Study On Big-Small Iteration Software Development Model And Define Method Of Iteration Scope

Software engineering is a set of methodology aiming at software, a product with special characteristics. It focuses on the lifecycle of software, from vision, design, coding, testing, running to maintenance. The lifecycle norm can be applied to different software products, which becoming a software development model. Due to the large scale involvement of human being and human's knowledge in producing software, software industry can be regarded as a brain-intensive industry. It shows the necessity of manage people and their manners, which becomes an academic subject called project management for software projects. Software project management mainly studies the management of time, people, cost, quality, risk and other aspects for productivity purpose. Software development model and software project management are two of the most important aspects in software engineering domain.The main study of this dissertation is digging into the details of the development of business and implementation driven large scale enterprise management information system, whose target release is obviously constrained. In order to improve customer satisfaction on the software project, two practical conclusions have been drawn as following:1. Big-small iteration development model is proposed the first time here. Nowadays, project customers and system users always show their lost-confidence towards IT application and the development team. Base on traditional iteration development model, the brand new model is to find back the trust between both parties. There are several useful points can show up customer key concerns, collected by the investigation on the user feedback: First of all, the lead time for development team to develop specific function and the round time from requests raised till system implemented is the great concern. Second, whether the release is what users want mostly and urgently. And third, whether the process of requirement handling, development is clear to user who raised requirement, can he/she track the status of his/her needs and understand what he/she or anyone else should do to follow up. Targeting towards above three concerns, a big-small iteration software development model is proposed, that is to say, with the constraint that production release date fixed, arrange the big iteration as traditional iteration for planning, requirement elicitation, coding, testing and so on, which may more focus on functionality having great impact to existing system or some brand new functionality as a whole solution, etc. On the other hand, arrange a small iteration at the late part of the big iteration, at the same time with the testing phase, to complete those items on usability, beauty, informative and other less-impact enhancement items. By doing so, engineers can understand the objectives of each time frame and adjust by themselves to fulfill the requests and provide high productivity. Simultaneously, all kinds of user requirements, large or small, can be arranged comparatively in balance, so that the product to be released can have maximum requirements covered.2. A new software requirement evaluation template is produced in this article to help iteration scoping. To solve the difficulties– how to select most useful, urgent, valuable requirements to come up to a stable iteration scope for a release, it is suggested that all aspects should be considered like business, development, testing, implementation and system support, and then to document down those useful information as the description of a requirement so that decision making group can make final determination easily.This dissertation makes a common study on the iteration development models widely used in today's software industry first. By analyzing the average responsible time from a requirement is raised out to coded, tested and ready on production, and studying the workload of engineers especially requirement analysts and development engineers, some room for enhancement is found and a new development model is proposed. Then, from the business behind a requirement, to a requirement's impact to development team, further to how to implement the functionality with a new software version ready, this thesis suggests a template with those valuable information listed which can help on priority of requirement judgment. Last, but not the least, a case study is introduced here which using the new development model and the requirement priority template, to prove that such proposal is workable in real project.