Too often, open source programs have bizarre, unhelpful names. This needs to change, at least for desktop applications, for Linux to really take off in that market.
For all my love for Linux, software names in Linux are probably the most cryptic, arcane names possible. To understand them, you need to have seen most TV shows and sci-fi movies in the last thirty years. Preferrably, these should be the ones that no one else saw. You must also have a large zoological encyclopedia on your bookshelf. At the time of this writing, the site where the newest open source software is announced, freshmeat.net, had listed the following new releases (this was a couple years ago now):

  • Kile beta 1.0
  • JDHCPSim 1.1
  • JavaGuard 1.0b1
  • Cerberus Helpdesk 1.1.5
  • TyGeMo 0.2.0
  • Gibraltar Firewall 0.99.3a
  • archmbox 3.0.2
  • Star Voyager 0.4.0
  • KDevelop 2.1
  • OpenVPN 1.1.0
  • SPIKE 2.1
  • php_jh_pdf 1.1.0
  • lintad 0.0.3
  • KDE 3.0
  • uClibc 0.9.11
  • samhain 1.4.7
  • phpTest 0.5.0
  • HylaFax 4.1.1
  • JBoss 2.4.5
  • phpmailer 1.60 (Stable)
  • wmmemload 0.1.2
  • Rolodap 1.0
  • SmartTAN 0.1.1
  • userinfo 1.6 (Stable)
  • Sidewinder Forcefeedback 2 driver 04102002

Look at those names. How many are readily understandable to a non-geek? Hmmm. Another thing is the version numbers. I’m not saying that they shouldn’t be accurate. These numbers reflect exactly where the program is in its development. The difference between these and commercial programs is that commercial programs usually have announcements of a new version on two occasions:

  1. there has been a recall or security patch or
  2. they have hit another whole number version (like 2 or 3).

To be fair, freshmeat is a site where you announce the “freshest” release. I think it’s partly because to me, and I thus assume it to be true of all geeks, it’s fun to install new software and try it out. And this way, you can have the very latest of everything. Many open source software are what you’d call server applications. They are really for the backend or administration of a website or network. And since those that care for those areas are geeks, it’s okay for them to be a little more exact.

Open source names often have puns (notice HylaFax – Halifax, Canada). They also often incorporate a bizarre word (samhain) or the name of an animal. Common names of animals as well as genus and species names are fair game, too. For non-geeks (and even geeks), this makes it necessary to read the description of everything. You’ll notice in the list that some kind programmers have put a helpful indicator in the name of their program. If it starts (or ends) with “php” it probably has something to do with the PHP server programming language, whereas a “j” would usually indicate a problem based in some way on the Java programming language. Those names with a capital “K” usually are designed for use with the KDE, the K Desktop Environment. Programs that start with a “g” are likewise designed for the Gnome desktop. “Gnome” itself is a bizarre mixture of fancy, puns, and acronyms. It is actually an acronym for GNU Object Model Environment. GNU in turn stands for GNU Not Unix, itself a programming joke because this acronym causes what’s called an infinite loop- the acronym repeats itself forever. This is amusing to me, but I’m a programmer. I think most people would find it odd. Gnome, however, is a fine name, but I don’t like to think of as an acronym. I like it anyway.

When applications are written, they should be better named. It’s not that the name needs to be descriptive, but it should be a name perhaps combined with an explanation. Mystery. Don’t make users have to deal with that. Next week we’ll talk about good names. The surprise users feel about your app shouldn’t be that they would never have guessed that an app with this name would do that– your app should take their expectations and exceed them with features that make their lives easy.