BlogBack

EVERGREEN 2013: LINUS’S LAW

By 2013 Evergreen was, to coin a phrase, “nominally complete.”  It had gained the features needed to check off most of the right RFP boxes, and so be considered alongside other ILS’s with a significantly older code base.  Acquisitions and serials, along with circulation, cataloging, authority control, and the (underrated, in my opinion) booking functionality were all in place.  By this point it had a modern, pluggable OPAC infrastructure, integration with many 3rd party products to expand its functionality, and was attracting attention via non-traditional use cases such as publishing house backend systems.  So, we developers were done, right?

Not at all.

In years past, the development team working on Evergreen had been small, and grew slowly.  In important ways, though, that began to change around 2013.  Previously, having more than twelve distinct contributors in a month submitting code for inclusion in the master repository was quite rare, and usually happened right around the time when a new release was being polished.  But from late 2012 through all of 2013, 15-25 contributors became the rule and less than that was the exception.  That is a solid 20-30% increase, and is significant for any project.

At the software level this was a period of filing down rough edges and broadening the talent pool.  There were few truly massive technological advances but there were many, and varied, minor improvements made by a growing group of individuals taking time to dive deeper into a large and complex codebase.  Importantly, this included ongoing contributions from a Koha developer on a now-shared bit of infrastructure, the code we both use to parse searches against our respective catalogs.

In short, 2013 is the year that we began to truly realize one of the promises of Open Source, something that is attributed to Linus Torvalds of Linux fame.  Specifically that given enough eyeballs, all bugs are shallow.  What this means is that as your project adds users, testers, and developers, it becomes increasingly likely that bugs will be discovered early, classified quickly, and that the solution will be obvious to someone.

In some ways this can be a critical test for an Open Source project.  Many projects do not survive contact with an influx of new development talent.  For some projects, that is political.  For others, it is a consequence of early design decisions.  Fortunately, Evergreen passed that test, and that is in large part a credit to its community.  After seven years and significant scrutiny, Evergreen continued to improve and its community continued to grow.

— Mike Rylander, President

This is the eighth in our series of posts leading up to Evergreen’s Tenth birthday.