Mad, Beautiful Ideas
Linux on the Desktop Requires Funding

A while back, the Linux Foundation was criticized pretty heavily earlier this month after this year’s Collaboration Summit for not supporting Linux on the Desktop. Admittedly, the core supporters of the Foundation: IBM, Red Hat, even HP and Dell, have more interest in Linux on the Server than on the desktop, after all the core of these companies business is on the Server.

However, Novell is a major contributor, and they have put an enormous amount of money in GNOME on the desktop. They have a Desktop Standards which has been working hard to standardize behavior of the Linux desktop such that application interoperability, even between KDE and GNOME. These days, I can run a handful of KDE apps that lack good GNOME equivalents (like Scribus) and still have copy and paste and session support. The work of the Foundation has helped Desktop Linux immensely over the years.

And now they’ve turned their eyes towards Device Drivers. Historically, this has been a problem with Linux, because devices can take a lot of work to get working, and many companies simply aren’t willing to take the time to support their device on Linux, when the driver interface potentially changing, and the potential legal issues involved in linking binary code against the GPL’d kernel.

So, the Foundation is suggesting that companies release their drivers as Open Source, which will enable the drivers to continue to work as the kernel evolves, as well as potentially opening the door to improvements to those drivers. Traditionally, this has been avoided by most companies, because they were attempting to protect their intellectual property wrapped up in their designs, and by releasing the drivers, they could be giving up control of that IP. Eric Raymond wrote years ago about the fallaciousness of this argument, but it wasn’t until recently that companies like AMD releasing documentation for ATI cards, and Intel doing the same for their GPUs.

Frankly, I’m glad about that. Ever since we bought Catherine’s Dell Inspiron 1420N, which Dell would only sell us with an Intel-based chipset, I’ve been hoping to see improvements to those drivers. Compiz doesn’t work, and any games depending on OpenGL, are basically unplayable. Admittedly, the laptop is primarily a work machine, but Catherine has been disappointed with that particular failing of the drivers. I’m sure it will improve, though if I could find a place to donate some funds to the project, I likely would.

Ultimately, this is where Linux on the Desktop is going to continue to struggle. Most Linux users simply aren’t willing to pay for software or drivers. Admittedly, I’m always looking for the FLOSS alternative, and I’ve very rarely offered financial support for a project that I’ve been using. There has only been one place that I’ve routinely supported Linux software, and that is games. If nothing else, if we can get more Games on Linux, I’m sure it will become easier to attract casual users to Linux, which will place further pressure upon the device manufactures to support the platform. In addition, with more users, we’re more likely to see other software on the platform as well.

However, a lot of companies aren’t willing to risk releasing their products on what they feel is an untested platform. Smaller companies, like Introversion and Basilisk Games are typically more willing, as they need to reach as broad an audience as possible.

Now, Runesoft, a game company with several Linux releases under it’s belt, uses an interesting mechanism to fund their ports. They require that a certain number of preorders exist for a game before they do the port. The current game is jack Keane, which they’ve already committed to the Mac. Considering Ryan Gordon’s comments in his FLOSS Weekly Interview claiming that, for UT2004 at least, there were more Linux than Mac clients, this is an interesting decision, but the difficulty in judging the Linux install-base (due to lack of a ‘sales’ model) makes the decision understandable.

This is a great model. Runesoft can guarantee that they’ll have their time paid for before they commit to the port, and we can help finance this and future projects. If we want Linux to survive on the Desktop, we as users need to be willing to financially support the projects. Jeff Atwood pushes this same position. It just makes sense. Software takes time to create. Yes, most Open Source software is created to solve an itch of the developer, but many successful projects reach a point where the developer begins adding features for others, and not just themselves. Ultimately, at this point either the projects suffers because the developer can’t afford to spend the time necessary to add the features that everyone wants.

Support Software. Support your Platform. I pay for commercial Linux software that I plan to use. I’m planning to budget a bit of money to donate to open source projects each month, so that they can continue to thrive. Without money behind the platform, Linux can’t advance. And if we want Linux to advance on the desktop, we, as desktop users, need to be willing to put forward some cash.