The blog of Kristóf Kovács

Lately I was thinking about quantities of code.

More specifically, how the codebase of a typical web application relates to the open-source code used in it?

For an example, let’s use a mutual-fund search web application we developed for a brokerage about a year ago. (Disclaimer about the design: they wanted feng-shui compatible, ok? We deliver.) It’s written in PHP, uses MySQL, uses an MVC framework (an in-house one, so in the calculation I’ll use CakePHP), uses jQuery, and is served by an Apache that runs on FreeBSD. All of these projects are on ohloh.net, so I’ll use the numbers (number of code lines, development cost in USD, development time in man-years) from there.

I’ve made a chart of how the code size measures up:

What? Let’s see it zoomed in, once again:

I don’t know about you, but I was blown away how true that statement is, “We are standing on the shoulders of giants”.

We can’t even imagine how expensive software development would be without open source components.

And I’m not even counting the development tools like source control (git, mercurial), editor (vim, emacs, eclipse, netbeans), testing tools (selenium, ab), and others…

Recommended reading:

Build, Buy, Open Source or Services? at CapGemini

How Linux is changing the cost model for embedded design at Electronics Weekly

Open Source Business Tactics in One Slide at Once More unto the Breach