OSS Inaugural Conference: Open Source Deployment and Development
OSS Watch [1] is a pilot advisory service set up by the Joint Information Systems Committee (JISC) to provide UK higher and further education with neutral and authoritative guidance about free and open source software and related standards. Although it is rather small, (a staffing of 1.25 full-time equivalent (FTE)), this new service has stakeholders ranging from IT directors and managers developing institutional IT strategies that acknowledge the role that open source software does (and will continue to) play; to IT staff deploying software across universities and colleges; and to software developers seeking advice on how to release their work as open source. It's a daunting task.
OSS Watch's initial focus has been to conduct a scoping study of the current situation in UK HE and FE in order to inform its plans for the coming years. The report from this study was released at a conference held in Oxford on 11 December 2003. Others then set the scene for open source in UK HE and FE. If nothing else, the conference revealed that one of OSS Watch's principal contributions may lie in facilitating opportunities to share best practice.
About 150 people attended from right across the spectrum of educational institutions, public bodies, software companies, and open source advocacies, and there was lively discussion at all times.
Programme
Plenary Sessions
The day began and ended with plenary sessions and between these were two parallel workshops with one strand following the deployment theme and the other the development theme. After a warm welcome from Reg Carr, Chair of the JISC Committee for the Information Environment (JCIE), the first plenary session, Open source in academia - where are we?, started with a presentation from OSS Watch manager Sebastian Rahtz and researcher David Tannenbaum. They spoke on the results of the scoping study on open source deployment and development in UK HE/FE which had been conducted in autumn 2003. The full report from this scoping study [2] is available on the OSS Watch Web site. Key findings included:
- 38% of HE and only 9% of FE respondents reported that their institutions have an IT strategy which explicitly considers OSS
- FE institutions lag behind HE institutions in their number of staff with the skills to deploy OSS
- HE institutions are much further advanced in the process of examining the potential of OSS solutions for their institutions
- Reasons for choosing OSS solutions in both HE and FE are principally financial
- The majority of institutions which develop software in-house (59%) have never considered the issue of licensing, and do not have an institutional licensing policy
No great surprises, probably, but it is important to gather more and more data of this kind to back up what are often simply guesses about the state of open source in the UK education sector.
After presentation of the study, Sebastian Rahtz outlined a programme of events for OSS Watch in 2004 to meet the recommendations.
The second presentation in the initial plenary session was from Jim Farmer, project administrator for JA-SIG's uPortal Project [3]. uPortal is a portal framework developed by a consortium of universities in the USA with funding from the Mellon Foundation; it is now being deployed at a number of sites in the UK. Jim brought out some important points about the funding model and how it is now supported by both open source academic programmers and also commercial partners. It was a salutary reminder of the range of open source projects, which nowadays go far beyond the simplistic model of lone enthusiasts working at night.
The plenary session concluded with a general discussion which warmed up nicely as people reflected on the history and meaning of open source.
Two Strands: Deployment and development
The conference then broke into its two strands. In the deployment strand, Making the institutional case, we saw presentations from Andrew Findlay of Skills 1st Ltd [4] and Brian Kelly of Web Focus [5], UKOLN. Andrew concentrated on the Secure Open Desktop Architecture (SODA) [6] being developed by Netproject [7]. With a scale from 10 to 100,000 desktops, SODA demonstrates the potential for open source desktop deployment on an institutional level. Brian Kelly's presentation, 'Open Source? No, Open Standards!', reminded participants that the most important issue for deployment was interoperability, and for that there really is no substitute for open standards. It does not follow that just because software is open source that it will conform to open standards. So when selecting software we need evaluation criteria to spot the appropriate solutions and avoid the inappropriate ones. Brian was the first speaker to raise the spectre of software patents and gave some salutary lessons about how they may work both for and against us.
In the development workshop in another lecture theatre, Jon Maber of the Bodington Virtual Learning Environment (VLE) Project [8] was speaking on 'How do you make an open source project?' Jon was the sole presenter in this workshop in the absence of Ben Lund, Nature Publishing Group, due to illness. Jon was very candid about the mistakes they had made and about the occasionally accidental nature of their success; but he also made it clear that a successful project needed a lot of hard work and commitment. It provided a useful lesson in the nuts and bolts of building up a piece of software to the stage where it could be used by other sites.
Another set of parallel workshops followed on after lunch. The deployment strand saw Paul Browning, Information Strategy Co-ordinator at the University of Bristol, report on some of his tales from the front line especially regarding the deployment of Zope [9] and, more recently, uPortal. Paul's presentation focused on some of the pitfalls encountered along the way and revealed another key aspect of OSS Watch's role - to provide an opportunity for colleges and universities to share experiences in open source deployment, whether they are successes or disappointments. One important point made by Paul was that an institution must make its own hard decisions about whether it can support a development process at all, regardless of open source and open standards.
Paul was followed by Henrik Omma, the founder of TheOpenCD Project [10]. TheOpenCD is a collection of quality open source software that runs on the Windows operating system - a perfect demonstration that open source doesn't necessarily mean anti-Windows. The project's work is to find stable software and package it with a good installer. The process of selection was interesting, in that they chose only one of each type of program ('best of breed'), and consciously limited themselves to mainstream programs, (Web browser, office suite, picture editor, etc). This looked like a healthy contrast to the glut of programs provided in Linux distributions, which not only provide the kitchen sink, but 5 alternatives to it as well.
In the development stream participants were focusing on Getting the right licence, with presentations from Susan Foster, Eversheds LLP and Andrew Charlesworth, Senior Research Fellow in IT and Law, and Director of the Centre for IT and Law (CITL) at the University of Bristol. Andrew's presentation concentrated on working through the variety of licences available, while Susan's focused on the broader institutional issues around intellectual property (IP). Questions from the audience indicated that the level of knowledge in this area is very basic and that practical guidance in future from OSS Watch would be welcome.
The final plenary session, entitled Does open source matter?, brought together presenters from two corporate giants. Jeremy Wray, Business Development Executive for Public Sector in IBM UK, concentrated on the pragmatic aspects of open source. He urged delegates to approach open source with an open mind rather than religious fervour, and to stay focused on open standards when asking whether any software solution is fit for purpose. Jeremy wasn't coy about IBM's involvement with open source. It's a business decision, and if they didn't think they could make money out of providing support for open source solutions they wouldn't be involved. He perhaps upset part of the audience by repeatedly saying that nothing was really free, ('if it looks too good to be true, it probably is'), which conflicted with the view of some open source work as comparable to charity in its selflessness.
Nick McGrath, Head of Platform Strategy, Microsoft Ltd, was the other presenter in this session. Although the promotional video with which he began may have been misjudged for this audience, the remainder of his presentation focused on Microsoft as a comprehensive solution to institutional IT needs. It may not have gone down well with the open source advocates in the audience, but inevitably this type of solution fits many local situations. It would be disingenuous to consider open source software without reminding ourselves of the real-world situation out there for IT decision-makers. Despite robust questioning, both the IBM and Microsoft representatives stood firm on the point that whatever decision you take, somebody pays.
We seemed to end up with a choice of corporate strategies: IBM work around open source by sweeping up lots of business behind it, and stress the role of open source as essentially a software development methodology; Microsoft prefer to offer a total alternative, and meet the free software movement head on. Whatever else may happen, one thing is sure - free and open source is centre stage at the moment, and everyone is taking it rather seriously.
After the plenary, a good many people stayed on for a reception, noted for its 'free beer' and heated debates on the quality of the Microsoft TV commercial.
Conclusions
When it was all over, we were confident that we had tackled a good broad range of issues during the day. Inevitably, not everyone had found what they wanted, or had explored their favourite subject far enough; but initial feedback was that the majority of delegates went away with a plan - software to look at, Web sites to visit, projects to evaluate, contacts to follow up etc. What more could one ask? OSS Watch itself has plans for 2004, to follow on from the conference.
It is going to:
- hold a workshop in the spring about licensing and IPR
- run a conference in the summer about support and training
- run an autumn workshop on interoperability and migration
- start a student-level experiment in desktop deployment with TheOpenCD Project
- work with Regional Support Centres to provide hands-on exposure to open source software for FE colleges.
It looks like being a busy, but interesting, year ahead.
References
- OSS Watch http://www.oss-watch.ac.uk/
- OSS Watch Scoping Study http://www.oss-watch.ac.uk/studies/scoping/
- The Java Architectures Special Interest Group (JA-SIG) http://www.ja-sig.org/
- Skills 1st http://www.skills-1st.co.uk/
- UK Web Focus http://www.ukoln.ac.uk/web-focus/
- The Secure Open Desktop Architecture (SODA) http://www.netproject.com/opendesktop.html
- netproject http://www.netproject.com/
- Project Bodington on SourceForge.net http://sourceforge.net/projects/bodington/
- Zope http://www.zope.org/
- TheOpenCD Project http://www.theopencd.org/