"Wake up call" for openSUSE as 12.2 is delayed
The release manager of the openSUSE project, Stephan Kulow, has written to the project's development mailing list, explaining the recent delays in the distribution's release schedule. "It's time we realize delaying milestones is not a solution. Instead, let's use the delay of 12.2 as a reason to challenge our current development model and look at new ways" says Kulow. It is not clear yet if openSUSE will move to a rolling distribution model, slow down its releases or entirely abandon scheduled releases.
According to openSUSE Community Manager Jos Poortvliet, the release candidate for 12.2 is unlikely to go out today as previously scheduled and the planned release date of openSUSE 12.2, 11 July, is also unlikely to be kept to. Poortvliet says that this and the email by Kulow "serves as a wake up call for openSUSE". The distribution's problem seems to be that it has attracted many more contributors over time and that its development processes have not scaled. An increasing number of contributors are submitting more and more smaller patches but according to the release manager's assessment, what the distribution really needs is "a crew of highly experienced core hackers to fix issues all over within a reasonable time frame."
On top of these growing pains, there are other issues that afflict the openSUSE release process. The build service needed to create images is breaking down frequently, causing more delays. All of this has led to every milestone of openSUSE 12.2 having been delayed so far. Poortvliet and Kulow seem to agree that these delays are not temporary and are symptomatic of deeper problems within the project that need to be addressed.
To solve these issues, several changes to the openSUSE release process have been proposed. One possible direction would be to turn openSUSE into a genuine rolling release distribution based on the Tumbleweed repository and based on SUSE Linux Enterprise (SLE). The problem with this plan, according to Kulow, is that Tumbleweed is not designed to be rolling forever and needs to re-base on a new release periodically. This would move much of the "major plumbing work" to SUSE when working on SLE releases. Another proposal sees the project scale down to one major release a year, a move that would give the release team more time to stabilise the source tree before a release. Following Debian's lead and giving up on scheduled releases completely would be another option but this could lead to long feature freezes and very out-of-date packages.
Poortvliet suggests that the most likely scenario for a new openSUSE release strategy is a combination of these ideas. Whatever is decided in the end, the openSUSE leadership seems convinced that a major change is needed.
- openSUSE Summit: Call for Papers announced, a report from The H.