why tiling window managers matter

I've realized much to my chagrin that I haven't written a post on about the Awesome Window Manager in a long time. It's funny how window managers just fade into the background, particularly when they work well and suit your needs. Why then, does this seem so important to me and why am I so interested in this? Funny you should ask.

Tiling window managers aren't going to be the Next Big Thing in computing, and if they (as a whole) have an active user-base of more than say 10,000 people that would be really surprising. While I think that a lot of people would benefit and learn from using Awesome (or others), even this is something of a niche group.

As a result, I think something really interesting happens in the tiling window manger space. First, the project is driven by a rather unique driving force that I'm not sure I can articulate well. It's not driven by a desire of profit, and it's not driven by some larger Utopian political goal (as a lot of free software is). This is software that is written entirely for oneself.

That's the way most (ultimately) free software and open source projects start. A lot of emphasis in the community (and outside) is made on the next stage of the progress, where a project that was previously "for oneself" becomes something larger with broader appeal. Lets start with the git version control system which started because the kernel development team needed a new version control system, but in the past couple of years has become so much more, by way of phenomena like github and flashbake. The free software and open source worlds are full of similar examples: the Linux Kernel, Drupal, b2/WordPress, Pidgin/Gain, Asterisk, and so forth.

But Awesome, and the other tiling window managers will, likely as not never make this jump. There is no commercial competitor for these programs, they're never going to "breakthrough" to a larger audience. This isn't a bad thing, it just effects how we think about some rather fundamental aspects of software and software development.

First, if developers aren't driven by obvious "us versus them" competition, how can the software improve? And aren't there a bunch of tiling window mangers that compete with each other?

I'd argue that competition, insofar as it does occur, happens within a project, and within the developer rather than between projects and between developers. Awesome developers are driven to make Awesome more awesome, because there's no real competition to be had with Aqua (OS X window manger,) or other Kwin and Metacity (GNOME and KDE's window mangers), or even other alternate X11 window managers like OpenBox.

Developers are driven to "do better," than the people that preceded them, better than their last attempt, better than the alternate solutions provided by the community. Also, the principals of minimalism which underpins all of these window managers towards simple, clean, and lightweight code, inspires development and change (if not growth, exactly). This seems to hold true under anecdotal observation.

While there are a number of tiling window managers in this space, I'm not sure how much they actually compete with each other. I'd love to hear what some folks who use xmonad and StumpWM have to say about this, but it's my sense that the field of tiling window managers has more to do with other interests. Xmonad makes a point about the Haskel programing language. Stump is targeted directly toward emacs users and demonstrates that Lisp/Common Lisp is still relevant. Awesome brings the notion of a framework to window management, and seems to perfectly balances customizable with lightweight design. While Awesome is a powerful player in this space, I don't think that there's a lot of competition.

Second, if there's no substantive competition in this domain and if there's a pretty clear "cap" to the amount of growth, how are tiling window managers *not entirely pointless.*

I think there are two directions to go with this. First I think we're seeing some of the benefits of these window managers in other projects like xcb (a new library for dealing with X11) and freedesktop benefit both directly and indirectly from the work being done in the tiling window manager space. Similarly, Xmonad is a great boon to the Haskell community and cause (I suspect).

The other direction follows an essay I wrote here a few months ago about the importance of thinking about the capabilities of programing languages even if you're not a programmer because languages, like all sorts of highly technical concepts and tools create and constrain possibilities for all computer users, not just the people who ponder and use them. In the case of the tiling window manager, thinking about how the people who are writing computer programs is productive. In addition to the aforementioned thoughts about competition and open source.

So there we are. I'll be in touch.

comments powered by Disqus