Fast Times for Debian

Ubuntu, DCC push platform in unexpected directions.

At about this time last year, I was expecting interesting times ahead for Debian. The long-awaited Sarge release of the popular, noncommercial Linux distribution was right around the corner, and Bruce Perens pending UserLinux project promised to brighten Debians enterprise prospects by bringing the distribution a new face and community.

Sarge remained AWOL for much longer than Id expected, although it finally showed up, in good form, earlier this summer. UserLinux still exists, for the most part, as a scantily detailed Web site and a ghost-town mailing list.

Nonetheless, although things didnt turn out quite as Id expected, the year has still been an exciting one for Debian—a case of me being wrong about how Id be right.

Around the time Id expected to see Sarge and UserLinux hit the streets, we did indeed get a free, current and supported new release, based on Debians bleeding-edge unstable branch. We got Ubuntu.

The finishing touches for Ubuntu were applied not from within the canonical Debian release process but instead under the control of Canonical Ltd.

This smooth finish, along with the fact that Ubuntu has accrued momentum so quickly, has led some to wring their hands about a possible Debian schism.

The fact that Ubuntu is now backed by a nonprofit foundation endowed with $10 million from Canonical founder Mark Shuttleworth does buy the upstart a measure of independence, but I think Ubuntus dependence on Debian as a foundation is too strong to expect the new distribution to cut ties with the project.

A big advantage that Ubuntu (and other popular Debian-derived distributions such as Knoppix) enjoy over other varieties of Linux is the large number of software packages that are precompiled for Debian—an advantage these distributions would lose if they allowed themselves to stray too far.

The birth of a new, whimsically titled effort (its called Utnubu) to promote Debian-Ubuntu cooperation from within the Debian organization is definitely a step in the right direction.

Its clear that the children of Debian are stronger together than apart—which brings me to the Debian Core Consortium (reported on recently by eWEEK. coms Steven J. Vaughan-Nichols). The effort, which calls to mind the (as far as I can tell) stillborn Linux Core Consortium initiatives, seeks to rally prominent Debian-based distributions, including Linspire and Xandros, around a common Debian core in an attempt to bring Debian to enterprises.

Time will tell whether the DCC ends up going anywhere. After watching UnitedLinux fall apart and waiting all year for UserLinux to go somewhere, Im withholding judgment for now.

I will say, though, that the stated DCC aim of providing compatibility with Red Hat Inc.s Red Hat Enterprise Linux and the RedHat Package Manager package format seems to me like a waste of time. If organizations want RHEL, they can buy it or have an RHEL clone such as CentOS for free. Enterprise Debian efforts should focus on Debian and on Debians strengths—a major one of which is the distributions native packaging system.

Frankly, its not that difficult for ISVs to build native Debian packages, and if they cant offer a .deb version, you cant very well expect them to support the RHEL-packaged version of their application on Debian.

Instead of labeling themselves as a Red Hat also-ran, the organizations that are interested in pushing Debian into the enterprise would do better to concentrate on highlighting Debians strengths versus those of Red Hat.

For example, Debian does not suffer, as Red Hat does, from any confusing cleavage between the offerings free and enterprise versions.

Red Hats Fedora Core is a solid and popular distribution, and it is attractively free of per-system license costs, but you cannot employ the excellent Red Hat Network to administer it—even if youre prepared to pay for the service.

Debian backers can argue that the core of their system is noncommercial, like the Linux kernel itself, and that companies can choose the service and support that best suit them.

The success of Ubuntu has demonstrated that by combining the foundation of Debian with some money and focus, impressive strides can be made. However, the window of opportunity wont stay open this wide forever.

Although Red Hat has to date been slow to make Fedora truly a community-owned project in the way that Debian is, the Fedora foundation that Red Hat established in June could change this and further solidify the enterprises dominance in Linux.

Senior Analyst Jason Brooks can be reached at