One of the major debating points about Debian has been the delay in releasing version 4.0 or etch. The release was targeted for December last year. There was unhappiness among some developers last year when Towns launched a project to pay the release managers in a bid to keep to the deadline. There are unconfirmed reports that developers who were unhappy with the payment plan deliberately slowed down their work so that the release would be delayed - and they could say that the plan had not helped.

One has to ask why a deadline for any release came about at all. Debian had, prior to etch, been a project that released a stable version when it was ready. After the last release, sarge, took nearly three years, an 18-month release cycle was planned for etch. To a large extent, the release cycle was influenced by the growth of the Ubuntu project.

Ubuntu makes a release every six months to coincide with the release cycle of the Gnome desktop environment. Ubuntu was initially based on the unstable branch of Debian but now has diverged quite a bit. While Ubuntu can be downloaded and obtained in CD format free, it is a commercial project; all the developers are paid by Canonical, a company owned by South African Mark Shuttleworth.

Despite not being commercial itself, Debian has come under some kind of artificial pressure to meet some kind of a deadline as it is felt (and practically every person who talks about it says the same thing) that the project is "falling behind". If one were to ask "falling behind what?" then the only answer possible is: "falling behind commercial distributions."

Thus, all candidates in the running for the position of leader have listed what they see as the issues which prevent Debian from adhering to some kind of timely release schedule. One of the issues that figures prominently is communication to reduce seemingly needless debate and flamewars. Another is the scrapping between sub-groups.

If etch is released in the next month or so as seems likely, it would have taken about 22 months, which is not bad at all considering the nearly 34 months that sarge took to emerge. It must be borne in mind that this is a team of 1000+ developers spread over 43 countries at work. And they always put out a high quality distribution which is ported to many more architecures than any other distribution. In 2007, one would be hard put to find a Linux distribution which can be installed over the internet after booting from a couple of floppies - but Debian meets that requirement too.

So, while there are areas which need a bit of a kick in the pants - which, perhaps, as Murdock outlines, a benevolent dictator may be better placed to do than a leader who functions as a true democrat - there are many areas of Debian which should not change else the project would lose its character. The temptation to chase behind commercial shadows is indeed great but in the end, Debian is unique among Linux distributions and too much of change would hurt the project rather than help it progress.

{moscomment}

Tags See All Tags

Linux  Open Source  Sam Varghese