I've been a software engineer on Unix and Unix-like systems since 1982. I've been writing iOS apps since iOS 2. I also write server scripts in Perl, both for system management, CGIs and data transformation.

I've written extensively about transparency in software projects and am currently working on a research project to understand how transparency in software projects is related to sexism, racism and other forms of unwanted stratification. My goal is to understand more about why there are so many/few programmers are that are male, female, people of color, or other ethnic backgrounds.

My hope is that the results will help organizations reverse the trend, and arm advocates of transparency and justice in the workplace with data they can use to show which kinds of practices are forms of stochastic sexism or racism: even if specific actions are not sexist or racist, their combination of effects can be. A company can inhibit transparency by failing to provide time and resources. If that is shown to correlate with sexism, it is an interesting finding to say the least. Does failure to invest in documentation by and for engineers keep women out of the software industry? I do wonder...

We have a lot of smart people in our industry, but we don't seem to understand why there is an imbalance. At least, there doesn't seem to be any formal research on this topic. Also, data do not exist to help us understand the relationship between transparency in a software project and sexism, nor data about why we see some groups of people so infrequently among the ranks of programmers.

I am trying to do something about this by creating a survey to collection the data needed to show at least that transparency in software projects correlates to an environment where anyone can thrive regardless of gender, race or cultural background. I happen to believe this is true, but I certainly can't prove it or even argue strongly without data.

I believe transparently run software projects are better at achieving their goals and I hope to show that too, at least by correlation. Causation is quite a bit harder, but we'll see what the data can show.

Read more at http://www.transparentsoftware.org