Code complete is probably the concept that is best known by everyone. When the software has been designed, programmed and then tested thoroughly to the point where there are no bugs, or at least everything that can be done in a time frame given has been done, then the project’s code is labeled code complete. The program will ship with the code as it is. There are a few steps to go through before getting to code complete. There is the first stage of designing the code, this consists of breaking down the elements you need to achieve the final product and is often compiled into a technical design document (TDD). After the TDD has been written, development will start on the code itself. Programmers will flush out the interfaces specified in the TDD with the code that is the body of the executable program. The beginning development stage is sometimes called pre-alpha and is when the bulk of the programming is done. At the time when the program’s functions have been finished and the program starts being tested in-house a program is referred to as being in an alpha stage. After internal testing is complete and the program is ready to move on to external teams that will try to find bugs that the internal teams overlooked or didn’t consider problems, a program is referred to as being in a beta stage. As software goes into the end of beta testing it goes through release candidates. This is where the software should be fully functional and possibly ready to ship as code complete. Once these tests have been finished and the code is considered code complete. Often software doesn’t go through each of these stages specifically, or it goes through some of them in rapid succession. Thoroughly taking the time to go through these steps will help the stability of the software though since it provides a lot of different input. There is a great book called, surprisingly, Code Complete by Microsoft Press that has tons of great techniques and lessons for making stable code and schedules. |
|||||||||||