Windows, Mac, and Linux version naming schemes explained

18.09.2012
Do software version names matter at all? Or are we all just curating a rich library of goofy things for future archaeologists to laugh at? Does anybody buy, use, or trust the code on their computer, tablet, or phone based on its name?

I wanted to get some real answers to these questions, and give the crafters of silly software names a chance to speak. I imagined hearing about some cah-razy IRC chat discussions that led to the naming of Ubuntu 11.10 as Oneiric Ocelot. I prepared to press some hard questions about how Fedora moved through versions 7, 8, then 9 as "Moonshine," "Werewolf," and then "Sulphur." I expected to hear something quite formal and corporate from Microsoft, and then nothing at all from Apple.

I traded emails with community managers at Ubuntu and Fedora. I hit up press contacts and known reporter-helpers at Microsoft and its PR firm, and they pointed me to a blog post. I reached out to Apple and, indeed, heard nothing back at all.

The best answer I've seen, or at least the most seemingly true and and simple, comes from , answered by an Android system engineer and voted up by other known Android community figures. Android works with "Tasty treats" for version names ("Cupcake," "Eclair," "Ice Cream Sandwich"), so as to "decouple the development name for an OS release from the final version number, since version numbers often will change for marketing reasons fairly late in the game." Was that an honest admission that engineers might think something is a .1 upgrade, but marketers in the same company might want to call it a whole new thing? That version names are really just an attempt to mark a point in time, rather than a statement of intent?

Then again, every software project has its reasons, and a history of decisions to guide or sidetrack them. Here are some of the most interesting.