Pages

Open Source Estimation

Open Source software development is an interesting phenomenon where the software development is collaborative and source code is open and available to public. The renowned literature of the ‘Cathedral and Bazaar’ by Eric S. Raymond was published more than 10 years back and still considered a pioneer work in the field of Open Source. Unfortunately, there are no well-defined processes and estimation techniques (models) in the Open Source world like the COCOMO, function points or SLIM models for the classical software development. Its very interesting that even with presence and support from the large organizations like Google, IBM, Oracle and tons of others Open Source world has not come up with any standardization. I agree that there may be well-defined processes and best practices in commercial open source software companies like Red Hat, MySQL and SugarCRM but certainly not a universally accepted standard. One reason may be that Open Source development is very unpredictable and largely depends upon the motivation of a handful of developers who are committed to develop such collaborative software. There are thousands of projects registered on SourceForge and FreshMeat (largest repository of the open source software projects) by a single developer, which never proceed further to development. There are also projects that are currently in the inactive stage.

I guess with boom and wide spread of Open Source today we certainly need some estimation techniques for the required number of developers (effort) and expected size of the community and thus an overview of the complete ecosystem around the software. Such models or processes will certainly help the community for better planning and effective execution of an Open Source project.

OSS Processes


Open Source software development is an interesting phenomenon where the software development is collaborative and source code is open and available to public. The renowned ‘Cathedral and Bazaar’ book by Eric S. Raymond was published more than 10 years back and is still considered a pioneer work in the field of Open Source. Unfortunately, there are no well-defined processes and estimation techniques (models) in the Open Source world like the COCOMO, function points or SLIM models for the classical software development. Its very interesting that even with presence and support from the large organizations like Google, IBM, Oracle and tons of others Open Source world has not come up with any standardization. I agree that there may be well-defined processes and best practices in commercial open source software companies like Red Hat, MySQL and SugarCRM but certainly not a universally accepted standard. One reason may be that Open Source development is very unpredictable and largely depends upon the motivation of a handful of developers who are committed to develop such collaborative software. There are thousands of projects registered on SourceForge and FreshMeat (largest repository of the open source software projects) by a single developer, which never proceed further to development. There are also projects that are currently in the inactive stage. I guess with boom and wide spread of Open Source, we certainly need some estimation techniques for the required number of developers (effort) and expected size of the community and an overview of the complete ecosystem around the software. Such models and processes will certainly help the community for better planning and effective execution of an Open Source projects.