If you look at the feature list for Debian 6, released on February 6, it's easy to be underwhelmed. This is especially true when measuring Debian against its offspring, like Ubuntu. Looks, however, can be deceiving — Debian is as relevant as it ever was, despite opinions to the contrary.
It's tempting to write Debian off as irrelevant, as my friend Steven J. Vaughan-Nichols suggests:
...as I look at the whole Debian situation, it seems to me that Ubuntu, with its leading the way from X Window to Wayland for Linux's foundation graphics and its new take, Unity, on the Linux desktop is now the ground-breaking Linux distribution that Debian once was. At the same time, Ubuntu is continuing to expand the Linux audience, while Debian continues to be a system that only hard-core Debian Linux fans will use.
Vaughan-Nichols concedes that "Debian is still important" he says "Debian is becoming increasingly irrelevant to the larger user community" being driven by Ubuntu and the other Debian derivatives. This is only true if you define relevance direct use — which is very much the wrong way to view Debian, and always has been.
Debian has never been a user-friendly distribution, or one that was really targeted at a mainstream audience. Debian 6.0 continues a long tradition of shipping a brand-new stable release that is already outdated, with little to appeal to new users. This was true more than 10 years ago when a Vancouver-based startup tried offering a prettified Debian with a simple GUI installer called Storm Linux. Stormix failed, but many others tried and finally Canonical has had a measure of success popularizing Debian with Ubuntu.
But it all revolves around Debian.
No Debian? No Ubuntu
You can't argue Debian's irrelevance using Ubuntu as an example. If you're using Ubuntu (or Linux Mint, or Mepis...), you're really using Debian with some enhancements. According to a presentation given recently by Debian Project Leader (DPL) Stefano Zacchiroli, only 7% of Ubuntu is directly derived from upstream projects, Canonical's projects, or other non-Debian sources. Of the rest, 74% of Ubuntu is rebuilt Debian packages, and 18% are patched and rebuilt Debian packages.
Debian is the raw material that is used to create Ubuntu, Linux Mint, and dozens of other Linux distributions that look more modern and easier to use. However, the reason that Ubuntu and the rest are able to ship fancified Linux distros that are easier to use is because they're able to start with Debian. If the Debian Project ceased tomorrow, it would be an enormous — possibly fatal — blow to its derivatives.
Debian doesn't get enough credit here, anyway. Yes, Ubuntu has appealed to a wider audience than Debian ever did — but it was Debian that inspired Mark Shuttleworth in the first place to create Ubuntu. As Brian Eno once said of The Velvet Underground's debut album, "Only five thousand people ever bought a Velvet Underground album, but every single one of them started a band." Likewise, Debian may enjoy a small percentage of the Linux market, but it's inspired one hell of a lot of people to start their own distribution.
It's not a zero-sum game. Ubuntu's popularity does not mean Debian is less popular. It means that Ubuntu has had the opportunity to gain new users. Ask 100 Ubuntu users whether they were using Debian before they switched to Ubuntu. Odds are, the majority were using Windows.
Debian has never depended on reaching the mainstream market, and never will. As long as Debian maintains its community — and I don't see much evidence that Debian is losing developers in droves to its derivatives.
It's not only the Debian derivatives that benefit from Debian. The Debian community finds and fixes a lot of bugs in packages that touch other Linux distros. Look through the changelogs in packages from Fedora or openSUSE (or Red Hat Enterprise Linux and SUSE Linux Enterprise) — you'll find Debian developers have touched a lot of packages and fixed a lot of bugs.
Community, not corporate
Debian's most important contribution to the Linux community may be simply that it's not controlled by a corporate entity. If 2010 taught us anything, it's that having a single corporate sponsor can lead to a lot of uncertainty at best and total disruption at worst. The OpenSolaris community (such that it was) got the worst of it, having their sponsor gobbled up and then being completely shunned by Oracle.
The Mandriva camp didn't fare much better, when the long-suffering company laid off most of its developers after years of financial troubles. The fork may do better, but the folks behind it have learned the value of independence.
Novell's state of uncertainty through 2010 wasn't as disruptive to openSUSE, but the acquisition had caused concern within the openSUSE community about the project's future and whether the new owner would be providing the same level of resources.
Debian, on the other hand, keeps (sometimes slowly) moving ahead. The direction of the project is set by the developers doing the work, and under a social contract that ensures that the project will prioritize the users and free software. Decisions are not made based on a need to rush products to market or by a "benevolent dictator for life." It's a trade-off, certainly. The corporate-sponsored distros tend to move faster, are responsive to (perceived) market demand, and have predictable roadmaps and release dates. All of that comes at the expense of ceding some measure of control to the company sponsoring the distribution.
Debian may not be the slickest and most polished Linux distribution, and the stable release may not be the most up-to-date — but it offers a number of perks that other Linux distributions do not. It supports more hardware architectures than Fedora or Ubuntu. This may not matter to a lot of users, but a number of users want to run Linux on PowerPC, PA-RISC, or MIPS-based machines.
Debian, irrelevant? No, not even to the users who've never heard of Debian. It's important technically and important to the larger contributor community that pushes Linux forward, even those distributions that don't derive directly from Debian. The stable releases from Debian are really just a byproduct of the project — it's the larger body of work that Debian produces that's important, and it's as relevant as ever.