After a slight delay, Fedora 20 Linux is now generally available with new features that affect desktop, server and cloud users alike. Fedora is a community Linux distribution that is sponsored by Enterprise Linux giant Red Hat.
Fedora 20 had originally been set for a Nov. 26 release, which was pushed back to ensure that it was fully stable. The first Fedora 20 beta debuted Nov. 12 and provided early adopters with a preview of what is now available. Fedora Project Leader Robyn Bergeron told eWEEK that it was a good feeling to get Fedora 20, code-named Heisenbug, out the door.
The Fedora Project is no stranger to rescheduled release dates. The Fedora 18 release, originally scheduled to debut in November 2012, was pushed out until January 2013.
"I'm just glad that we can let people spend their winter or holiday breaks without the pressures of an impending release this year," Bergeron said.
A key addition in the Fedora 20 release is full support for the ARM architecture as a target platform. ARM support now is completely integrated, and the Fedora Project had been working on it for several development cycles, Bergeron said.
"We've added significant quantities of build infrastructure for the ARM effort, and a lot of work was done to ensure that the process more or less appears to function in the same way to the packager or developer as it would for our other historical primary architectures," Bergeron said.
ARM is increasingly becoming an important architecture for hyper-scale computing deployments. Red Hat, the lead sponsor behind Fedora, has been working with ARM vendors, including Calxeda, to help further ARM support in recent years.
Cloud is another key area of focus for the Fedora 20 release, which now includes what the project refers to as "first-class cloud images." The new cloud images aren't so much about what the images themselves do but rather how the images get into the hands of the user, Bergeron said.
A key focus in Fedora 20 was with a change called "Visible Cloud," which is about ensuring that Fedora cloud images are presented as options in a really obvious way to end users that are visiting http://get.fedoraproject.org, Bergeron said.
"We've also made sure that the images are now getting testing done as part of our overall release process, and we've made further improvements to the build process to ensure that the cloud images are built with the same tool chain as the other products," Bergeron said. "It's really all about ensuring that our cloud offering is handled with the same care and love as the rest of our delivery options."
Fedora 20 includes a number of virtualization enhancements, as well, such as a new virtual machine (VM) snapshot user interface that is intended to make it easier to take and manage VM snapshots. There is also a new logical volume manager (LVM) thin-provisioning capability to improve logical storage management.
On the desktop side, Fedora 20 integrates the GNOME 3.10 desktop as one of its options as well as KDE 4.11.
The Fedora 20 release also has a strong focus on big data and specifically the open-source Apache Hadoop project. Fedora 20 has packaged Hadoop to run out-of-the-box in an easy-to-install and integrated manner.
With Fedora 20 now out the door, the project is looking forward to Fedora 21, though Bergeron said that, first and foremost, the project will be taking a bit of breathing time to go through the process of defining the future and how it is produced.
At the Fedora Flock conference in August, the Fedora community discussed new potential approaches for building a more agile Fedora Linux distribution.
"We're expecting that January will be spent wrapping up product requirements for what we expect to be three offerings, rather than 'one to rule them all'—separate desktop/workstation, server and cloud products," Bergeron said. "But that work around product definitions will need to be backed up by some serious rethinking of how we build, automate, test and utilize our infrastructure in the process of producing those offerings."
The Fedora Project recently celebrated its 10th anniversary, which has also triggered some additional introspection into how things were done historically and how they can be improved.
"We're honestly not unlike so many of the other businesses and IT shops out there; growing complexities, growing demands, not necessarily always accompanied by a threefold or more increase in the availability of people to support that work," Bergeron said. "So I expect that we're going to be doing a lot of work around laying the foundation to help us work smarter, not harder.
Sean Michael Kerner is a senior editor at eWEEK and InternetNews.com. Follow him on Twitter @TechJournalist.