Desktop Linux software
Unlike the previous two topics where it either works or fails, Linux software is something open to interpretation. This means even though the software provided might work exactly as advertised, you may find that its not meeting with your own personal expectations.
Speaking for myself, I have found that I am largely satisfied with the selection of applications available to me. However, someone coming from Windows might find that they are either running into limited compatibility problems or worse are not able to locate an open source alternative to their previously used legacy application.
The first instance is generally had with office suites such as MS Office migrating over to Open Office. While the latter can support the Microsoft file formats, more often than not formatting goes right out the window.
Then theres the issue of compatibility. Ever try to take something from MS Publisher over to Scribus? Now theres a fun way to spend an afternoon. So even though Scribus is an awesome desktop publishing application, the previous application has the end user locked into a file format that is making the migration extremely difficult at best.
As you can clearly see, a combination of circumstances in software can have drastically different results from one end user to another. Even though it was actually fairly painless to make the switch myself a few years ago, a brand new user committed to legacy applications will likely need to run extra software such as WINE or VirtualBox just to make the switch to Linux a permanent one.
Desktop Linux today and onto tomorrow
Finally, the last point of contention is that desktop Linux is forever playing catch up with the two popular desktop operating systems.
To those who feel this way, I would point out the following. First, Linux was the first OS to have provided working USB 3.0 support.
As projects such as rt2x00 and Intel work to put the final polish on 802.11n support for wireless, 802.11n was just to be ratified. It is expected that both chipset types will be working great with the next release of Ubuntu. As it is now, Intel already provides integrated 802.11n support while chipsets like rt2870 are ironing out the speed bugs to make sure it is running full steam ahead in the next Ubuntu release.
Then we have non-FoSS software. Proprietary applications are slowly making their way into the world of Linux as well, much to the dismay of some users. Skype just released its latest release designed to work specifically with PulseAudio (works great) while LightZone provides a Linux friendly option for those coming away from Adobe LightRoom. Regardless of what I might think, the fact is that we are seeing growth in proprietary software for this platform.
What about the computing needs of tomorrow? Will the Linux desktop collectively be able to keep pace with proprietary operating systems as things progress forward with each passing year? Yes, absolutely.
Both touch screen technology and even the brand new augmented reality concepts are already available to anyone who would like to try them. Seems obvious to me that the platform is well positioned to keep pace here.
The real future challenge for the Linux desktop is getting a handle on driver regression, providing a managed model that non-tech types can subscribe to that provides an out-of-the-box support system including remote desktop support.
And finally, finding a happy medium with open source applications and those of a proprietary nature. Hit those bases, I see every year forward as being year of the Linux desktop.
ALSO SEE: KDE's Expanded Desktop vs. Online Apps