Matt O’ Rama

Starting back on Scrum

Right around the time of our first deadline, we decided to stop doing iterations and start just banging on whatever bugs we found. After we didn’t go live we never went back on the iterations, which was a mistake. Today we had our first sprint planning in a few months. Unfortunately it’ll only be a week sprint since most of the team is going out to San Francisco next week for a vendor conference. Of course it’s hard to feel too bad since we’ll be in San Francisco, my favorite city, but it did make for a somewhat half-hearted planning meeting. We used to use Scrumworks but our licenses are gone and we never really liked it anyway so this sprint is going to be tracked using a spreadsheet and a whiteboard. After we get back I’m planning on using VersionOne to track things. I’ve been using it on a trial basis and I like it a lot, I just need to get the team using it to see what they think.

For this half sprint we decided to take a couple of core areas we’re still struggling with bugs in and hit those as hard as we can. Another teammate and I will be hitting the invoice bugs that somehow are still popping up with alarming regularity. The others will be hitting one of the other statements and a few of the high-priority financial issues. I’d like to see a good number of these things knocked out before we leave so the testing team doesn’t have to deal with them. Testing a billing application is ridiculously time consuming and any one bug can cause hours and hours of testing to be wasted if it affects a balance somewhere unanticipated.

Before the planning meeting we got together and did a short session of planning poker as well, our first. It went pretty well. Once I explained how it was supposed to go, we moved through a bunch of issues really quickly. One problem we had was that the stuff we’re working on mostly now are bugs so it’s hard for everybody to have an idea of how hard something will be to fix. I think we’ll try again but the planning poker will really be helpful once we’re doing new development work again.

First Java servlet

As I mentioned on twitter today, I wrote my first Java servlet today. I had written part of a webapp using the Tapestry framework a few months ago but stuck and haven’t gone back to it. This was a basic servlet to take a chunk of XML and send it via JMS to a queue for testing a new JMS integration architecture we’re working on for work. It seems like everything in Java has some weird name that always makes me think it’ll be harder and more complicated than it is. I’m used to Perl where you just drop a file in Apache and use CGI.pm to pull out the parameters. Turns out writing a servlet is almost that easy, although the setup to get all the files and classes deployed in Tomcat is much more complicated. Luckily I’ve done enough with Ant to hack a deployment script together and we have an Ant expert I can ask for anything hard. As usual, what I did was find a couple of examples and hack the heck out of them until I understood them and they did what I wanted. I did go through part of the first chapter on servlets from my Head First Servlets & JSP book that I had barely cracked the cover of before, so that’s more research than I usually do ahead of time. Overall this basic servlet was pretty easy, although I know I just scratched the surface of the power and complexity available. Tomorrow I start working on figuring out how to send JMS messages. Luckily my team is incredibly good and I have good examples to hack on.

Bringing people around to Agile

This is an email I wrote my department yesterday, talking about keeping on with Agile. The team has decided we want to keep going with the successes we’ve had with Agile but we obviously have to convince the people above us. Since I’m not a manager I have to put out my ideas and work more behind the scenes to make changes. I don’t know how useful this might be to anybody but if you’re looking at introducing Agile to your work, just think about putting your ideas out there to stew around in people’s heads.

Since the email is pretty long, click Read More to read it if you’re interested.
More »