A word of advice to FOSS geeks:

If you must recommend Ubuntu Linux to others, recommend nothing later than 10.04, the last LTS release.

10.10 saw a number of minor but irritating bugs creep in that show a significant shortage of testing and forethought. There were countless small things like context menus no longer working after returning from a suspended state or new window positioning that’s completely counter-intuitive. Some of them, like changing sides for window buttons or listing indecipherable package descriptions above package names in Update Manager, were deliberate (and conceivably, in some universe, necessary), but most of the changes were clearly mistakes. When these are combined with long-standing bugs (like Network Manager arbitrarily deciding to disable the Save button) and inconsistencies, they begin to weigh against Ubuntu’s many virtues.

In 11.04, Unity, combined with an increase in the number of stupid bugs (that spiffy state-of-the-machine motd message is FUBAR’ed now on console login) clearly indicates that Ubuntu is more interested in new and shiny than they are in quality. A quick scan of Launchpad (itself a new product designed to simplify bug maintenance and supplant the competition, but which has done neither) shows that there are, on average, 100 open bugs per project.

Ubuntu is slipping out of control. Canonical have stopped listening and – more importantly – working with the community. The number of defects is growing, but Canonical’s response is to make it harder for mere mortals to submit bugs. They seem to think that strong guidance is needed for their product to grow in new and interesting ways. Fair enough, but they’re confusing leadership with control. They’re simply imposing their views because they don’t value the discussion. They’re treating criticism as opposition and shutting themselves off from valid feedback.

Worse, they simply don’t have the number of skilled developers they need to achieve their goals. When I look at the bug queues on some packages, I shudder in sympathy with the poor souls who are expected to wrangle them. Canonical is clearly embarked on an impossible task, but nobody’s either got the guts or the vision to spell this out to Shuttleworth and co.

Getting buy-in and active participation from the community is a pain in the arse at the best of times, but the alternative is far worse. Heaven knows that the GNOME dev camp are… special, to be nice. But it’s clear that, given the choice between getting a partial but workable success through compromise or taking their ball and going home, Canonical has consistently chosen the latter.

This cannot end well. It will, however, end sooner than later.