The web noise
When I´m again starting a new mega-over-ambitious web project, I ask myself the same question: “What´s is the f&*%ing reason for doing a web oriented project, if this project should be a very tiny, nice desktop project”? Maybe because doing it in a web way is sexy, a tech fashion. Maybe I´m starting to become a old guy with old ideas, but when we stop to ask reasons and start to do things using “best practices” or “experts recommendations” without thinking about, something is very wrong.
Some great pros of thin clients are low deployment costs, low hardware requirements, distributed and parallel platform almost for free.
Some cons of web solutions: the everyday fight against the HTTP stateless protocol, the browsers lack of standards.
Some pros of rich clients: rich user experience (you are in control of every pixel on the screen).
But there are great means to low the deployment cost of desktop applications. The Java Web Start Technology is under used, but a great to glue all together.
The nature give the most simple answer which works, and in development isn´t different. If the desktop development was the first to appear, it was because
it is the most simple solution which works. The tendency to create only web applications today is increasing the projects costs,
and helping its younger death.
When starting a new project, think about. What your opinion?
Great RCP toolkits for Java: Netbeans and Eclipse.
Great Web applications for Java which FAKE a desktop: WingS and Echo2.
