While announcing the name of Ubuntu's next release for October, 2007, Ubuntu founder Mark Shuttleworth spoke of a new version of the Ubuntu distribution to be released alongside Gutsy Gibbon. This release, he says, will not include any firmware, drivers, applications, and even sounds and imagery which are not completely open-source and include full rights of modification and redistribution. This ultra-orthodox version of open-source sounds very attractive to open-source supporters, but is it practical?


In any Linux distribution, a large majority of the source meets the requirements mentioned. However, there are components to many distributions which are proprietary and thus not usable in a so-called ultra-free distribution. If a component is of enough importance, it could bring the usability of such a distribution into question.


For many proprietary components to a Linux distro, there are alternatives which are completely open-source and meet this ultra-free definition. However, this is not always the case. For example, the binary ATI display drivers for Linux do not include source and would thus be disqualified from inclusion in an ultra-free Ubuntu release. Since ATI has yet to make any suggestion of the release of the source to their display drivers, alternatives would have to be considered.


Unfortunately, the alternatives are sometimes simply not adequate—or do not exist at all. Especially for new hardware, alternatives may have not been developed enough for full release. Features found in their proprietary counterparts may be basic, unstable, or nonexistent in the open-source alternatives. New hardware which is immediately supported in the manufacturer's closed-source software may take many months to be supported by open-source equivalents. For some applications, these limitations may not matter; not all machines need wifi support or 3D-acceleration. However, for other applications, it might make the entire system unusable.


While it is admirable for the Ubuntu team to work to free Linux from proprietary code, it might simply be an impossible task without much more support from manufacturers and developers of closed-source software. As long as users have to rely on proprietary software to easily and reliably use new, unusual, or otherwise unsupported hardware, running such a distribution in a practical environment may have to wait.