Way back in 1998 a number of confidential Microsoft memoranda discussing Linux and other open source products were leaked and then published on the web. Called the "Halloween Documents", the memoranda identified open source software as a potentially serious competitor to Microsoft's product lines and discussed a variety of strategies for combatting it. One of those tactics, colloquially known as "FUD", involved spreading "Fear, Uncertainty and Doubt" about the viability and enterprise-readiness of open source. According to this now-defunct Geocities page, FUD is "scare-mongering" that is "used via 'gossip channels' to cast a shadow of doubt over the competitors offerings and make people think twice before using it."
Now, 18 years is an awfully long time in technology, and the days when a savvy marketer could delay open source adoption with FUD are long gone. In fact, even Microsoft has long since given up trying: in the years between then and now they have made substantial contributions to Linux kernel development, and Windows 10 now comes with a Linux subsystem that allows you to run Linux applications directly on Windows.
But hey, just because the rest of the world has moved on, that doesn't mean Library Tech has to. We are, after all, an island unto ourselves and if we want to stick with our closed, monolithic proprietary software stacks then that's just what we'll do. Anyhow that's my takeway from Carl Grant's latest screed, wherein some rather FUD-like tactics are brought to bear on FOLIO, one of the more interesting products to emerge in Library enterprise space in quite some time. To me it seems a bit sad that the tired old arguments of decades past are still being trotted out in our backward little corner of the technology industry. It would be nice if we could move on, and to that end I'll offer some counter arguments (along with the usual disclaimer that these arguments are my own, not my employer's).
Of course, Mr. Grant doesn't argue that open source is a bad thing. It's simply not possible to do that anymore and retain any shred of credibility. Instead, he begins by noting that he is a "huge supporter" of open source software, because his library runs a lot of the standard packages, like DSpace, Fedora, and Drupal. That's nice, but I feel compelled to note that simply using open source does not make you a "huge supporter" of it. To support open source you need to resource its ongoing development, through contributions of staff time and/or money. And a true "huge supporter" probably wouldn't expend time and influence trying to derail someone else's open source project.
Mr. Grant's introductory section raises no substantive criticisms so there isn't much to respond to there. Mostly he raises "questions" about different aspects of the FOLIO project, a standard rhetorical dodge used to imply "problems" without actually defining any. For example, he identifies "Questions about the suitability of micro-services as the core architecture" as an "issue" without saying what the questions are, what the answers are, and whether and why those answers are good or bad.
The follow on sections are somewhat more substantive. I'll take them in the order they come.
1. Can the library community support the vision this project entails?
Most of this section boils down to the observation that libraries will be unable to supply much development assistance for FOLIO because of our pre-existing commitments. Right now this is probably true. But since EBSCO is underwriting FOLIO development at this stage, it doesn't matter. The open architecture of FOLIO is important because it allows vendors besides EBSCO to build products that work in the FOLIO ecosystem, and will become important for libraries after they have adopted the product. FOLIO enables community participation, it doesn't depend on it. Carl Grant argues "Enterprise software requires enterprise level development teams" and I agree. Surprising then that Mr. Grant doesn't mention FOLIO's team includes Index Data, who have been writing first-rate open source library software for over two decades, backed by EBSCO who have a strong and proven track record of delivering enterprise grade tech.
Then there is the observation that FOLIO is late to the game and that competing products (which Mr. Grant identifies as Ex Libris Alma and OCLC Worldshare) will be moving forward while FOLIO is in development, so FOLIO will never catch up. That may be true, but I will observe that was not the case for the previous generation of ILS software, which has been moribund since the early 2000s. Open source ILSes such as Evergreen and Koha did eventually catch up; even if you're late off the mark it's not difficult to win a race in which your opponents are standing still. Software stagnation is a well-known effect of lock-in: it is not uncommon for enterprise application vendors to under-resource development of their existing product lines when they've achieved market dominance. That's what happened to the previous generation of Library tech and I see no reason to believe it won't also be the case this time around.
As those of us who have been watching this space know, FOLIO grew out of the failure to launch of an earlier open source Library Services Platform called "OLE". Mr. Grant advances a number of arguments that I'll summarize as "OLE failed so FOLIO will too." But this is a false equivalence. FOLIO is not OLE: there are different players, different leadership, a completely new architecture and codebase. It is a different project, and it will succeed or fail on its own merits. To be clear: some open source projects do fail. But then, many proprietary software development projects also come to naught - for example, ProQuest is no longer developing their Intota LSP. Like anything else, open source can be done well or badly, and FOLIO is off to a good start.
The last argument in this section is rambling and difficult to parse. I believe it grows out of Carl Grant's (deliberate?) misunderstanding that FOLIO development relies on volunteer labour and so, lacking the firm hand of benevolent and all-knowing corporate authority, will never be able to settle on a "common defined product definition." This is a not-very disguised version of an ancient critique that open source software development models cannot deliver products of the same degree of complexity as the top-down approach common in proprietary software development shops. Nobody makes this critique anymore because experience has shown it to be wrong.
Mr. Grant asks, "Does anyone besides me see the vicious cycle here?" Well, now that he mentions it, I do see a vicious cycle playing out, just not the same one. I see a cycle of libraries moving back into walled off proprietary enterprise software gardens, after some years when we achieved a degree of standards-based interoperability that allowed us to, for example, choose an ILS and a discovery layer from different vendors. This move to a one vendor/one library model, commonplace in the 90s, is indeed a backwards move for our profession, eliminating the marketplace competition that would otherwise keep prices down and foster innovation. If FOLIO succeeds it will offer a way out of that vicious cycle, and for that reason it deserves our attention and support.
2. Listen to what you’re REALLY being told by the “organizations” involved!
This section is pretty short. The even shorter version is "it's likely going to be several years before FOLIO is ready for mainstream deployment." And that's probably true. So what's the hurry? The vendors keep telling us that running our current generation of enterprise software is terribly costly and inefficient and we have to upgrade as soon as possible ... Which is kind of odd, considering they're the ones that developed those systems in the first place. But the truth of the matter is the old systems aren't that bad, and the new ones aren't that good. Given the costs of migrating and the organizational disruption that will pose, there really isn't a compelling case to be made for upgrading right now. We can wait.
This section ends with a /really/ odd argument, a quote from the incoming ALA president James Neal to the effect that "I propose that by 2026, there will be no information and services industry targeting products to the library marketplace. Content and applications will be directed to the consumer, the end user." Carl Grant muses, "Think about that and the implications it has for FOLIO (and really, a lot of library software)." OK, here's what I think: James Neal is wrong. This is the same tired futurist nonsense about disintermediation that we've been hearing for years. In 2026 Libraries will still require a variety of software applications to get our work done. If we have money to spend, I guarantee you there will be vendors lining up to help us spend it. There are no implications for FOLIO, this is a red herring. So much so that Mr. Grant himself contradicts this argument in the next section, asserting that Libraries will need a whole bunch of new software applications in the years ahead.
3. “Follow the money”
Carl Grant's analysis of EBSCO's motivations in supporting the FOLIO project is likely correct as far it goes, but it leaves out the most important detail. Which is that EBSCO's major competitor (ProQuest) recently made an aggressive move toward marketplace consolidation that could result in a hit to EBSCO's bottom line, and FOLIO is one of the ways EBSCO is fighting back. (And by the way, notice how Carl Grant continues to refer to Ex Libris as though it was still an independent company, not a subsidiary of ProQuest). Major library vendors competing for market share is not a problem for consumers. If anything, this bodes well for FOLIO because it means that EBSCO, its major backer, is fully committed to the project.
Mr. Grant then criticizes EBSCO for developing a Library Services Platform instead of developing a whole bunch of other things that Mr. Grant believes we need more. Assuming Mr. Grant knows how markets work, this is a bad faith ploy. If libraries do in fact have money to spend on research data management systems, knowledge creation systems and the like, that presents a sizable business opportunity. Someone, if not EBSCO, will undoubtedly come forward to supply our needs. Just as ProQuest is free to use some of their profits to acquire Ex Libris rather than investing that money in software development, EBSCO is free to determine their own strategic priorities without first consulting Mr. Grant.
Then Mr. Grant, flying bravely in the face of every economist who ever lived, goes on to argue that marketplace competition is bad for customers. "We don't need more alternatives, the marketplace has already spoken" is his battle cry. I'll just note here that the technology market, even in our own little corner of the world, is not static. The marketplace has never "already spoken", in fact that idea is more than a little bizarre. Marketplace decisions are never final - if they were we'd all still be texting on Blackberries and gaming on Commodore 64s. Disruptive innovation is normal in tech, the market is in a state of ongoing flux, old orders are regularly overturned. The only ones who don't want more alternatives are the established players, because fewer alternatives means they get more business.
4. What is the governing structure going to be for this effort?
In this section Mr. Grant asks a number of questions about governance of the FOLIO project, asserts that we don't know the answers and says this is a cause for concern. Actually, he does ask some good questions and he is correct the answers aren't readily available. I am however less concerned than he is, as I suspect the answers are still being worked out. Right now, it appears to me that EBSCO is taking the lead on administration and defining the overall objectives, and Index Data are making the decisions about architecture and implementation. This makes sense given that EBSCO is footing the bill at this stage, and they're paying Index Data to do the work. In any case you really don't want "governance by committee" - or far worse, "design by committee" - in the early stages of a new software development project. You want good people making good decisions and writing good code. I'd be far more concerned about FOLIO if it had some kind of elaborate charter and byzantine committee structure to ensure every possible voice gets heard before any decisions get made.
The Open Library Foundation has some kind of role here as well but according to American Libraries back in April "details regarding how the foundation will be organized are still under consideration". All I would observe here is that there are many examples of successful open source projects involving multiple players governed by foundations, so it's not unreasonable to think the model could work for FOLIO too. The argument that OLE had governance problems is not relevant; as noted above FOLIO is a different project.
Conclusion
As you might imagine, I disagree somewhat with Mr. Grant's conclusions. I do agree with him that there is a place for mass community collaboration around open source software in Library Tech, although calling this the "FOLIO concept" seems odd to me. We already have such collaborations, many of them: Evergreen, Koha, DSpace, Fedora, Hydra, Islandora, Archivematica, Open Journal Systems, LOCKSS ... the list goes on. FOLIO is just one more, though a potentially important one.
The notion that governance must be worked out before anything else is done is contradicted by the history of open source itself. Imagine Linus Torvalds saying "No, no, I mustn't release the Linux kernel before the Open Source Development Labs exists to steward it." Of course, that didn't happen: the first version of the Linux kernel was released in 1991. The OSDL (now merged into the Linux Foundation) dates from 2000. Governance is obviously an issue, but no, a formal governance structure is not a prerequisite to open source success. If anything, I would argue that prematurely adopting a formal governance structure could do more harm than good due to the agility-limiting effects it might have. See the point about "design by committee" above.
Mr. Grant makes much of the fact that $5-$10M was "wasted" on the OLE project. Yeah, too bad about OLE, but the fact is that $5-10M is peanuts in the library automation space, particularly when it's spread out across multiple institutions and multiple years. Open source software has saved our organizations many multiples of that amount already, so I'm pretty sure we can eat the loss. This is hardly the bankrupting, slink off with tail-between-legs disaster that Carl Grant paints it as.
Finally, Mr. Grant ends with a plea to redirect our efforts toward solving "some of the many real problems that we really need to be solved." I would counter that FOLIO addresses the most significant problem in the library automation space right now: serious market consolidation, with all the problems that anti-competitive vendor monopolies pose for innovation and cost containment. If we don't get those factors under control, we open ourselves up to the real possibility that we won't have the resources to deal with all the other areas demanding our attention. We have only to look at the journal publishing space to see what happens when we have no choice but to buy from a dwindling number of suppliers with the ability to extract monopoly rents. We don't need to let the same thing happen in the library automation marketplace, but it could easily happen if we all passively follow Mr. Grant down the road of least resistance.