Marshall Breeding’s Open Source LTR
January 6th, 2009ALA Techsource, a publishing arm of the American Library Association, just published Open Source Integrated Library Systems, a Library Technologies Report by noted librarian-technologist Marshall Breeding.
Overall this 32-page report is of value to the open source community, but get it while it’s hot from the oven. Most of the report focuses on present-day knob-and-dial functionality, which means that “the shelf-life of this LTR [is] likely to be measured in months,” as Roy Tennant noted in a recent issue of Current Cites. Also, Breeding’s report is limited to integrated library systems, which means it does not address the myriad open source library projects such as Jangle, SOPAC and Vufind (which collectively would make a good Techsource report on their own merit).
However, Open Source Integrated Library Systems does provide a snapshot of the state of the major open source integrated library systems in production (Evergreen, Koha, NewGenLib, and OPALS), while introducing and explaining some of the core concepts of OSS development.
Parts of Open Source Integrated Library Systems are more fruitful than others. Breeding is on shaky ground when he devotes two large paragraphs to the virtues of MySQL and then adds a curt coda about PostgreSQL, the well-respected, very powerful database environment used by Evergreen. But Breeding acknowledges the strengths of OpenSRF, a communications layer in Evergreen, and he does a commendable job of outlining the differences in licensing models between (and within) proprietary and open source software. This latter discussion alone will be helpful to many in our field, and extends the usefulness of this report.
In other sections, Breeding lingers to think on the page, and this is often interesting.
I confess when I heard this report had been published — and it was unheard-of within the Evergreen community until shortly before its publication, when its review in Current Cites come over the transom — I had been bracing myself for Breeding’s oft-repeated suggestion that the total cost of ownership (TCO) for open source could meet or exceed proprietary products.
So it was a pleasant surprise to see Breeding more nuanced on TCO than he has been in the past (we are frequent flyers on similar speaking circuits). Perhaps he has mellowed on this point due to his involvement in the “academic” open source ILS exploratory project, OLE, for which Vanderbilt, Breeding’s home library, is a member, and for which Breeding participates directly as a member of the Connecting with Other Projects working group.
The OLE project has only recently risen from the fertile soil of LibraryLand. Like Extensible Catalog, a project out of the University of Rochester, OLE is in a ramp-up period of design and development. (Extensible Catalog plans to have a working product by mid-2009.)
While Evergreen stakeholders keep a watchful eye on the distribution of technology grant funding, it won’t fatally damage existing OSS projects to have yet another project on the table, and the many discussions among the projects — such as OAI-PMH development for Extensible Catalog, and collaborative discussions with OLE — can only help Evergreen’s future development.
Returning to OSS and TCO, the most pragmatic and most common argument for open source in libraries is the predicted lower total cost of ownership, and TCO tends to dominate the discussion in Breeding’s report, as well.
TCO can be a problematic argument. It teeters dangerously close to the rationale that OSS is “poor people’s software.” (Or as one speechifying vendor put it last year, OSS is only good for third-world countries — as if third-world countries were not deserving of good software.)
There are many other, equally pragmatic reasons to move toward open source that might have been valuable for this report to discuss at length, had space and/or printing costs permitted. (Remember when Butterfingers were a nickel, and LTRS were 60+ pages long?) These reasons include the flexibility of easily-customizable software, the native interoperability of software whose code and indeed, development can be viewed by other software partners in the OSS and proprietary field, the security that comes with transparency — it is hard to snow customers that “Taos is on the way” when the code is out there for all to see — as well as more easily achieved customizations and rapid application development.
Given proper leadership and encouragement, open source projects can engage far more developers than any one company could ever hope to. That doesn’t preclude the value of a “seed company” such as Equinox, Media flex, Liblime, or in the broader world, companies such as Red Hat. It is just a mathematical truth that there is major FAIL written all over the model where perennially cash-poor libraries pay licensing fees to closed-code companies, who then try to scrape R&D out of the same small pot providing support, sales, salaries, and other corporate essentials. It’s just not a winnable war. We had it right the first time, when we were building our own code back in the 1970s and 1980s; we just didn’t have the file-sharing technologies to make it possible to collaborate. Some folks have said if the Internet had existed back then, libraries would have invented open source.
Compared to closed code, OSS is a far more flexible economic model. In OSS, development can be produced for money (also called a “bounty”), as a quid pro quo from interested institutions, or (less often, but it happens) individuals in the field. Every good line of code counts toward the effort–sometimes immeasurably. In this sense, OSS in libraries resembles the micro-economics that benefit (ahem) third-world countries: a margin of effort in librarianship counts more than out there in posher corporate environs.
Another, less tangible benefit of OSS lightly touched on in Breeding’s report is that it puts libraries back in the drivers’ seat. There are significant long-term strategic advantages (fiscal, professional, and pedagogical) for libraries to return to the place we were in for over a century, where we designed and built our service-delivery tools.
Tool design forces humans to think and to contemplate, and in our evolutionary history, helped lead us to larger craniums, civilization, and the ability to accessorize. In librarianship, tool design, from the standard card size to Serial Keyword Index of the 1970s to the catalogs we designed through the 1970s and 1980s, compelled us to ponder and think about what we were doing.
This collective intellectual effort — the stuff, Andrew Abbott argues in The System of Professions, that makes us a profession in the first place — was temporarily lost in the period where without benefit of easy file-sharing or a body of thought related to the value of collaborative development, we handed our code over to companies who in many cases had every intention of partnering with us, but were stymied both by the intellectual-property models that drove software revenue back in the day, and by the growing syndrome of what Lori Ayre calls “ten years of learned helplessness” (it’s closer to three decades) that threatened to turn us into the shoe clerks of the information age.
(Shameless horn-tooting: this was roughly the topic of my five talks for VALA and CAVAL in November — how our creative endeavors and tool-building gave us a century of empowerment, and how the renaissance of open source development puts us back in the driver’s seat.)
In any event, perhaps it is out of scope for a short report to think so broadly, but these more abstract benefits of open source are nontrivial, and are quite obvious to those of us who have seen it both ways.
There are a few remaining quibbles. Breeding’s attempts to distinguish the acquisitions requirements of “small public libraries” overlook that many of these libraries are just as interested in EDI as the largest ARLs, and have complex acquisitions workflows in place to support their equally complex needs, often as nodes in a larger organism that is a shared consortial catalog.
In a couple of places the document contradicts itself. Breeding states that “large municipal libraries have not been moving toward open source products,” then notes quite accurately that King County Library System “has publicly announced its interest in moving toward an open source ILS.”
Then again, the largest municipal libraries, though they may be nimble in many respects, really cannot move on a dime, at least when it comes to the massive, complex legacy migration issues. In any event, as Breeding notes elsewhere, right now large municipal libraries aren’t “moving” anywhere — most right now are limping along with legacy systems while they try to keep boots on the ground and books on the shelves.
Once in a while, I blinked, as I did when Breeding commented in a discussion about the growth of open source, “The classic polemic casts Microsoft as a monopolistic domineering company against the open source companies that free the world from its stranglehold.”
That statement was awkwardly squeezed into an otherwise-reasonable discussion; perhaps it was there for the zing factor. Regardless, if you look at the website for any library open source project from Archon to Zebra (and not overlooking our favorite vowel along the way, “E”vergreen), you will find precious little polemic but plenty of well-supported arguments for the value of open source to libraries — an argument Breeding largely accomplishes in his report.
Open Source Integrated Library Systems has some small errors suggesting a second edit, perhaps with some accompanying phone calls, might have been in order. On one page Georgia PINES, the mother ship for Evergreen, has 266 libraries, and a page later, more than 270. The latter is correct; 275, to be precise, though to use that number in press releases, where no doubt this fact was gleaned, would sound too Urkel-like.
But overall, Open Source Integrated Library Systems provides insight into the juggernaut of open source in libraries; it’s a quick ramp-up for library professionals trying to get up to speed on open source. Furthermore, this report and accomplishes something that for many in our profession, only a printed book can do: by virtue of its being published at all, and through Breeding’s stature in our field, this report, however slender, casts the imprimatur of sober respectability on a development model that has its historical precedent in well over one hundred and thirty years of innovation in our profession. If Breeding does not entirely close the loop to electrify that point, he can be forgiven. The Book can now be placed on the Shelf.



















