by Dr James A J Wilson.
Open source software seems to promise a great deal - reducing total costs of ownership, avoiding the risk of being locked-in to a proprietary vendor, improving interoperability, even offering better security than proprietary software. Almost all generic business functions, from creating presentations to running web servers, can be done using open source software. And it's free. So why aren't more organisations using it?
The fundamental characteristic of open source software is that it is free, both in terms of price, and in terms of freedom. The source code that constitutes the software can be altered by anyone with the requisite programming skills, and their contribution then fed back to the development community or redistributed to others.
OSS Watch, the UK's advisory service for open source software in further and higher education, recently conducted a survey (http://fastlink.headstar.com/oss5) of IT directors in the sector to find out what people thought of open source software, and their reasons for using or excluding it. Although our work was undertaken in the education sector, the opinions we found are likely to be more broadly indicative.
The principal concern respondents had concerning open source software on desktop machines was training needs. Some 79 per cent of those surveyed indicated that this was an important factor which counted against migrating to open source. The second most important factor was user expectations. Users expected to have the familiar industry-standard software on their machines, and IT directors were cautious about presenting them with unfamiliar tools. A number of other issues were mentioned as concerns by over half of those polled: the costs of the migration; whether there would be interoperability problems; the potential need for third-party support; and the time involved in simply identifying the relevant software.
In open source software's favour, 60 per cent believed it would provide lower total costs of ownership, and more than half thought it would overcome the risks of vendor lock-in associated with proprietary software.
Respondents were generally more bullish about installing open source software on servers than desktops, presumably due to the greater technical know-how of the staff dealing with server maintenance, plus the presence of well-established and respected open source software for servers.
So, are people's concerns about open source software justified? Firstly, it is worth noting all of the issues raised would, in fact, apply equally when switching between different proprietary software systems, and several of the concerns are common to any organisational change.
With regards to training needs, most users familiar with Microsoft Office should in practice have little difficulty migrating to OpenOffice, which has a similar feel and duplicates most of the basic functions. Indeed, as the software market has matured, many applications have become standardised, so users familiar with one package will usually be able to get to grips with another serving the same purpose without too many difficulties. This is not to suggest, however, that some retraining would not be required, especially for less technically-minded staff. The Linux operating system, which is an alternative to Windows, is very powerful, but in many respects quite different. While it is becoming easier and easier to use, some training would almost certainly be required if it were to be rolled out across an organisation.
As for requiring external assistance, several open source companies base their businesses on paid third-party support, however this is rarely very expensive, and, because communities of users and developers tend to congregate around popular open source projects, there are invariably online forums in which help and assistance is also given without charge.
Interoperability is generally a strength of open source software rather than a weakness. Whereas proprietary developers may have a motive to make it difficult to access data in software produced by their rivals, open source developers have no such motivation. Instead, they tend to use open standards that can be easily interpreted by other developers.
Questions about migration costs cannot be answered without proper cost benefit analysis of the specific software solutions under consideration.
If migration is well managed, and staff are kept informed and involved in the decisions, the process is always cheaper and easier than it might otherwise be. Studies suggest that staff buy-in can be crucial. User expectations should be managed, not ignored. Open source software can often result in longer-term savings, but these may have to offset an initial loss of productivity.
As with all procurement decisions, it would be unwise to foreclose a significant part of the market, whether that part is open source or proprietary software. Schemes such as the 'Business Readiness Rating'
(http://www.openbrr.org/) and 'Open Source Maturity Model'
(http://fastlink.headstar.com/oss6) have been developed to help assess whether particular open source software packages are mature enough to meet an organisation's needs, although these are not so good for comparisons between open source and proprietary alternatives.
The use of open source software can bring substantial benefits to an organisation, and the perceived barriers to adoption are sometimes exaggerated, but time and care must indeed be taken to choose the right software, and the potential costs must be weighed against the freedoms.
NOTE: Dr. James A. J. Wilson is OSS Watch Research Officer and Intute Arts and Humanities Content Editor, Research Technologies Service, Oxford University Computing Services.
