After spending a bit of time in Bugzilla it becomes very clear that a lot of time is spent watching the lizard eating bugs. And as the the queries grow larger and larger it becomes harder and harder, to get an overview of the state of the project (in this case: Fennec) or to drill down into specific details. So unless you are Gavin you probably need a bit of help like me.
What I came up with is a small app using a combination of Python and JavaScript that you can find here. A cron job uses a Python script to query Bugzilla for the data in order to make loading the page faster, with the downside – of course – of the bug query being up to 5 minutes old (in this case). The actual front-end consists of a few JavaScript functions, flot, jQuery and some markup.
The app itself consists of two parts:
- a grid based data overview, where you can drill into the details of every element and
- a graph based overview, with various presentations of the data
For now, I have chosen a variety of presentations of data for both the grid and graph overviews but I am very interested in hearing about other ways of looking at the data, other cross-sections, different graphs, better ways of interacting with the data, etc.
So please try it out and write a comment about what could be changed to make it better.
PS. The last two graphs “Graph Opened/Changed” and “Graph Bug Age” are both zoomable ie. you can mark a section of the graphs and the graph will display only that part.
Update: The source code is now available here. It still needs to be cleaned up, but is functional.
#1 by Gordon P. Hemsley - April 25th, 2009 at 04:21
Any chance you can package up the source and put it online so that I can download it and modify it for use with Bespin?
Thanks.
#2 by Dave Miller - April 25th, 2009 at 04:26
In the course of writing this, did you run across any APIs that you wish Bugzilla had that it doesn’t? We’ve been working on trying to spec out the APIs a little (particularly the xmlrpc stuff) and if it’s missing anything needed in the real world we’d love to know.
#3 by christian - April 27th, 2009 at 19:37
@Gordon P. Hemsley: I will clean up the code and put it up in a scm somewhere this week. I will update the post with a link, when I get it up.
@Dave Miller: I didn’t actually use any xmlrpc stuff for this, which make the data kind of limiting. I expect to start using the xmlrpc stuff very soon, to see if I can overcome some of those limits.