Rich User Interface, a real need

Just saw an iphone, the design is quite cool, but if I have one, after a week, it will look the same for me. But the user interface... I've to say it blows me, to be specific, the ideas amaze me. The accelerometer, I heard that Samsung has a motion sensor in one of its phone, but this accelerometer and most importantly the implementation on iphone reminds me that user interface is indeed very important (I really don't know how to say how I feel the first time I saw it).

In the case of application that the users use in PC, as a developer, it have to be designed such as that it will make the user feels like to use it not being forced to use it. Does this mean RIA is the answer? ... I guess I have to say most probably yes. My tendency was to build a simple but effective UI, but the experience I had today changed my mind, the user has to feel like using the application from the first time he/she looks at it and it will ease user acceptance of the application.

Small thing matters!

Small thing matters, for the last few days I've been thinking about it. Read here.

Adding new project in subversion

To add a new project in subversion is very simple

1. Go to the repository directory
2. Execute : svnadmin create project_name
3. Check the permission of the new directory (and all files/directories inside), change it if needed (so that svn user can access it)
4. Checkout the project
5. In the directory where the project is checked out, add directories as follow : branches, tags and trunk
6. Add and commit it to svn

Your subversion project is ready to be used.

To schedule or not to schedule

This one I learned it the hard way and it indeed turns me from a not must have schedule to a must have schedule for every project! This really means every project! Even if it is an ad-hoc project (no time for requirement gathering, not enough time for analysis, not enough time for every aspect required for a normal project) No matter ehow important that project is.

The last project I've been working in has a very short "deadline" based on when it is represented to the user (not by me) along with a set of features to be implemented. Instead of creating a detailed schedule to measure the effort, what I did was motivating my developers that they can finish it....

Indeed that was a BIG BIG mistake that I made, now knowing that creating a schedule for it might help lessen the expectation. maybe the name of the game is managing expectation... a hard lesson and it must be one of many important lessons that I learned.

Many arguments passed my mind, but failure is a failure, no excuse.
Fyi, the current situation doesn't allow me to have more developers, enough time for testing, analysis, user feedback, etc... Reminds me of a suicide project once a friend told me about, either died along with the project or avoid it, which the later option isn't possible for me, but I'm going to survive with a lesson learnt.

JSF strange behaviors (Lesson - 2008/02/26)

JSF
------
The story :
Invoking action method in a JSF page does nothing, no exception, no clue... nothing.

Lesson :
If you create a getter method and it is referred in the page, you better create a setter method too. I still don't find the explanation yet.
See link here for more details.