There have been some very interesting posts on open source licenses in the past few days – prompted by a complaint about IBM's strategy in integrating the Symphony development work with the OpenOffice project.  

What is interesting about these blogs is that they make good comparisons between the Apache (permissive) and GPL (restrictive) licensing paradigms.  Ian Skerrett leads off with a blog in which he takes issue with the claim that “LGPL … keeps developers 'honest'; the Apache-License 2.0 will not”.  Both Ian's article and Matthew Aslett's blog make a strong case that permissive licenses (Apache et al) are slowly displacing copyleft/restrictive licenses.  This is certainly something that we see in OpenNTF – where the authors of over 90% of our submitters have chosen to use Apache rather than one of the GPL family of licenses.  (And those who do choose GPL tend to have far more issues with incompatibility of components – but that's another story.)

Finally, the most comprehensive of the blogs is Douglas Heintzman,s explanation of IBM's strategy of transitioning Symphony development to the Apache-based OpenOffice project.  In developing his argument, Doug provides a good description of the viral mechanism that copyleft licenses (such as GPL) use to “enforce disclosure of code modifications” - and contracts that with the permissive licensing regime – and why the latter is “more attractive to corporate vendors, and facilittes corporate investment of resources.  

Well worth reading.


comments powered byDisqus