THE GREAT LINUX debate still rumbles on, and looks set to continue for some time.

While market share has been creeping up on the desktop, it can't match the impact and use of Windows - yet Linux advocates will not tire of pushing their beloved Tux onto a public tired of Microsoft's efforts.

But the operating system just isn't ready for day-to-day use by non-experts. A mass of Linux fanboys will comment below and strenuously deny this, but those of us who are a little less biased, and have a chunk of common-sense, realise it simply isn't user-friendly enough to provoke a mass move from the much friendlier Microsoft Windows.

GUI Eye Candy
Like it or not, eye candy, special effects, translucent windows, etc, are inevitably going to attract 'normal' PC users.

The effect of these seemingly nice-to-haves should not be under-estimated in the battle to attract the masses to a new operating system.

Similarly to a eye candy, a swish, good-looking, highly usable and, most-importantly, intuitive user interface will bring users to an OS in droves.

Apple's OS X has begun to record advances in market share, and this has everything to do with an easy to use, attractive interface.

A number of Linux advocates will state that eye candy should be second to technology, form and function, but yet bemoan the lack of uptake of their chosen distribution.

Mark Shuttleworth of Ubuntu has recently come out and said similar things, asking the Linux development community to advance their interface to something along the lines of OS X.

Surely Shuttleworth is in a position to not only suggest this type of development, but to actively promote the development and push forward with his thoughts?

Too much hot air is wasted, when people in positions of power could be pr actically advancing development.

Again, people will flame nd comment that they are attempts at developing the GUI to have additional eye candy - but these effects are not standard in any large distribution, nor are they exactly bug-free or easy to install.

Unified Packaging
Seriously, installing applications on Linux is awful. Don't flame me or berate me with examples of how easy it is - it's not.

Half the time it's full of trials, tribulations, problems, and manual hacks - this is simply not good enough for a simple non-technical user.

If that's not bad enough, there is a variety of different packaging versions available - in which each major distribution seems to each adhere to its individuality. Learning one Linux distro does not necessarily mean you'll be able to work easily on another.

Ubuntu uses Debian packages, Suse and Redhat uses RPM, etc. Even the package installers work markedly differently.

Linux developers and packagers need to sort the situation out - embrace a standard packing movement and stick with it.

Each distribution should adhere to the standards set out, allowing Linux-developed software to be installed on every supported distribution without any problems.

When I say without any problems, I mean NONE. No manual hacks, no editing of obscure configuration files, no rebuilding of the kernel, no manually installing other dependent packages.

Run installer, install, installed. Like Windows. *Shudder*.

Support Major OEMs
It's pretty hard to support every device and manufacturer out there, but the Linux distributors could at least focus on supporting the major OEM product lines with installs that work correctly out of the box.

A recent experience with a mass produced, popular Sony Vaio left an installation with a wildly erratic touchpad, non-working sound, and a variety of other problems. Support requested on forums went either unanswered or gave vague answers which left the machine in a similar unresolved state.

You never know, you might see further OEM take-up of Tux.

Professional Support
Resolving issues on a per-issue basis, or yearly licence would provide Linux distributors with a source of income, and an increased user-base.

Some providers already provide a host of professional support services, but tailoring support to individual desktop users with a relatively cheap support cost can only benefit and help an increase user-numbers in the long term.

User Testing
I can't believe the majority of distributions go through any focus group sessions for feedback on their product.

I'm sure they'll say they do, but I can't believe the majority ever come-off favourably compared to OSX or Vista.

Get some marketing and technology-led focus groups together, formed of simple non-technical users, and try to evolve your product's design from a typical user's point of view - not a set of staunch 50 year-old Linux developers who've been programming away for Linus since Torvalds first whipped up an implementation of MINIX.

There's a long way to go before a single distribution is friendly and supported enough for the mass of typical non-techie users, but it isn't a far-fetched dream.

The Linux community should find common ground in the issues above, unite, and subsequently resolve the inherent, unaddressed problems with desktop Linux for the good of the masses - who are calling out for a usable, unbroken (read: Vista) operating system.

Stop splintering, fragmenting, and focusing on non-core issues, get to grips with the basics outlined above, and do us all a favour; replace the newest Windows with something other than another Microsoft iteration of Windows Me.

It just needs a collective use of the brilliant minds now at work behind the variety of distributions, applications, and core-OS, to come together and finally compete properly against the major commercial operating systems for the benefit of everyone. µ