In this webinar, attendees learnt how projects can transition to Agile from non Agile environments. This session threw light on the unique set of challenges and trade-offs that involve cultural, technical and process issues while transitioning to Agile Development.

The PDF of the slides can be downloaded for further reference. Below is the video of the webinar for your viewing pleasure.

Top 10 Questions & Answers
You can download the rest of the Questions and Answers here.

Q: Is it mandatory to have a story wall the way it has been shown? Cant it be a huge Excel sheet in a common location?

Either a physical wall or a tool like Mingle is recommended. How do people move cards from one stage to another in Excel ? I guess it can be done in a somewhat cumbersome manner. The key point here is that the manager is not the sole custodian of the wall.

Q: It would be nice to see a practical implementation of Agile and its transition. Is there any story of a practically successful Agile implementation ?

There are some case studies here : http://www.thoughtworks.com/our-clients/case-studies/case-studies.html

Q: Can we use Agile development for fixed price projects? Most of the projects are approved by clients based on initial estimates like how many man hours will this take to implement. How do you estimate in Agile, since it is incremental?

Please see http://martinfowler.com/bliki/FixedPrice.html

Q: What happens if the bug fixes cannot be completed within a particular iteration? Would you carry over the defects into the next iteration or would the iteration be delayed?

Fixed iteration duration should not be viewed as a target date for completing functionality but rather as a regular heartbeat for taking stock of where the project is and doing course corrections.

Q: How does Agile fit into the Onshore-Offshore development model with the core stakeholders at onshore locations ?

Please see http://www.thoughtworks.com/pdfs/distributed_agile_faq.pdf

Q: What is the smallest cycle for an interation?

Typical SCRUM sprint is 4 weeks. Typical XP iteration is two weeks. Agile teams typically use books as guidelines and are empowered to deviate.

Q: How do we estimate in Agile software development ?

You could estimate in "real days", "ideal days" or "story points". Key is to not focus on precision. See http://www.youtube.com/watch?v=fb9Rzyi8b90 and  http://www.infoq.com/articles/agile-estimation-techniques

Q: How is Change management done in Agile ?

Changes are analyzed and accommodated pretty much on the fly. There are two keys to realizing this. One, if something new comes into a release plan, something old of similar size has to be de-prioritized.  Second, the premise of XP is that the cost of change curve is nearly flat. This is achieved by superior development practices like continuous intergration, fast and frequest builds, TDD (or BDD), domain driven design and continuous refactoring. Otherwise, change will result in big impact.

Q: In Agile, all the time they talk about functional testing, so what about system Integration testing ?

Test strategy is usually decided early on, by iteration one. In the following iterations, testers typically do a bit of exploratory unscripted manual testing and a lot of automated testing. See http://testing.thoughtworks.com/articles for more extensive advice.

Q. All these Iterations are overlap each other, so I feel sometimes we lose control of the project. Please advice how handle these situations.
Q: Testing/QA starts in the next iteration? Can it or should it be done in the same iteration ?
Q: As in each iteration we are testing for last development so when we are fixing bugs. I mean to say does Dev - 2 is independent to Dev-1 if so then when we are fixing bugs for dev-1 ?
Q: You mentioned that the testing completes in the next iteration. Does that mean that all the stories get completed always in the next iteration?
Q: How do you manage/incorporate defect fixing resulting from testing of the previous iteration ?

Iterations don't overlap. I think you mean that the testing of nth iteration overlaps with the development of (n+1)th iteration. Consider it as a series of pipelines (analysis, development, testing, customer sign-off) of variable capacity. Measure velocity as the throughput over all the pipelines. Adjust capacities of pipelines to avoid work piling up in the middle. Of course it works even better if you can manage to test in the same iteration (e.g. some manual exploratory testing at least) Bugs are typically picked up by development team as soon as they are reported. Ideally there should be few bugs at this stage if there has been good communication earlier.

Contact Us : mcs-online@thoughtworks.com
Phone : +91 9902043964

Master Class Series