Some Recipes to Improve Your Google Web Toolkit Development

While struggling with GWT debugging and development I found a pretty good article which has a good reasoning why and how a GWT-based project could be divided into "independent" parts from the development perspective of view: Some Recipes to Improve Your Google Web Toolkit Development. Will try to follow these guidelines right after tomorrow.

Happy Coding Hour with Philip.

I missed all deadlines for making this post, but had a good reason, that I'll explain further:

So, back to the coding hour: it was just great. We worked on my GWT telemetryviewer, but most of the time we spent discussing and drawing on the board. Philip asks me questions that I wasn't able to answer and which forces me to learn more. I do appreciate this, it reminds me class teaching experience: you seems to know all, but student asks you are very "dumb" (in your opinion) question - ... and you are lost. You just can not answer straight, you are puzzled by a question because it so simple that it really cuts down to the bones of the subject through all the meat. That's it. I've learned the lesson: even if GWT claims that it hides all DOM and client-server interaction from the developer, I must have an idea how all this stuff works.

Happy coding with Hongbing.

During coding hour (in fact we spent about 90 min) we were working on the UI design today. We tried to follow my instructions about installing the environment and failed since GWT seems to keep absolute paths within its config and datafiles. We had to change hardcoded paths manually.
After stuff was installed and we got mockup working we were trying to understand the GWT hosted and compiled modes along with hosting the mockup within the Tomcat. This task turned out to be pretty difficult since I am using JFreeChart jar and servlet along with JCommon. Basically we gave up and proceeded further.
The last timeframe we spent discussing my UI design. Hongbing was developing HackyStat for a long time and helped me to figure out which way of abstractions would fit previous HackyStat architecture (I want reuse some code instead of hacking by myself). Overall this part was a most productive.

Thinking Linux?

Okay, I was using SuSe for more than three years now, both Enterprise @HPC and Open @home. I like it, I think Novell rules, and would recommend it for everybody, but boy, there are only 2% females out there in the real life using SuSe??? Something wrong out there, because I have two females and two males in my family using SuSe, i.e. 50/50, WHat's up with others 27458 users taking the survey?