I less than 3 Pychart (or how I turned tracking 10 bugs into a programming task)

We are getting closer to release! Beta2 freeze is tomorrow. Quality in 12.04 is looking very good today. However, we still see hundreds of bugs get fixed across desktop and server between now and April 26th. In the past, I've found that in the flury of activity it's easy to lose track of the most important bugs in all that noise, and then some scrambling ensues.

To counteract this, at least for myself, I had a couple of calls, with Jason Warner (Desktop Engineering Manager), Robbie Williamson (Server Engineering Manager), and Steve Langasek (Foundations Engineering Manager). We talked about what bugs we had (that we know about now) that would actually keep us from releasing as scheduled. We have a term called "release blocking bug", but in point of fact, almost none of them would actually keep us from releasing. The kinds of bugs that would truly make us slip a ubuntu release are ones that cause problems with existing OSs in multi-boot situations, serious bugs in the installer, serious bugs in update manager, bugs that result in a loss of networking, etc... Bugs that can reasonably fixed in an update do not block the release.

We decided that the best way to keep track of the very few bugs like this is to continue to track them as normal, but to set their importance as critical.

There is another set of bugs that I also ask the team to focus on. This set is more aspirational. I want us to fix all of the upgrade bugs that we find from automated testing, or at least all of the High and Critical importance ones. I would sincerely love to see every upgrade go smoothly for all of the millions of people who will be upgrading to Precise.

So, when I am going to start talking about pychart? Right now, in fact! Keeping tabs of bugs is boring, so must be automated, and I love automating things with Python. So, I wrote a program that scrapes the data from those 2 pages, store the info in a sqlite database, and generate a line graph each time I run it.

You can see all the code here if you want, but I doubt you do, it's pretty hacky. But, it was fun to bring together the ecellent json, HTMLParser, sqlite3, and pychart libraries.

Here's the pychart money shot:

        xaxis= axis.X(label=_("Date"), tic_interval=1,format = format_date)
       yaxis = axis.Y(tic_interval = 2, label="Open Bugs")
       ar = area.T(x_axis=xaxis, y_axis=yaxis, y_range=(0,None))
       plot = line_plot.T(label="upgrade", data=graph_list, ycol=1, tick_mark=tick_mark.star)
       plot2 = line_plot.T(label="blockers", data=graph_list, ycol=2, tick_mark=tick_mark.square)
       ar.add_plot(plot, plot2)

       can = canvas.init("/home/rick/Documents/bugs.png","png")
       print ar.draw(can)
       self.ui.image1.set_from_file("/home/rick/Documents/bugs.png")
def format_date(ordinal):
   d = datetime.date.fromordinal(int(ordinal))
   return "/a60{}" + d.strftime("%b %d, %y")