My History With Open Source

From CPAN to Design Patterns

Throughout my career I've benefited greatly from being able to utilise open source software that other developers have produced and made freely available.

Some of my earliest commercial project work benefitted from libraries made available for Perl via the Comprehensive Perl Archive Network (CPAN). It sometimes felt like our company had a huge advantage over organisations that used VB Script for developing ASP pages, as they seemed to be tied into the world of closed source and needing to pay to use libraries that other organisations had developed as a licensed product for sale.

In the early two thousands I was continuing my university studies as a part time student while working as a software developer. One of the distributed systems courses gave me some exposure to JBoss and Tomcat, which made me question why we were paying to use commercial application servers for some of our clients' projects in my day job.

Aside from the common day to day helper libraries such as JUnit and log4j, Tomcat was probably the first major open source Java system that I brought into my work, proving out that we didn't need EJBs and all of their standards for getting some data out of a database and onto some web pages. At around the same time we were probably dabbling with Apache JMeter as a mechanism to validate that this new kid on the block (well, our block at least) was going to cope with what we anticipated was going to be thrown at it.

Although we didn't use any particular library for it, I would also consider design patterns as an example of shared knowledge that really helped us to achieve our scalability and performance goals. Safely caching data that was being read frequently, but only being updated only a few times each day. If you went skiing in New Zealand in the early 2000s then I can almost guarantee that you checked snow reports using code that I developed.

Giving Something Back

Open source licenses can be a legal minefield for the owners and the users of products and libraries.

Working in large corporations often involves policies and even clauses in employment contracts - along the lines of "If you develop on company time, or on company hardware then anything produced as a result is the property of the company" and / or "The use of any open source software is expressly forbidden unless it has been formally approved by the XYZ committee".

Even smaller companies need to be aware of the differences between GPL, MIT, Apache and other variations of licenses before building a product up.

So far my main contributions to open source projects have mainly been limited to minor improvements to the documentation, and a couple of minor bug fixes for some smaller projects. Correcting typos and improving grammar can be a small way of helping out - providing that it isn't pedantic or debatable whether the new phrasing is better. So far I have had all of my contributions accepted with gratitude, as the original developers sometimes have English as a second language or just slipped up a little in the rush of getting something out and released.

Personally, I also find that by contributing to explaining how something works I can improve my ability to understand and recall that information later on. So, as well as being a good way to make an initial contribution to an open source community, consider that by improving your understanding you will also be moving some way towards being able to contribute to the code as well.