Five problems Linux still needs to overcome

Big business loves Linux for servers, and they seem to like it more than you might expect for the desktop, too. That said, enterprises still have some concerns about Linux. Here are the top five as picked by people who responded to The Linux Foundation's recent corporate and government end-user survey, "Linux Adoption Trends: A Survey of Enterprise End Users."

But before diving into these problems, I'd like to point out something. These are the opinions of business people who are mostly already Linux users. Questions like whether KDE or GNOME is the better desktop interface, or just how cool Ubuntu 10.10 is, matter a whole lot less to them than they do to Linux fans or programmers. Instead, they care about how they can use Linux to advance their work. They don't love Linux for its own sake. They love it because of what it can do for them. So let's get on with their list of concerns, in their order of importance.

1. Drivers

It's 2010, and almost any device you can name inside or outside a computer has a Linux driver -- yet 39.4% of business users still have concerns about Linux drivers. Sigh.

For years now, the Linux Driver Project (LDP), by Linux kernel developer and Novell engineer Greg Kroah-Hartman, has been creating Linux drivers for anything that any vendor brought to the project that needed one made for it. Kroah-Hartman and his crew of open-source developers charge nothing to create Linux hardware drivers. Despite that, a handful of companies still won't release Linux drivers. Other companies, like Wi-Fi chip vendor Broadcom, that have been slow to release Linux drivers has recently taken to making them. So what's the real problem?

I think there are several problems hiding under this issue. First, even now some hardware doesn't have any Linux drivers, or, more commonly, the drivers aren't that good. That's true of Windows as well, but people seem to give Windows a pass for this kind of thing.

On Windows and Macs, peripherals like scanners or all-in-one printer/fax/scanners also come with a dedicated software program to help users get the most out of them. Very few hardware vendors deliver that same kind of value-added extra to Linux users. While these programs aren't drivers, people tend to think of them as part of the driver package. The only solution here is for everyone to keep asking hardware manufacturers not just for Linux drivers, but for complete software packages.

2. Interoperability with other platforms

Plenty of progress has been being made here too, though much to some Linux fans' displeasure. They look at Novell and Microsoft's partnership on Active Directory (AD), virtual machines and, in particular, bringing .NET to Linux as Mono and they want to thrash Novell for betraying Linux to the Evil Empire.

Be that as it may, businesses want more interoperability between Linux and Windows. Like it or not, this trend is going to continue. It's corporations, not individuals, that buy Linux. I foresee other Linux companies besides Novell working with Microsoft as time goes on. It won't be pretty so long as Ballmer is in charge, but it will happen.

3. Talent to support Linux

There are a lot more Linux-qualified technicians and system and network administrators than there used to be, but there's still not enough of them. If you're in IT and you want a job, I have a couple of suggestions for you. One: if you're a programmer, get involved in a Linux-related open-source project to show that you have the chops to deliver quality code. Two: if you're interested in administration, start working on servers and networks anyway you can and go after a Red Hat Certified Engineer (RHCE) or other high-end Linux certification such as the Novell Certified Linux Engineer (Novell CLE).

4. Fragmentation

Bad news for Oracle with its "new" Linux: businesses don't want any more Linux choices. On the other hand, this also doesn't bode well for LibreOffice, the OpenOffice fork.

It's not that businesses don't appreciate choice between vendors -- they do. But there's not interested in choosing between half-a-dozen different Linuxes, two or three is much more their speed.

I don't see this as being a big concern. The last important Linux distribution to arise was Ubuntu back in 2004. I don't see any other major new Linux distributions arising in the future for either the desktop or the server. Mobile devices and tablets may be another matter. Android is doing well, but MeeGo may yet turn out to be an important portable Linux, and, who knows, perhaps another one will emerge or an Android or MeeGo variant will emerge.

5. Legal Issues

Yes, you'd be hard-pressed to find a pulse from the corpse of SCO, but Microsoft continuing to rattle the old patent saber has kept anti-Linux legal FUD alive. It also doesn't help any that Microsoft actually is suing Motorola over Android-related patent issues.

Here, all I can do is remind people that for all the threats, no one has successfully sued Linux for any intellectual property violations. The reverse is not true. The Software Freedom Law Center (SFLC) has socked many a proprietary company for putting Linux-related code into their programs and products.

Yes, when it comes to mobile devices, if it seems like everyone is suing everyone else -- well, that's because they are. This really doesn't have anything to do with Linux's legal integrity. It's all about companies using our fouled-up patent system to win cash or a business advantage. I don't see these fights spilling over into the more mature server and desktop markets. Thank God!

Even with all these concerns, one fact remains clear: big business is moving to Linux. CEOs, CIOs, and company boards all have worries, but none of them, either separately or together, is enough to stop them from moving to making Linux key to their business IT infrastructure.

FREE Computerworld Insider Guide: Five IT certifications that won’t break you
Join the discussion
Be the first to comment on this article. Our Commenting Policies