|
| Article Details |
| |
|
|
Dallas, June 2, 2003 – Today there are a remarkable number of interesting C# presentations at TechEd. The day started with “Extreme Programming with C#” presented by Ken Schwaber. Of course is Extreme Programming as a method independent from any programming language, however, it was interesting to listen to Schwaber’s experiences in this area. No longer waiting for twelve months for applications that nobody really wants, but every thirty days working deliveries, slowly but surely moving towards the end goal. It becomes accepted that it is impossible to specify large, complex software projects entirely. According to Schwaber it makes much more sense to have some idea of what an application’s functionality needs to be and to refine that idea by frequent iterations. The goal of the iterations is to deliver working solutions from day one with a frequency of about four weeks. In this way it is possible to always have a running system. After some time the functional requirements become clearer and clearer. The amount of work per iteration is not too much and having regular feedback with the client is a clear benefit. Schwaber is a big promoter of software code that is documentation at the same time. Test first development is absolutely necessary, with the advantage that the test automatically becomes a low level design of the functionality. During iterations a software development team has its own responsibility for reaching milestones, delivering good quality software and delivering the desired functionality. The team decides in a scrum (a short meeting) what the next things to implement will be. This approach has also been used during the development of .NET functionality at Microsoft.
Dan Fernandez’ presentation “C# past, present and future” started by looking back to the young history of C#. Fernandez said that six months ago C# was used by 7% of all developers in the USA. Today about 12% of all US developers are using the language. The expectations are that 24% of the developers will use C# one year from now. According to Gartner there will be around 2.5 million C# developers in the year 2006. Fernandez believes this is a reason for each developer to start today with C#. It is a great career move, because today’s developers will have at least 3 years experience with C# in 2006. Something that is very remarkable is the presence of a large developer community around C#, something we also see for Windows CE.NET. Fernandez explained that Microsoft developers also have to eat their own dog food. Large parts of Visual Studio.NET and the .NET Framework have been written in C#. Also a number of server products and MSDN, MSN and GotDotNET have been written at least partly in C#. However, according to Fernandez it is not likely that all Microsoft code will migrate to C#. There is simply too much code to even consider this. Looking to the future of C#, one of the new cool features that will be implemented are Generics. These parameterized data types are comparable to C++ templates. It is already possible to experiment with Generics. Microsoft Research has published tools to use Generics today. Another cool new feature will be support for partial types. This means that it is not necessary in the future to write all code of a class in one single source file, but that classes can consist of multiple source files. Partial types have several advantages. Right now Visual Studio adds lots of designer generated code in between our classes. Our own code would be much cleaner if all that designer generated code could be stored in a separate file. Partial types are also ideal to develop large classes with multiple developers. With this feature it is no longer necessary to work with several people on the same source file. |
| | |
| Back |
|
|
|
Do you want to learn developing a full blown Windows Mobile Application? This article and accompanying multimedia content will help you to do so. It will be extended over the upcoming weeks / months, so check back regularly.
| | | Read Full Article
|
|
|
|