Scrum defined in under 10 minutes

Hamid Shojaee from www.axosoft.com put together a great video to introduce the core concepts of scrum and Agile practices called "Scrum in under 10 minutes"

[kml_flashembed fversion="8.0.0" movie="http://www.youtube.com/v/Q5k7a9YEoUI&hl=en&fs=1" targetclass="flashmovie" publishmethod="static" width="400" height="300" play="true" loop="false" quality="best" allowfullscreen="true" allowscriptaccess="always"]

Get Adobe Flash player

[/kml_flashembed]

I really like what he said particularly the focus on the importance and simplicity of the burndown chart which helps to :

  • Visually show the hours remaining, and help anyone see "are we there yet?"
  • Allows decision-makers to make adjustments if needed

A simple burndown chart quickly shows the impact of some of the most common efficiency killers or blockers

  1.    late-breaking requirements changes
  2.    waiting for business decision
  3.    waiting for tech. partner (ssl keys, ops. vendor, etc)

He mentioned the standup, but didn't cover the 3 questions (What did you do?, What are you going to do? What do you Need in order to accomplish it?), or the idea that only people that have work can speak (Chickens vs. pigs)

I also liked his take on bugs, that you should plan a sprint to tackle them, and any bugs that come up during a sprint should be tackled immediately.

That's actually something that a lot of teams struggle with, "How do you plan for and estimate the time necessary to fix bugs?"

I always say Bugs Can't be estimated. While you may be able to plan how much time you will allocate to working on bugs, you can't really estimate how long it will take to fix bugs unless you have already looked at them, and have figured out the problem. From my point of view its the "looking at it and figuring out what's wrong", that takes the longest, and is the part that needs to be accounted for.

How do you plan for bugs?
Do you let bugs accumulate and then have a final sprint before a Release where you tackle all of the bugs? (How would you know that you can fix all the bugs in time for the release?)

Do you tackle all High-Priority bugs as soon as they come up, and leave the medium and small priority bugs for later?

Do you have a 'zero-bug' policy, where you fix any bugs that are opened?

UPDATE: I've expanded on this discussion a little more in a new post Bugs Can't be estimated, and after having some discussions around how best to plan for bugs, I was asked to do a last minute Security Analysis, and I wrote a post about How to avoid the last minute security review, and I think that the planning part can actually apply to how you manage a big list of bugs as well.