I love Mozilla, as an overall project, as a source of fine projects and as an exemplar of the FOSS movement. But, seriously guys, unless the house was really on fire did you absolutely have to roll out an upgrade on December 19th? We spent an ugly last half of the week, and until midday today weren’t sure if people were going to be canceling portions of their XMas vacations to fix the issues it caused for us.
Where I work we are encouraged to push these sorts of things out to the world, to help others who might be having the same kind of problems and to help push the ball down the field. Here’s an example of that. Even though this isn’t open source, learning lessons I’ve picked up hanging around that movement helped us out ultimately. Rather than just posting about the problem, when I came in this morning my highest priority was to get a test case that would help us define the boundaries of the problem. By having that and getting it to the Active Widgets guys, they were able to use it to debug the problem. We got a fix better than our workaround and on the same day to boot. I like how on Beast, every time someone reports a problem they basically refuse to listen unless you include tests. “Submit a failing unit test case” is the standard reply to all such posts.
Ultimately, the issues are solved and all is well. It was touch and go for a while, but now it’s back on track. But, seriously, couldn’t the Firefox 18.104.22.168 / 22.214.171.124 upgrade have waited until January 2nd?
This post complies with my work blogging rules of the road.