Can Dell really offer Linux on its desktops? That's a question that a number of industry observers are asking. Most conclude that it can't -- but are they right?

It all stems from Dell's troubled financial state. Up until now, Dell's business model has happily chugged along, based on the direct sales model combined with highly efficient just-in-time inventory and build-to-order systems. The Texas company has managed to squeeze costs out of the system in order to retain the typically paper-thin margins of the commodity hardware business.

But the growth in commoditisation, a slowing domestic market plus additional efforts by key rivals to maximise market share have all combined to dump Dell's margins, net revenues and share price in the doldrums.

As a result, Dell's had to reach out to markets it might not ordinarily consider -- which is how it found itself asking Linux enthusiasts which distributions they'd like it to offer. More specifically, more and better support for open source software was at the top of the company's IdeaStorm site when it asked what people wanted from it; in particular, Ubuntu, Fedora and OpenSUSE.

The trouble is whether Dell, if it decides to tap this market, will end up losing money. The phenomenon of a company asking for its customers' views, giving it to them, and then finding out that a vocal but tiny minority are just that -- tiny -- is hardly unheard of. Many people make the point in the discussion on Dell's site that Vista is doing reasonably well in the consumer space, and that people are not necessarily crying out for an alternative to Microsoft.

Totally unscientifically, take my own experience as a computer-literate individual who supports to some degree or other about a dozen friends and relatives. All are in Dell's target market but none is clamouring for Linux. All simply want their computers to just work and don't care about the OS.

Rather, they care about the applications and about ease of use. And they certainly wouldn't want to start messing about with a conf text file. I don't imagine that my experience is anything out of the ordinary in this regard, any more than my "clients" are.

So will Dell's offer of Linux gain it extra sales? Possibly fewer than it thinks.

The second issue is whether Dell can possibly support the sheer volume of distributions on offer. The answer has to be no -- not on the kinds of margins that consumer market hardware vendors are subsisting on. In which case, Dell will almost certainly only offer one or two distros, thereby slicing a minority of the PC buying customer base even smaller.

There's huge religiosity attached to Linux and its various flavours, to the point where ten minutes trawling round forums finds acres of evidence that many, highly vocal Linux enthusiasts are blind to the way that most people use their PCs. It seems unlikely that they'll switch distros to buy a Dell.

Even bundling OpenOffice with its hardware -- probably the easiest of the steps the open source community is calling for -- could make a lot of sense from many users' perspectives. Yet Dell's accountants would be none too pleased, as it means that co-marketing income from vendors of its bundled software -- Windows, anti-virus software and the like -- will fall.

Instead, as my Infoworld colleague Neil McAllister points out, Dell needs to concentrate on the enterprise. Offer one flavour of Linux, support it to the hilt, and fill the world's call centres and public sector clerical workers' desks with a rock-solid but limited Linux-based application set. The volumes involved should help make up for the loss of money from Microsoft and the rest.

In asking the question about Linux, what Dell has done is laudable in many ways. It's raised the visibility of the desktop Linux debate -- even if it's shed precious little new light upon it. And it's got people talking about Dell in a positive manner, diverting attention from its limp balance sheet.

But by raising expectations that seem highly likely to be dashed, it may, in the long run, have done itself few favours.