Forgot your password?
typodupeerror
Cloud Linux

Old Arguments May Cost Linux the Desktop 591

Posted by CmdrTaco
from the water-under-the-bridge dept.
itwbennett writes "The old Linux arguments that pit one tool against another — Evolution vs. Thunderbird, LibreOffice vs. OpenOffice, and GNOME3 vs. Unity vs. KDE vs. everything else — may cost Linux its shot at the desktop, opines blogger Brian Proffitt. 'We can compare LibreOffice to OpenOffice.org to Office till the cows come home,' says Proffitt. 'But what happens when Google Docs gets truly robust enough for business and high-end document production? Or Prezi gets enough mindshare to start an upwards trajectory of user numbers?' It should be the case that increasing reliance on cloud software will make it easier for businesses to choose Linux, but for that to happen, Linux communities need to stop fighting the old fights, says Proffitt."
This discussion has been archived. No new comments can be posted.

Old Arguments May Cost Linux the Desktop

Comments Filter:
  • by Sycraft-fu (314770) on Wednesday August 10, 2011 @12:41PM (#37045690)

    For Linux to ever have a shot on the desktop, it would have to stop being Linux. Namely it would have to get some standards beyond the kernel. It would have to become a system where a lot more was standardized and you could rely on features, packages, UIs, etc being in all distributions. To desktop users, an OS isn't a kernel, it is a rich experience that comprises, well, everything you find on a Windows or MacOS disc. Until that happens, it'll never be an OS people want to use on the desktop because people don't want choice, they want consistency. That doesn't mean it couldn't still be flexible, just that it would have mandatory features and defaults.

    Along those lines it would have to do away with having source be something a user had any idea existed. No distributing programs as source, no recompiling the kernel to make something work, all binary all the time for users. Again, wouldn't mean it would have to get rid of source, just that the user experience couldn't include it. That would have to be all nice guided installers that are fast and easy.

    These things aren't important to servers, and completely unimportant to embedded devices, hence Linux has done well there. However they are what people need on the desktop.

    Notice that the end-user facing Linux that has had the most success by far is Android and it does precisely those things. It gives users and developers a consistent environment and set of tools that are guaranteed to be there, since they are a part of what Android is. It provides easy, binary-only installs for users so they just click on what they want and get it.

    That's what Linux as a whole would have to do to have a chance of capturing a significant share of the desktop market. So long as the answer to problems remains "Oh just use a different distro, that one doesn't have feature X and Y," or "Recompile your kernel with these options to make that work," it'll be the sort of thing that there isn't widespread interest in.

The meta-Turing test counts a thing as intelligent if it seeks to devise and apply Turing tests to objects of its own creation. -- Lew Mammel, Jr.

Working...