Dorothea Salo

Introit

Thank you for that extremely gracious and wholly undeserved introduction, and thanks also to Robin Rice for the opportunity to be here in lovely Edinburgh and speak to this brilliant gathering, and to Les Carr of Southampton for very kindly taking me under his wing yesterday.

I promised CONtroversy—hm, I suppose here that’s conTROversy, but either way, I’m supposed to bring it on. Well, I’ll see what I can do. It’s a little intimidating, really, because as I said, Les took me around to the Fringe yesterday, and I saw what I’m up against—this city is crawling with professional entertainers, people who do this for a living, and if that’s not enough it’s also crawling with pubs, so on the one hand people here have had their standards elevated quite high, and on the other hand, well, some of them have been hitting the pubs, and it’s a tough crowd, nobody here is interested in me! It’s a little as though I’ve been set up to take a tremendous pratfall—I cannot win here!

Any resemblance between what I have just said and the situation of the institutional repository is—entirely premeditated and intentional.

In the spirit of the Fringe, let me say that you are not required to just take my chaff lying down—well, sitting down. I have a bad reputation for being just a terrible, terrible heckler when I’m sitting at talks, so now’s your chance—heckle back.

L’IR, c’est mort

The old IR is dead. C’est mort. It is an ex-repository; it has joined the choir celestial… yes, all right, no more Monty Python. I will do an autopsy on it first of all, in hopes that that will be controversy enough, but what I really want to do is talk about the reinvention, renovation—the renaissance of the institutional repository.

So if the institutional repository is dead, who killed it?

We did. We all did, librarians and developers alike. The IR died a-borning, folks, because we started out with a single underlying idea for it—anybody know where I’m going with this? All together now: If you build it, they will come. This was a hopeful, audacious, beautiful idea. I love this idea. I have always loved this idea. But a lot of other assumptions grew out of that first one, and they turned out to be unfortunate:

• We did all the planning. So everything will be just fine.
• Resources? Staff? It’ll mostly run itself.
• All we’ll take is the peer-reviewed literature.
• Everybody wants open access!

And that ruined our planning processes. Because we thought they would come, when we did all that planning and wrote about the planning process, we didn’t consider what to do if they didn’t! We didn’t staff these things up—we didn’t think we needed to! We planned for the peer-reviewed research literature because that’s what we wanted. Never mind what faculty wanted or needed, or what they’d be willing to give! And we wrote some things—and this hits at some people in this room, and I’m sorry for that, except I’m not, really—we wrote some things that were utterly fatuous. “Everybody wants open access.” Yes, well, I want a pony. Where’s my pony? Look, sure, they might say they want open access—though I don’t know about you, but I’ve only ever heard a few of ’em say that—but the end of that sentence is “as long as I don’t actually have to change anything that I already do.” Worse than fatuous, this nonsense was counterproductive; it got in the way of what became a struggle for content and staff and resources and basic, basic respect for IRs.

Developers, it’s your turn on the hotseat! These are two of my personal favorites; they have been spoken of the same repository platform:

• Customization? It works out of the box!
• Oh, come on, they can just hack it to do that!

We didn’t need customized solutions, because we were assuming that IRs were about open access, and “everybody wants open access,” right? We built software that worked for PDFs and not a damn thing else, and we didn’t build in APIs or two-way protocols or plugin architectures, because we knew what we needed the software to do and it did it! But hey, if that doesn’t work, it’s open source, you can just hack it!

Well, no. There is nothing quite like talking to a developer at a conference, telling him that you are an extremely amateur self-taught programmer newbie, but you’d like to contribute something to the IR platform you use, and having him tell you airily that you can just rewrite the whole browse system from scratch! This actually happened to me, cross my heart and hope to die. Some of these people have a really weird notion of ‘extremely amateur.’

• Dublin Core is plenty good enough.
• Document versioning! But all we want is the final version, right?

And some general silliness circulates about Dublin Core, as well as the notion that we shouldn’t give people things like document versioning that they—hello!—are asking for, because they don’t fit into our narrow little conception of the IR universe. Our software and services don’t have to satisfy anybody! Because everybody wants open access!

A lot of this boils down to the assumption that utterly destroyed IRs: “Sure, they’ll type keystrokes!” I don’t need to add to that, except to say, again, that it’s a nice idea. Nothing wrong with it as an idea.

These ideas informed what we built. But they were wrong. So we built it, software and workflows and policies and procedures, and they did not come. To us here at this gathering, this is not news. But it still is news to our library administrators, to some of the top thinkers in open access, even to the very programmers who make our software! The thing is, none of them gets blamed when uptake is low. That’s on people like me: the service that I am the public face of fails. Sure, it fails because it was built on wrong assumptions from the start, but realistically, who will they blame? They don’t know you people. They know me.

And that has consequences.

I started in my current position a year ago March. Two months later, I learned that the statewide library system that funds my position had cut the repository’s budget for the 2008–2009 biennium by one-third. (“Essential infrastructure in the digital age,” says Mr. Lynch.) Now, I don’t think these two events are connected in any way except temporally, but who knows? Maybe they knew I was controversy on the hoof.

Anyway, in October 2007, I was asked to chair a working group to figure out the future of the institutional repository I run. Oh, and incidentally, they also told me to figure out how they should fund me, which of course implies explaining to them why they should fund me at all! Suspend that sword of Damocles! Just a little over to the left, or you’ll miss the artery.

What happened then? At this point I am going to leave you in suspense.

With this weird systems-analyst brain I seem to have, I asked myself what was going on and what I could do about it. So I want to talk for a moment about faculty, because our ideology has simply left the ones I’ve worked with in the dust. With our eyes on the brave new world, we ignored the world that faculty are still enmeshed in. Academia has astounding inertia. It changes at a glacial pace. It is also a viciously competitive environment with very narrow boundaries.

IRs are new. They do not fit either the old boundaries or the current world. Faculty have questions, and lot of the questions are legitimate, though some are pure paranoia:

• Will my publisher be upset?
• Will someone plagiarize me?
• Will someone violate my copyright or steal my idea?
• Will I be sued for violating copyright?
• Is my institution trying to be Big Brother?
• What is the authoritative version of record?
• Will this count toward tenure and promotion?
• How do I cite this?

It all amounts to a great wild cry of “I don’t understand!” And because academia is conservative and timid as a culture, that means they’ll ignore us with great avidity until they have answers to their questions. But we have done a really good job of ignoring those questions too. And we want these people to hand off their professional lives to us? Really? So the next time you feel the urge to say “Everybody wants open access!” please don’t. It’s not that simple. Faculty fear and apathy are a real downer.

Another thing that let the institutional repository down was its software, which assumed that it only had to exist and halfway work to be popular. It didn’t have to be flexible, accommodating, attractive, or usable! It didn’t even have to be useful! So the top two words I find myself uttering about the software platform underlying the repository I run are “I’m sorry.”

• I’m sorry, I don’t have download statistics per item or per author.
• I’m sorry, I can’t manage versions of your article for you.
• I’m sorry, there’s nothing I can do with your dynamic website or CGI app.
• I’m sorry, the software won’t let you edit the metadata after deposit.
• I’m sorry, I can’t help you digitize your analog materials.
• I’m sorry, I can’t display your scanned book pages as a page-turning app.
• I’m sorry, I don’t have a gallery view for your images.
• I’m sorry, I can’t stream your videos.

I’m so tired of saying I’m sorry. I’m tired of being everybody’s whipping-post for inadequate software and inadequate ideology. Yes, the IR is dead, and I helped kill it, and I am not sorry.

Death throes: the shape of opportunity?

“Only in silence the word,
only in dark the light,
only in dying, life,
bright the hawk’s flight
on the empty sky.”

—Ursula K. LeGuin, A Wizard of Earthsea (emphasis mine)

So we can mourn the IR, and heaven knows I’ve done plenty of that, or we can see the shape of opportunity in the ashes. So let’s look at those ashes a little more closely. Just now this bad, bad, bad software—and I’m sorry, it’s all bad, just some of it’s less bad than others—has got us in a hammerlock.

We’re stuck with workflows that don’t work. I’ve given up trying to explain DSpace’s review steps to people, because they are utterly senseless. They bear zero resemblance to what people actually do when they make digital stuff.

We get defaults and designs that make no sense: neither MIT nor Southampton should be telling the world what an IR should look and behave like—yet they have.

We get protocols that don’t do enough. OAI-PMH is, I’m sorry, a disaster. You can’t search with OAI-PMH; it doesn’t do queries. You can’t retrieve anything other than an item or a full set, and it’s the data provider rather than the harvester who’s defining the set, which for a harvester is a recipe for getting what you don’t want. OAI-PMH aggregators know that most IR metadata reeks, and they go through heroic contortions to clean it up, but they can’t do anything about the badness, because you can’t pass metadata corrections back over OAI-PMH; you can’t even tell the data provider they have a metadata problem!

Now, we librarians, and we repository-manager types, are nice people. (Really. I’m the exception that proves the rule.) We live to serve! And we get frustrated beyond belief when our software gets in the way of our ability to serve. You can’t tell me the software hasn’t burned some good people out, all by itself.

And now, developers, I hope you are wincing. Because I don’t know of any reasonably well-funded IR that isn’t hacking the living daylights out of its crappy software. And a lot of the time they’re hacking in parallel to do the same damn things on different installations, because the Lord High Inner Circle Developers won’t address problems! If I see one more LDAP stack or embargo hack for DSpace, I will just scream! It is stupid that we are wasting all this effort that we could be putting into making better software and better services. Because that’s the opportunity, right? Better software and better services. That’s what we want in the new IR.

There’s also our relationship with the larger information universe to consider. The Web is whizzing past us, and we are standing still. Mashups, for example: what is my library supposed to think of me, when I can’t easily roundtrip metadata between the IR and the catalogue? And I can’t. I can’t easily accept catalogue metadata either, which is uncommonly annoying, especially around electronic theses and dissertations (ETDs).

I don’t want to embarrass anybody here, so no examples, but most IRs are dead ugly, I don’t care what platform they’re using. I went to a user-experience conference once where research was presented indicating that people form lasting impressions of a website in well under one second based solely on the design. If that’s true, and I believe it, no wonder the IR is pushing up the daisies.

We’re missing opportunities here. We’re missing opportunities to make ourselves useful and appealing. We can’t afford that! Arguably we never could, but we’re in the situation we’re in, so here’s a way we can make it better.

And then there’s us. There’s quite a lot of “us” these days, and we have a lot of different titles. We report to different people, we have different priorities, and we don’t all speak the same language:

• Data curator
• Researcher
• ETD coordinator
• Repository librarian
• Research programmer
• Systems administrator
• Systems analyst
• Scholarly communication coordinator
• Metadata librarian
• Liaison librarian
• Grant administrator

We have been doing a lousy job of communicating and working together across these boundaries, and of being a coherent collective voice. This has just killed us. I myself feel out on a limb all the time; it doesn’t seem like there’s anyone in the larger profession acting as my backstop. I don’t have a magic cure for this, I’ll tell you right now, but it ought to worry us. Who are we losing? We need to do better, and there’s opportunity for that—as this very event makes clear.

Let’s take a step back—then two steps forward!

Now, it’s hard to give up our cherished beliefs and goals, code we’ve worked on for years, our sense of the world. A lot of the arguments I see in the community lately feel like warmed-over leftovers, honestly. We really need to move on.

Beyond green open access

Repositories were built because of and for the sake of green open access. Unfortunately, we haven’t made inroads there, for all the reasons I’ve described. Self-archiving will not save us, quod erat demonstrandum. We have to take a step back from it and two steps toward other things. Here’s what we’ve started to learn.

Whether it’s data curation, gray literature, ETDs, professional-identity management—we’re learning to own and be proud of projects that go beyond the peer-reviewed literature. In a larger sense, that means that we’re looking at a great deal more of the research process than the final publications, and we’re discovering that there’s real gold there for us, gold that’s a lot easier to pan for because we’ve never harvested it before, so faculty don’t have the same preconceptions and habits that we have to try to break down.

This leads to a reconceptualization of green open access—and it’s worth pointing out that we’re not walking away from open access just because we’re stepping beyond it! We’re just being more strategic in how we approach it. Faculty have shown us they don’t care about open access per se. So we’re finding things to give them that they do care about, and that achieve open access as a byproduct.

Beyond silos

A problem I run into a lot where I am is the whole silo thing. We have the IR, and we have our digital-library platform, and honestly they contain a lot of the same kinds of stuff. I have images. They have images. I have conference proceedings. They have conference proceedings. I have faculty research. They have faculty research. But when a project comes up, whether it goes to the IR or to the digital-library platform is a decision made on really flimsy grounds. It ought to be made based on what platform will do the best job presenting and preserving the content. But instead, we have a mishmash of politics and procedures and rank snobbery and the like—I’m sure all of you are familiar with this—and it’s really silly. Ideally, it ought to be about the stuff, not the platform. But it still isn’t, and our stuff is still locked in silos.

We have to learn to be profligate with our stuff! We’ve focused so hard on acquiring stuff that we’ve ignored all the other roles we could play, and we’ve focused so hard on where the stuff lives that we’ve ignored all the other things we could do with it besides just warehouse it. It’s long past time to step back from that.

We especially need to take two long steps toward the role of helper, because if we’re going to overcome that great huge “I don’t understand!” bubble, we’re going to have to do better than “Here’s an IR. Fill it.” And this kind of help, whether it’s copyright help or file-format help or data-management help, has to happen on an institutional level—national and disciplinary repositories don’t have the reach, the staff, or the patience. This is what’s beyond the silo: the service.

Beyond eternal demos

When I go to conferences around institutional repositories, and I’ve been to quite a few by now, I see a lot of cool demos. Look at this GIS mashup! Look at these researcher pages! Look at this automatic conversion tool! But that’s all I can do. Look. Wistfully. Half the code written won’t work on my repository platform. Half the remaining code is some kind of grotty hack that my systems administrator will quite rightly turn up his nose at. Half of what’s left won’t be supported in the next version of my platform. Half of what’s left is—well, not much!

Ladies and gentlemen, this is broken. While a very few of us leap ahead to do great things, most of us are spinning our wheels. We have got to step back from this inner-circle, million-hack development model we’re stuck in, because it is not helping us. We need to step forward to component-based software that enables faster innovation that actually spreads, becomes mainstream, instead of forming the basis of a really cool demo and then dying.

This is about more than code. The other thing we’ve done a really bad job of mainstreaming is our knowledge about what works and what doesn’t. Just last May I presented at a conference as a favor to a friend, and while she was driving me to the bus station to go home, she said that her tiny little institution was gung-ho about creating an institutional repository. Why? To capture their faculty’s peer-reviewed research literature. And how was this going to happen? Why, self-archiving, of course! And how much staff would it need? Staff, who needs staff? You see my point here, I hope. How many more people are we going to let stumble all over themselves and fail?

Because when they fail, we fail. At ASIST Annual last year, I met a librarian friend who talked about their fledgling ETD program. The library management there was wholeheartedly behind ETDs, but according to her, her boss said “This won’t be one of those institutional repositories, right? Because they all fail.” Coming back to me again—how am I supposed to advocate for a service that has these kinds of headwinds blowing against it? How?

And with that, let’s return to the story of the repository I run. I want to finish that story now; I think we’re ready.

At a meeting in March—and Les Carr can back me up on this, because he was there—the super-amazing working group I was chairing looked at the problems that the repository was facing, and everything I’ve laundry-listed here came out, and they said, “Okay, why do we have this thing? We have a fantastic digitization and digital library program with a phenomenal faculty response. We’re looking at moving its storage platform to more flexible software (namely, Fedora). The IR silo only confuses people. Let’s kill it and add the content in with the digital library!” One step back. Big step back. You would think, anyway.

But you don’t have to tell me twice that I can stab DSpace in its black, black heart. So we turned in a draft report this month, in which we suggest that the repository should rise from the ashes as something that might actually help people! The exact shape of that—well, we don’t know yet. We suggested a lot of things, from ETDs to copyright training to adding born-digital materials management to our regular digitization program to digitization training. What sticks—well, that happens way over my head. But I’m confident something will.

Variations on a theme in H

So on that hopeful note, I want to offer some ideas on what I think the reborn repository will look like. And for some bizarre reason I can’t explain, my thoughts fell into words starting with H.

Harvesting

The content is out there, Scully… It’s in Slideshare and CiteULike and WordPress.com. It’s on faculty’s individual websites. It’s in course management systems and disciplinary repositories. It’s there! We know it’s there! And we’ve got to stop pretending they’re going to stop putting it there and start putting it in IRs instead, because I’m sorry, they’re not. So why aren’t we harvesting it? Why aren’t we intercepting it on its way there? Why aren’t we picking it up once it appears there?

Well, there are some challenges here. The first challenge is automating this, because it cannot be a manual process; we don’t have the human resources to even think about that. So we need APIs. And we need to be really really careful about building our own APIs, because as the history of Google and OAI-PMH shows, we’re not good at mainstreaming library APIs.

The second challenge is rights, and it can be even more annoying than you’d think. Take a hosted blog on something like Blogspot or WordPress.com. The blogger owns the content, no problem there. Who owns the blog design? The blog site. Annoying.

The third question is how we build the networks not of technology, but of cooperation among repositories, commercial services like Slideshare, and disciplinary repositories. This is starting to happen; Microsoft and arXiv just announced that they’re working out how to make arXiv support SWORD, which is fantastic. A lot of these folks will be tough nuts to crack, though. It’s going to take sustained pressure to make something like the Social Science Research Network see reason; those folks see the content they attract as their chief business asset, and they guard it like the crown jewels. We will have to convince them that working with us benefits them. It would be nice if we could do that sincerely, without some of the strongarm tactics and bloviation that have marked our relationships with publishers and even faculty.

The last challenge is purely technical, but we can’t ignore it because of the incredible social pressures operating on faculty. We need that statistics exchange and compilation. It is not fluff, not insecure ego-boosting; it is not optional.

Healing

It’s also time to heal some wounds we left in our mistaken haste. Our metadata is a disaster, folks. Now, I could go on and on about how bad metadata affects end-users, and I’m sure some folks listening to me right now are rolling their eyes because Google rules all and metadata just doesn’t matter, but honestly, that’s not what I’m on about. Bad metadata gives us a credibility problem within our own communities. It looks really bad to libraries. It looks bad to faculty, even faculty who think “authority control” is the administration reining in their local department chair.

What looks bad to both technologists and librarians, and deservedly so, is that our technologies didn’t make any provision for fixing bad metadata! We’re swapping it around all over the place, and with OAI-ORE we’ll be swapping whole items as well, but the thing is, nobody on down the chain can say “Hey, oops, you’ve got this author name here, and it’s just wrong, and here’s how to fix it, would you please?” This is a solvable problem! Let’s solve it.

We also need to heal some of the bizarre dislocations that make us unattractive to faculty depositors. If a faculty member brings me some CD-ROMs and a pile of paper, I am stuck! I can’t do anything with the paper. But this is ridiculous—it’s all part of the same problem, which is capturing that faculty member’s work. There’s an interesting meme in the archival community in the States called “More Product Less Process.”[1. Greene, Mark A. and Dennis Meissner. “More Product, Less Process.” American Archivist 68:208–263 (2005).] I think some of their experiments with digitizing stuff fast and on the cheap could be adapted to repository work. Let’s deal with the paper as efficiently as we deal with the data.

For our next trick, we can ease the worries faculty have about putting their stuff in so many different places! Unified statistics is part of that, but so is an exchange mechanism so that faculty who job-hop can take their repository content with them. Believe it or not, I get this question pretty often! The subtext is that faculty need to feel that they have control over their stuff. Whatever we can do to reassure them that they have that control is a good thing.

As I hinted previously, we have a silo problem. The more we can do to align ourselves with similar efforts, the more support we’ll have and the more useful we can be. That means aligning our interfaces, aligning our APIs, aligning the message we present to faculty so that we’re genuinely in harmony with our environment. And the same for the wider Web; we can do much better working with what’s there than stubbornly insisting that we’re better!

Finally, I want to give everyone permission to say that institutional repositories haven’t been a sterling success so far. It’s okay! Really! Don’t be scared of that, don’t hide from it, don’t try to cover it up. Let’s just say it. Le IR, c’est mort. See? Not so hard.

But that does mean we have some fences to mend, and some credibility to think about restoring. Honesty can only help with that. Additionally, I am here to tell you that we’ve got a serious morale problem in the repository-manager ranks—and yes, it’s sort of hidden, because the other thing about repository managers is that, as a class and with exceptions, we’re afraid to speak up about things. (With exceptions! I said with exceptions!) So it could be ignored, but we don’t want to ignore it, because we need each other. Let’s heal it. We can.

Helping

So we built it, and they didn’t come. And we were bruised by that, because there’s a lot of wonderful idealism running around in repository circles. But we can refocus that idealism, I think. Instead of saying “open access is useful to the world,” we can say “we, as IR managers, are useful to our administrations and to our faculty.” The idealism is the same… but the way it works itself out in the world is very, very different.

We don’t entirely know how we’re going to be useful yet, just as I don’t know what the university system I report to will decide on as far as services I should be offering. But we have a lot of options, and I expect to see many of them discussed here in the next two days. Data management and curation is coming onto the horizon, and a theme I have heard over and over again is that early intervention in the research process is necessary to do this right. Libraries in the States, at least, have been slow to warm to this, but that may change. It’s definitely something to think about.

One add-on service I see coming has to do with all the current and future mandates coming down the pike, from funders and institutions and governments. Now, contrary to what you may have thought, researchers do not like deposit mandates. Oh, sure, they’ll comply if their funding is on the line. But you would not believe the howl that went up from our campus over the NIH mandate. Harvard, yes, I know… but I would watch Harvard closely if I were you. I don’t know that the researchers there know what they’ve gotten themselves into.

So imagine a researcher, maybe three to seven years from now, who has two major and one minor funder, all of whom have deposit mandates. Imagine she works at an institution that also has an institutional mandate. Let me tell you, this researcher is not a happy camper. Can we help? Yes we can, and we should. We can and should offer one-stop shopping. Local deposit should be all that researcher has to accomplish. We can check the grants database for who the funders are, and we can see that the file gets to the external repositories it needs to go into, and we can tell the researcher we’ve done it. Judging from our experience with NIH, researchers will be thrilled to receive the help. On our campus, we offered to do PubMed Central deposits for anyone who asked. We’ve done maybe a few dozen so far—but we earned unbelievable goodwill just by offering. So there’s an opportunity here. Let’s get in front of it now!

Funny thing, though. NIH deposit wasn’t a service anybody ever envisioned being part of the IR. So maybe—and here’s the real radical suggestion—we don’t need to know right now what services we’re offering. After all, historically we’ve drawn much too tight a line around them. We have talented people running these repositories; a lot of them are in this room. Maybe what we do is free them from the ideology, free them from the software, and turn them loose. We’ll get open access. We’ll get preservation. What else will we get? Good question! Let’s find out.

Hope

The last H is Hope, because good things are happening and we should celebrate them! Funders are not backing down, which gives us opportunity to help faculty who may not ever touch our institutional repositories. That’s not a downer; it’s genuinely hopeful, because it gives our people rather than our silos relevance. It’s sad that I even have to say that, but honestly, that’s been a weakness in the IR movement all along: celebrating software instead of people.

We also have examples, finally! Southampton, Imperial College, Harvard, Stanford. Faculty are monkey-see-monkey-do emulative. Trailblazers matter!

Librarians are starting to push past their laserlike fixation on the end-products of publication. This is all to the good! And a lot of what I see in the software- and standards-development communities lately is fantastic, absolutely moving in the right direction.

And finally, just the fact that I of all people am standing up here before you indicates that we’re willing to ask ourselves hard questions, and rethink the assumptions we initially held. We haven’t always been willing to do that. That’s hopeful.

Starting to fit in—finally!

Now, to finish up, I want to offer a (necessarily incomplete) sampling of a few people in this space I look to for wisdom. Keep in mind here that I’m an ugly American, and so what’s happening where I am lags two or three years behind where Great Britain and Australia are. Likewise, wisdom being as important as it is, we all think we have it, but truth be told, one person’s wisdom is the next person’s idiocy. So some of the people I’m going to show you actually argue strenuously with each other about things! Which is wonderful and necessary and completely healthy; nobody has all the answers. But I want to posit that a single theme runs underneath all these people and all their arguments, and that theme is fitting in.

Imperial College London is a fantastic case in point. Faced with the same adoption problems everybody else has got, they said, “let’s quit jawboning people and see how we can fit ourselves into the landscape!” So they trawled for publication details from library databases instead of asking faculty to make lists. They built online CVs for their faculty, because doing so encouraged faculty to give them a hand dealing with publications. And when they built an ingest workflow? Faculty saw one page. One. For getting all the workflow nonsense out of faculty’s way, Fereshteh Afshari and Richard Jones and the rest of the Spir@l team, I salute you. Well done!

I don’t know why anyone uses DSpace, honestly I don’t; ePrints has got it whipped in so many ways that matter. The key one to me is usability, a piece of software that one doesn’t have to apologize for. In an email to me, ePrints developer Christopher Gutteridge talked about their refreshingly common-sense approach:

“We… apply a Spock approach—the needs of the many outweigh the needs of the few. So end-users are most important, followed by depositors, followed by editors, followed by librarians, etc.”

“EPrints 3.1 is a little different… to make administering it easier. We’ve introduced web-based configuration editing, many more configuration files… reload if changed (saving restarts), and… an issue discovery system (duplicate titles etc).”

And then there’s the inimitable Les Carr, with his suggested ePrints slogan: “EPrints: sucks less than Hotmail!” Now, yes, that’s just funny, but there’s also a self-critique there that we could do with more of in this business. So for humor, for clarity of vision, and for real commitment to the people who have to use and maintain institutional repositories, Chris and Les, I salute you.

Chris Rusbridge is looking at how all the services repositories could offer fit together to create a real value proposition for depositors. Also, I give a shout-out to Chris for being even blunter than I am about repository stakeholders. I won’t quote you, Chris, but I do recommend looking at the notes from the JISC Innovation Forum, and from one incredibly outspoken person to another—Chris, I salute you!

Now, Andy Powell and I have clashed a couple of times. I’m not including him to curry favour; I’m including him because he makes me think good and hard about why I don’t agree with him, when I don’t—and often I do. Andy’s concern is where repositories fit, and into which organizations. He’s not sure institutional repositories are a good fit with researcher practice, and he’s got sharp observation to back up his contentions. He’s also wondering where the content in repositories fits into the larger content universe, and he has many, many great things to say about this. Andy, for vision and deep thought, I salute you.

And Andy leads beautifully to Paul Walk, who wants repositories to fit better into the Web. He interrogates Andy thusly: “I wonder if the user-centric/institutional/global debate around repositories is just symptomatic of a tension about to become apparent all over the (institutional) Web?” Which gets to a lot of what I said, much more longwindedly, about the huge variety of places valuable information lives, and how we can ensure that it stays available, stays usable, stays discoverable. For common sense and excellent systems analysis, Paul, I salute you.

And this presentation just wouldn’t be complete without a nod to the people who find us money and who are our voices in the larger world. SPARC Europe and DRIVER just announced that they’ll be putting on a big marketing and outreach push for IRs across the continent, and while my views on the efficacy of marketing and outreach are well-known, even I have to admit that at the right level—namely, ‘as high as possible’—it does good. It also helps repository managers feel less isolated, certainly a great good thing.

As for JISC, they’ve backed more winners than I can fit onto a slide. They are absolutely the best and sharpest funders in this area out there. And they’re backing another winner in Repository Fringe, so for that, let’s all salute them!

We can do this. We’ve stumbled, we’re getting back on our feet, and we’re going to charge. So in these next two days—have sense, have fun, and have at it! Allons-nous! Thank you.