ICANDO Bulk Chemical Ordering
The ICANDO Chemicals system is a replacement for the diverse existing systems throughout Europe. This presents a particular challenge, in that it will not be possible to run the new ordering system alongside an existing system. There will need to be an overnight replacement of a system for a particular market.
The IT managers know the risks of non iterative development, and therefore they propose a series of internal releases, to be put through rigorous quality assur ance. They decide to focus on one market and aim to deploy in that one market once the system has been thoroughly tested and passed all acceptance criteria. They will operate the system for a period before rolling out the system in other markets
The initial requirements analysis determines a set of essential use cases that must be implemented for a usable system. There are a number of desirable use cases and other use cases. These are indicated in Table 5.1. The ‘must have’ use cases will be the minimum that could be released for operation, but ideally the first release should incorporate the ‘should haves’ The project breaks the development into smaller iterations, with three itera
tions covering the ‘must haves’ and two covering the should haves’. iterations have their own Moscow prioritization. They are time-boxed at 3 months each, with the iterations overlapping in time, as indicated in Figure 5.10. The aim is to achieve all the musts and shoulds in the first 12 months, but the system will be deployable if only the musts are achieved.
551 CHARTER FOR ITERATION 1
The aim of this iteration is to get all of the user validation (log in etc.) and customer registration implemented. There will inevitably be new use cases deter mined to supplement the key use cases identified. A Moscow prioritization for the iteration is given in Table 5.2. The motivation for this is that without the ability to register customers and manage users the system is effectively unusable. These are also fairly straightforward use cases, and as the first iteration is Metalwho usually a substan tial learning phase it is advisable to keep the use cases as simple as possible.
The iteration will run for three months, and all the use cases will have been tested within the iteration. The iteration will cover the requirements through analysis, design and implementation to the testing of the use cases.