulted
appropriate stakeholder groups), and whose staff and software
successfully accomplish those things."
Complex, because (clearly, from this thread!) this is not the normal
state of things in many if not most IRs. I have YET to understand what
concrete, measurable goals many IRs (particularly in the States; I
think things are a bit clearer in the UK) have been founded to
accomplish. Believe me, I've asked. Most of the time, sponsors look
back in bemusement and say "I dunno; what do YOU think it's for? You
run one, after all; you surely must know!" This abysmal excuse for
buy-in and support in high places is obviously not ideal! Should I try
to articulate something, they nod their heads, say "okay, y'all repo
people just go on and do that, then" and walk away without engaging on
what would actually be needed to meet whatever goal I have outlined.
My regular depositors (now that I actually have some, which I do) are
all over the map:
* a vanishingly small percentage -- approaching zero, really -- care
about open access (sorry, but it's true)
* a very few want permanent, citable URLs for something (and aren't
stuck on DOIs)
* quite a few want decent care taken of born-digital or digitized
materials they don't have means or brainspace to look after (from
master's theses to local newsletters to institutional records)
* quite a few want their stuff to be Googleable (and the IR I run has
pretty good Googlejuice compared to a random uni website)
And many would-be depositors want things that I can't give them: a
home for massive files (digital video is my bĂȘte noire), usage
statistics, a bullpen for in-progress work, easy-deposit connections
to their regular working environments, version control, "reserved"
permalinks (so that they can put the permalink on the poster or in the
paper before upload), display environments tailored to different
content types (such as image and poster galleries, or pageturners, or
EAD viewers, or something like Sudamih's Database-as-a-Service). So
all of the blue-skying in this thread is fun and all, but also quite
frustrating to read, because *the software is not there, people; it's
just not*. Worse, many of us IR managers, self emphatically included,
*have completely run out of credibility* to ask for development and
collection-workflow resources, because for some unimaginable reason we
haven't thus far managed to meet goals that were never articulated for
us in the first place!
As for sponsors, especially at the outset of IR planning they seem to
want the IR to Magically Do Something about the serials crisis or
records management or effort reporting or gray-literature collection
or web archiving or the Death of Internet Culture or the Data Deluge
or whatever. Explaining until blue in the face that IRs are not magic
pixie dust, collection-building and culture change are *hard* and
*slow*, and repository staff (especially when that's a single person)
are not infinitely scalable never seems to register with such folk...
despite over half a decade of painful IR experience which has been
plentifully and (by and large) honestly published about.
The sense I get from many IR sponsors is that they're not sure whether
they want the IR to be a COLLECTION, a SERVICE, or an ADVOCACY TOOL
(this last on various levels; some want IRs to advocate for the
institution, some for open access). There's nothing wrong with any of
those goals. Each of them just has to be policied, staffed, and
organized rather differently. Lack of clarity on this point means that
IRs and their managers try to serve as two of these, or perhaps all
three, and they (we!) don't do particularly well at any of them. Which
is, I can say with authority, a horrible hole to be in,
professionally. I like to do things well. I don't know that I ever
have, in this space -- in fact, I long since concluded I and the
repositories I've run are dismal failures.
So, perhaps another definition for a good repository: "one that
doesn't make its manager tear out hanks of hair regularly in pure
unadulterated frustration." If there's one of those out there, I don't
think I've found it yet... and I know a fair few repo managers, I do.
Dorothea
|