LWN Weekly Edition
Front pageSecurity
Kernel development
Distributions
Development
Announcements
->One big page
This page
Previous weekFollowing week
| |||||||||||||
DistributionsMageia nears its second releaseThe Mageia distribution was formed in September 2010 when a group of former Mandriva employees combined forces with motivated community members to fork the software and form a non-profit association to manage the distribution and its assets. The project founders cited a lack of faith in the faltering Mandriva S.A. company, which had recently laid off much off the development team. As we discussed last week, the state of Mandriva is still unclear, while Mageia made its first stable release in June 2011, and is now nearing its second. SoftwareGauging the health of an open source project is a tricky undertaking. Oliver Burger recently announced that the release schedule for Mageia 2, would be pushed back, with a final release now slated for May 15th. To some that would indicate cause for concern, but to be fair even large and well-established projects occasionally slip on their release schedule, as happens frequently with Fedora. In the announcement, the revised Mageia 2 schedule was described as the result of outstanding bugs resulting from migration to systemd for initialization and dracut for initial ramdisk creation, as well as usability changes to the DrakX installer. Moving to systemd and dracut will bring Mageia more in line with Red Hat and Fedora, which should simplify packaging and development for the smaller project in the long run. The very first Mandrake release was based on Red Hat, back in 1998, and the distribution remains roughly compatible in other respects (including use of RPM as the package format). On the packaging front, though, Mageia also chose to embrace RPM 4.x, while Mandriva took to the rival RPM 5. While the original Mandrake shipped with KDE in place of GNOME as the desktop environment, a distinction still maintained in the latest Mandriva release, Mageia offers both. The latest development build for Mageia 2 (beta 3, released April 18), is provided as a live CD in both GNOME and KDE versions, for both 32-bit and 64-bit architectures, and in five different language packs covering distinct regions of the globe. If that is not enough options, there are also three DVD ISO images (each of which includes KDE, GNOME, and LXDE): one for 32-bit systems, one for 64-bit, and one with both. ![[Mageia 2 control center]](/images/2012/mageia-controlcenter-sm.png)
Beta 3 ships with kernel 3.3.1, KDE 4.8.2, and GNOME 3.2, which puts it slightly behind in its GNOME packages. There was talk of migrating to Grub 2 and btrfs as well, but both moves appear to have been pushed back until after Mageia 2. There are also several outstanding issues regarding video driver support (detailed in the Mageia 2 errata). I tested the GNOME version of beta 3 (since I am more familiar with GNOME) on a 32-bit system. Although I was bitten by the video driver bug described in the errata, the desktop experience is snappy, and Mageia provides a good, well-rounded GNOME 3 experience. The live CD version includes most of the same standbys you will find in any contemporary distribution: Firefox 10 (the extended support release), LibreOffice 3.5.2.2, Evolution, Empathy, and so on. Mageia does differentiate itself by offering its own system and configuration tools, derived from Mandriva. They do a better job than most distributions of presenting configuration options in human-readable form (e.g., "Set up a new network interface" rather than "Network Manager"), although die-hards might object to the pervasive use of the word "administrator" in place of "root." The Rpmdrake software-update tool is also unique to the Mandriva heritage; it is a tree-view driven installer and update manager akin to most others, but it also separates applications into multiple levels of categorization and allows you to filter updates based on whether they are security fixes, bugfixes, or general updates, and to distinguish between GUI applications, non-GUI applications, and metapackages. ![[Mageia 2 Rpmdrake]](/images/2012/mageia-rpmdrake-sm.png)
Governance and communityPerhaps a better way to assess how a new distribution is getting along is to look at the vitality of its community and contributor pool. On this front, Mageia appears to be in good shape. The distribution hosts both mailing lists and web discussion forums for users, as well as the typical IRC option and a number of developer-team-specific mailing lists. Both the forum and community list are active — there are about 1500 registered forums users, and the list tends to generate on-topic threads that involve multiple active parties. That is in contrast with the scenario we have all seen from time to time when one or two people tend to dominate every discussion. But although the list traffic is healthy, one thing does stand out about it. Even on the mageia-discuss list, which is "about Mageia general discussions: future plans, community, association, ...," development discussions account for the bulk of the traffic. Installer issues, casual bug reports, build notes, and debates over package selection are commonplace. What this suggests is that Mageia's user community is still largely composed of developers and packagers, rather than end-users. This is a good indicator that the contributor pool is motivated and engaged, but it also indicates that the distribution has yet to break through to the mainstream public. The mageia-dev list is even higher traffic, particularly package maintainers push updates in preparation for the Mageia 2 release. The project's organization is interesting; it has an elected board that takes responsibility for the governance of the non-profit foundation and overall mission-and-direction questions, and a separate community council in charge of planning, coordination, and production. Board members are elected to one-year terms, but the seats on the community council are filled by three representatives from each team (development, packaging, QA, documentation, translation, infrastructure, etc). There are currently 13 members. That governance model is in the minority among large, established distributions. OpenSUSE, Fedora, and Ubuntu all include corporate appointees in their governing bodies. Debian does not, naturally, as it does not have a corporate sponsor, but its technical committee is charged primarily with technical policy duties, and can only overrule a developer decision by a 75% majority vote. Mageia's board and council appear to function rather casually, as befits a young project. Dispute resolution and major technical crises have yet to strike the distribution in a big way (although there were decisions to be made leading up to the Mageia 1 release) — but if history teaches anything about open source projects, the disputes will come. Mageia's founders wanted to insulate the distribution against business concerns that would endanger the future of the software product itself — with good reason, considering the recent financial woes on Mandriva's side. But a small distribution needs more than legal protection to make itself successful in the marketplace. As a result, Mageia's recent moves to bring the distribution back in line with the system stack used by its RPM-based contemporaries are just as important as its community structure; they make development more sustainable by relying on healthy upstream projects, rather than expending limited resources on roll-your-own components.
Brief items Distribution quotes of the week
Maybe it’s too much to ask that we look over every little decision we make
where there’s disagreement and attempt to find every last bit of common
ground that we can (There were certainly times when it seemed to take
forever to make a decision) but what about decisions that are close votes?
What about decisions that have days-long threads as part of their
backstory? In these cases, consider the proposal that the majority agrees
on to be a strawman. A starting point from which to start chipping away to
see what changes can be made that are still acceptable to the majority
while addressing many of the issues that the minority has. Remember that
the goal is to craft a compromise that addresses as many concerns as
possible.
-- Toshio Kuratomi
12.04 being an LTS we’ve been minding our P’s and Q’s, but many of our
quality-oriented practices from 12.04 LTS will continue into Q
territory. We’ll keep the platform usable throughout the cycle, because
that helped hugely to encourage daily use of the release, which in turn
gives us much better feedback on questions of quality. And we’ll ratchet up
the continuous integration, smoke testing and automated benchmarking of the
release, since we can do it all in the cloud. We have, so to speak, stacks
and stacks of cloud to use. So quality is quotidian rather than
quarterly. And it is both qualitative and quantitative, with user research
and testing continuing to shape our design decisions. The effort we put
into polishing Unity and the rest of the platform in 12.04 seem to have
paid off handsomely, with many quondam quarrelsome suddenly quiescent in
the face of a surge in support for the work.
-- Mark Shuttleworth
Fedora architecture promotion requirements adoptedAfter extensive discussion, the Fedora Project has adopted a set of requirements that must be met before a secondary architecture can be promoted to primary status. More discussion has inevitably followed; the most controversial item seems to be the requirement that the architecture must support installation with Anaconda.
Red Hat Announces the Beta Release of RHEL 6.3Red Hat has released a beta of Red Hat Enterprise Linux 6.3. For details see the technical notes on the changes and the release notes.
Scientific Linux 5.8Scientific Linux 5.8 has been released. "We want to thank all those who have contributed helping us build and test this release. Finding bugs early has helped make this an excellent release." As usual the release notes contain additional details.
The new Minty Swift Linux is now availableThe Swift Linux distribution aims for a lightweight and user-friendly experience with good support for older hardware. Version 0.2.0 uses the Linux Mint Debian Edition (LMDE) as a base and comes in several editions: Regular Swift, Diet Swift, Taylor Swift, Minnesota Swift, Chicago Swift, and Silicon Valley Swift.
Distribution News Fedora Fedora 18 Release name voting and pollThe Fedora Project has opened the voting to decide the code name for Fedora 18. There is also a poll on whether releases should have code names at all. "This cycle, the Board is also asking contributors to let us know if we should continue to have release names for future Fedora releases. Even though the interface is the same, this portion is intended to be a poll rather than a straight up vote. The Fedora Board will look at the answers to determine if enough contributors value continuing to create release names to make it worthwhile in the future. If it does seem desirable, the Board will likely look into forming a working group to come up with a new method for creating release names for future releases."
Newsletters and articles of interest Distribution newsletters
People behind Debian: Samuel ThibaultRaphaël Hertzog interviews Samuel Thibault about his work on accessibility and the Hurd. "The Debian GNU/Hurd port can almost completely be installed from the official mirrors, using the standard Debian Installer. Some patches need some polishing, but others are just waiting for being uploaded… Debian GNU/Hurd can start a graphical desktop and run office tools such as gnumeric, as well as the iceweasel graphical web browser, KDE applications thanks to Pino Toscano’s care, and GNOME application thanks to Emilio Pozuelo Monfort’s care. Of course, general textmode hacking with gcc/make/gdb/etc. just works smoothly. Thanks to recent work on ghc and ada by Svante Signell, the archive coverage has passed 76%."
Shuttleworth: Quality has a new nameHere's the traditional naming post from Mark Shuttleworth; the next Ubuntu release will be code-named Quantal Quetzal. Various other bits of information can be found there as well: "Rumours and allegations of a move from Upstart to SystemD are unfounded: Upstart has a huge battery of tests, the competition has virtually none. Upstart knows everything it wants to be, the competition wants to be everything. Quality comes from focus and clarity of purpose, it comes from careful design and rigorous practices. After a review by the Ubuntu Foundations team our course is clear: we’re committed to Upstart, it’s the better choice for a modern init, innit."Those who are interested can also read Lennart Poettering's response.
Page editor: Rebecca Sobol |
|||||||||||||