OSS Watch survey findings: interoperability a major factor in educational Open Source adoption.
JISC's new Open Source Software Watch (OSS Watch) pilot advisory service was formally launched last week with a conference and the presentation of a scoping study of free and open source use in UK Further and Higher Education. In the study, the single biggest factor that Higher Education respondents mentioned as a motive to choose open source was interoperability through better standards support. Intriguingly, the single biggest factor holding back open source software adoption is ... interoperability and migration concerns.
The Oxford University based OSS Watch pilot service is set up to provide UK further and higher education with neutral advice and guidance about free and open source software. The new service wisely started off with a fairly wide ranging survey-style study of the kinds of uses Free/Libre/Open Source Software (FLOSS) are put to, what kind of skills in its use exist and what the attitudes towards and plans for FLOSS are like.
The results were presented at the inaugural "Open Source Deployment and Development" conference in Oxford last week. The conference also included talks by people ranging from the Bodington VLE's Jon Maber, via Jim Farmer of JA-SIG fame, to Nick McGrath of Microsoft Ltd.
The scoping study's main findings include a noticeable gap between FE and HE: 73% of HE respondents, for example, reported that their organisation has either looked seriously into OSS and/or has already made some decisions on its deployment. By contrast, 61% of FE institutions said that very few members of their organisation were more than slightly aware of open source concepts, and only 15% had made decisions on deployment of OSS. Which, incidently, makes one wonder what all their websites are running on, given Apache's dominance of the webserver market.
The FE / HE gap is also noticeable in the reported skills in using and developing FLOSS: 88% of FE correspondents reported that very few of their staff had such skills, while 59% of HE correspondents reported that staff had moderate to good skills with FLOSS.
Lastly, the actual motivation for 'going open source' also shows a gap: 26% of HE correspondents cited "interoperability & open standards" as the main reason to adopt- the biggest factor in that group, just ahead of 'cost'. FE correspondents overwhelmingly -53%- cited total cost of ownership as the main attraction of free and open source software.
Where it gets interesting is the main concerns about choosing FLOSS solutions: both FE and HE cited 'Interoperability & migration concerns' as the top worry.
The report's author, David Tannenbaum, claims that this apperent paradox can be explained by any combination of the following factors:
Of these, the first two are virtually unknowable, as it depends on the research each of the individual implementors did prior to installation.
- that those who adopted OSS for this reason were operating under false beliefs
- that those who are concerned about interoperability do not have full information about OSS applications
- in some application areas interoperability is a concern, while in others it is not.
- in this question interoperability is was [mixed, WK] with migration concerns, whereas in the earlier question it stood alone
The issue of patchy support for standards in different FLOSS application categories, does look like an important factor: of the nearly fifty VLEs listed on open source development site SourceForge, for example, only a handful have or claim any e-learning standard support. By contrast, the dominant open source web browser, Mozilla and its many variants, is widely acknowledged to have the deepest support for web standards of all modern browsers.
There are some good reasons for this, to be sure. Modern, web-based VLEs are relatively new as a type of software application, while webbrowsers have been around for quite a while now. Also, it took the Mozilla project's many contributors over four years to get from the decidedly non-standard Netscape 4.x codebase to the nearly complete W3C compliance that we have now. From a user perspective, VLEs may have yet to hit upon the painful lack of interoperability that plagued the web back in 1998.
More generally, the link between open source and open standards is not as obvious as it may look initially. It is true that there is not much incentive for open source developers to lock users into their products, and therefore more incentive to play nice by supporting standards; particularly since many open source projects rely on users-turned-vendors. On the other hand, the availability of the source code makes the danger of vendor lock-in virtually non-existent, and even interoperability between systems of different kinds is in many (but not all!) cases fairly easy: you can dive right in, and tweak until it works.
The unfortunate side-effect of this freedom is that it appears to lead some FLOSS developers to think that they don't need to support standards at all. This is amplified by the open source community's abhorrence of duplication of effort. To your average FLOSS developer, fifty VLEs looks like a fantastic waste of time and effort. Two or three is just about tolerable, but everyone should really focus on the one and only definitive package. Who needs interoperability when you all ought to use and develop the same thing? Hubris, it seems, can amplify this further: my open source application is the greatest, why would anybody use anything else?
The Mozilla project, by contrast, couldn't afford such ego building luxuries. The only way it could compete with the overwhelming dominance of proprietary browsers was to be scrupulously standard compliant, and support all the quirks of the competition as well. Again, unless we get our act together, the VLE market may yet have to hit that 1998 state of affairs.
Tannenbaum's fourth factor --the mixing of interoperability and migration issues in the survey's 'worry about open source' question, but not the 'appeal of open source' question-- may indeed also have been a considerable factor. Considered sequentially, it makes perfect sense: college system manager is trapped in proprietary solution, looks wistfully at the flexibility and standards support of FLOSS solutions, but doesn't dare switch, because she knows that all sorts of stuff will break, because that's what the proprietary stuff is designed to do. Certain widely used closed source file formats and file & print protocols, for example, were specifically engineered not to be interoperable. So unless you can move the whole thing in one go, there is the possibility of something breaking.
And that, of course, is one of the reasons why the OSS Watch could play such a vital role in the UK FE and HE sector. The scoping study already recommends, among many other things, that more nuts and bolts information about migration and interoperability issues should be provided. Given the pragmatic approach of the OSS Watch so far, and the sector's stated preference for a mixed proprietary and open source environment, that seems eminently sensible.
More information about the new JISC pilot service is available on the OSS Watch website. There is also an executive summary of the scoping study, and a full fat version (17 kb, pdf). You can also download the conference's presentations from the presentation page.