Enterprise Open Source

I wrote my longest TechSource post of last summer while sitting in a very warm home office in Palo Alto, sweating in my skivvies. Now in late summer 2007 I find myself in a day-job where the A/C is so efficient I drape a my “office blanky” around my shoulders in the afternoon (I'm essentially a snake, so I'm cold wherever I go in the well-refrigerated South). Things change, don't they?

Things have changed in LibraryLand as well. We've seen further consolidation in the library automation industry. Earlier this year we watched the surprise smash hit, Now, Voyager; then CSA nibbled up ProQuest; meanwhile, Sirsi-Dynix devoured Docutek, and then Vista-Equity devoured Sirsi-Dynix (which has a hyphenated name due to an earlier consolidation some of us antique librarians remember). It's all so very Old Testament!

Meanwhile, the 4.0 release of MetaLib feels faster, has nifty faceting, and identifies the full-text sources. (A full-text limiter would be even better, but labeling full-text is a big leap.) Furthermore, as Roy Tennant wrote over on Library Journal, integrated library system vendors are rolling out “unified finding tools” such as Encore, Primo, and Rooms that hold out hope for a single search interface for library information.

But the truly significant activity in LibraryLand technology hasn't been vendor-driven. It has been the maturation of what I call “enterprise open source”: products such as Evergreen and Koha that are robust, well-implemented library automation packages with strong development communities and equally strong funded-support models. (Next month I'll discuss other open-source up-and-comers, such as Sakai—yes, there is an alternative to the laborious hell-in-a-web-app that is Blackboard!—and small but lithe products such as LibraryFind.)

Olde-Tyme Open Source

For a while, open source had a shaky reputation (when it had a reputation at all) as software written by a guy in a torn teeshirt sitting in his parents' garage. If you used open source software, you were on your own if it didn't work (unless you happened to have a programmer in your library familiar with the code the product was written in). Too often, unless you were talking about massive projects such as Apache, the best you could say about open source software, compared to commercial products, was that it was “free” (as in “free kittens,” as Eric Lease Morgan first put it).

I feel that way about a few open-source products I've had to work with. You can't compare commercial sound-file software such as Soundforge with the limited and irritating Audacity. The best I could say about Audacity, particularly after struggling with it for a couple of hours, is that it is “free” (though my time certainly isn't).

But I don't feel that way about, say, Evergreen—and neither do the members of Georgia's statewide PINES network, who are enjoying far better software, for far less money, than could be bought off the shelf. The Georgia State Library faced upgrading its ILS for $10 million. Instead, their developers built Evergreen for $1.5 million. I don't know about you, but I think it's a good day when I save $8.5 million, particularly when the product I bought for $1.5 million actually does what I want it to do.

And then there's the support. Companies such as Equinox and Liblime exist to support open-source library software (respectively, Evergreen and Koha). These companies provide the configuration, installation, and maintenance activities of any typical library vendor.

Foundation garments

But third-party support is a completely different support model with a very different customer dynamic. Unlike the traditional vendor-support model, open source software support isn't an ancillary post-purchase service where you write the check and pray for good support; good service is the business Equinox and Liblime are in. There's no reason other companies couldn't spring up to compete with them for customers, as has happened for other open source products such as Linux, and they behave accordingly.

Customers are coming on board. In addition to the Georgia PINES network, Equinox —only in existence since July 1 of this year—has signed an agreement with the British Columbia Public Library Services Branch. Liblime has a number of agreements in place and has recently signed several partnerships with membership organizations such as Pennsylvania's Palinet and INCOLSA in Indiana, and has signed up several large groups of libraries, such as the Indiana Shared Library Catalog.

Information wants to be... well-supported

When you ask why open source is important, some developers and open-source enthusiasts might wave incense in front of your nose. One argument is that you can modify the code. But most libraries don't have developers, so this argument isn't often compelling. Another argument is that open source software is “free.” But I've worked in an organization where we poured money into developing “free” open source software (and once we stopped pouring money, the development also stopped), so that argument isn't persuasive.

The primary value of open source software is the open nature of its code. Those of us who have worked with traditional vendor software, with our hands and feet bound by nondisclosure agreements, know the frustration of not being able to share development issues with the broader library community (let alone actually look at the private code driving the software purchased, in most cases, with public dollars). In the open source world, transparency is a strength, not a threat: problems are pondered and solved in public. In the same vein, in the open source library software world, the “secret sauce” isn't the software code; it's the companies that support it. Service is everything; software code is just part of the means to the end.

Another significant value of open source software is that its survival depends on a community's needs, not the whims of a corporation. I don't begrudge anyone the need to earn a living—I too have a mortgage to pay—but too many vendors have left librarians at the altar while they pranced off to follow their bliss (can you say “Taos”?). As we librarians know all too well, there is no greater protection for intellectual property than to place it in the public trust.

Finally, I'm going to contradict myself a little by saying that another advantage of open source software is that you can modify it. True, many of us do not have developers who can do this work. (I often dream of a world where every library had one programmer on staff. Imagine tens of thousands of librarian developers, working full-time—and collaboratively—on library software!)

But even if you wouldn't hire a programmer for your library—permanently or for a few hours—the ability to modify the software changes the balance of power. It says this is software of and by the people; it's a statement about ownership.

In a library automation market of massive consolidation and buy-outs, enterprise open source software is above all disruptive. It opens up opportunities to choose between traditional closed software and open source software—or even to write your own. All that is good for all of us.

Next month: other open source packages get attention. What's your favorite? Write in!