What's happening at Ubuntu: from X.org updates to upstart [LWN.net]

What's happening at Ubuntu: from X.org updates to upstart

Posted Sep 1, 2006 2:37 UTC (Fri) by sbergman27 (subscriber, #10767)
Parent article: What's happening at Ubuntu: from X.org updates to upstart

"""If there is a silver lining to the error, it is that it happened during the one week in six months when we have the core distribution development team together in one place. This gave us the opportunity not just to analyse and fix the issue, and to talk about the sequence of events that led to the problem, but also to discuss the processes we must improve to further reduce the likelihood of a repeat."""

I'd certainly love to have seen this! A room full of developers seriously discussing, after 2 years of releasing patches, whether it might not be a good idea to do some sort of testing on those patches before releasing them to a user-base of newbies.

It's pretty obvious that no serious testing was done on the xorg patches before they were loosed upon the world.

I like and use Ubuntu. But for shame, Ubuntu devs! For shame!

(Log in to post comments)

What's happening at Ubuntu: from X.org updates to upstart

Posted Sep 3, 2006 13:42 UTC (Sun) by cjwatson (subscriber, #7322) [Link]

Obviously, we deserve and are content to receive criticism that this upgrade broke a lot of people's systems. All the same, while it's true that the testing of the X.org patches in question was clearly not *sufficient*, I think you're maligning the Ubuntu team somewhat by skipping over the fact that it only failed on certain classes of machines. The developer responsible for the upload tested it on a number of his own systems with substantially different hardware configurations before release, and had a number of interested users try out the changes as well, and no problems were found in that process. The problem was that it wasn't tried out on a wide enough range of hardware, and that we didn't flag the patch as potentially risky during approval for dapper-updates, not that it was uploaded blindly with no testing.

Much of the discussion within the development team was more about best practices for dealing with emergencies once they arise, rather than "huh, you think we should do some testing, then?". Of course, there are changes that we can and will make to the testing process for stable release updates to make this sort of thing much less likely in the future; there will be details of that in the post-mortem report.