Open Source TCO: look at the COSPA recommendations (part 2)!
The EC funded COSPA project defined two frameworks to better understand open source software adoption and cost of migrations in public administrations. COSPA findings below reported grounded as they are in practical cases, form a basis for management and policy-makers, interested in this area.
Today’s recommendations by shinyai
The “Report evaluating the costs/benefits of a transition towards ODS/OS for each key task related to personal productivity used in the PAs under study” – available thanks to the Wayback Machine – benchmarks the effectiveness of the deployed OSS solutions through a statistical and cost/benefit analysis using those frameworks.
The study has been conducted applying the framework on assimilation theory, identifying the most relevant factors. Then such assimilation factors have been differentiated into facilitators and inhibitors to OSS adoption. The last section of the document eventually present a study on the migration costs and the cost of ownership.
COSPA work compares in detail 6 PAs (Beaumont Hospital – Ireland, SGV – Italy, Extremadura – Spain, Skopje – Macedonia, Pisa – Italy and Törökbálint – Hungary) in terms of assimilation models and migration costs.
On the basis of its findings, COSPA has formulated the following recommendations.
Recommendation 1. (achieving a general level of OSS deployment) To achieve a general deployment of OSS, COSPA recommends that PAs focus on the specific facilitators and inhibitors to OSS assimilation we have identified, prior to migration. Specifically, COSPA suggests recognising that technological benefits of OSS outweigh its disadvantages – e.g. ability to tailor to precise needs, transparency – as these are important facilitators in the assimilation of OSS. In contrast, it is important to overcome the perception that employees might feel their work is under-valued if using ‘cheap’ OSS products, and also the perception that changing operating models to OSS might be problematic – e.g. no contracted maintenance support.
Recommendation 2. (savings on costs) To base the decision to migrate to OSS to save on licenses costs alone is unrealistic as they are only initial costs, all too easily influenced by inflation and market fluctuations over time. COSPA recommends the decision to be based on two related evaluations: costs of migration and costs of ownership. The former involves high investment for a shorter period, while the latter foresees expenditure for maintenance over a period of at least five years. In the migration, COSPA findings report that a substantial factor are the intangible costs such as costs for peer training. COSPA also reports that there are no extra costs due to lack of productivity arising from the use of the OSS solution. Although training costs are a substantial part of the migration costs their benefits can be realised over the long term in terms of costs of ownership. People are more conscious of the software they work with when they have been trained on open source code. This gives more power to them in negotiating fees for consultancy and maintenance.
Recommendation 3. (barriers in migration to OSS). As any new radical IT innovation, a transition to OSS involves the discussion on barriers to migration. COSPA analysis has reported that barriers may arise in several areas: a lack of knowledge/experience in relation to what OSS products are appropriate and how they might be deployed. COSPA recommends a policy of both ad hoc and periodic training to help achieve the benefit of a transition to OSS. In the COSPA findings, some of the technical reasons that determined the success of a migration were exchange of documents in an open shared format (ODS), utilization of old hardware in high schools, being independent of software vendors even when creating a distribution or an application for local needs. COSPA recommends considering this factor before deciding to migrate, as the migration costs might not be really affordable and other reasons may need to be taken into consideration.
Reply