Andrea's Home Page

Why writing another window system?

I'm writing this page because a lot of people thinks that starting a project from scratch is a crazy idea, especially when we are talking about a window system. You should analyse the evolution of some projects, and you'll see modifications of basically standards, tons of libraries, that creates interdependencies of a lot of software, which leads sometimes to leak of standards, and projects becomed so big, so it's far from the idea that was started from. Indisputable, radical changes are needed sometimes.

Here are some reasons for writing another window system:

leak of standards: there are hundreds of X-windows toolkits, tons of window managers, and so on.
there is a very small number of object oriented window systems. It is unconceivable to write a non-object oriented user interface in our days. And this is not just a caprice. Object orientation have a lot of advantages, especially for writing user interfaces; you'll see below why.
the UNIX world desperatly needs easy to use and easy to install systems (programs, window systems, and so on); X-Windows is hard to configure even some commercial servers provides some more help (I'm thinking to Metro-X and so).
X-Windows partially satisfies the actual demands:
- hard to install and configure;
- reversed client-server architecture;
- the network support can be improved;
- application level leak of standards;
- a lot of adaptations that leads to a very big project;
- does not provide a coherent printing support;
- X applications are a quite hard to write;
- hard to extend and even to maintain;
- high level of resource requirements;
- operating speed can be improved.
the need of overtaking an obsolete technology and the necesity of a new vision in the world of user interfaces and window systems.