On Tuesday, I’ve attended a couple of panels and 2.5 sessions.
The morning panel was titled “Principles of RIA” and was about bringing animation and effects to your RIA to make it more engaging. For some reason it didn’t get me excited as I have to deal with more prosaic issues on a daily basis.
The next hour and a half I spent sitting on the floor in the overcrowded room where Deepa was reviewing new features of Flex 4, which were not Spark components.
During the lunch I was invited to speak at a conference in Mexico and discussed specifics of working as a consultant on government projects.
Then, I made a wrong choice because of the misleading title – I was expected to see more advanced stuff, but it was an intro type session. C’mon, stop writing ANT manually – check out our free plugin FX2Ant that generates ANT script from your Flex project in seconds.
My next selection was the session on Test Driven Development (TDD) by Elad Elrom. This was the best session of the day hands down. There are situations when TDD can save project development cycle. Writing tests before the actual application is written forces people to better understand the application being developed. Ten year ago using UML-based code generators would have a similar effect – we had to think before code, but we’d had to test code manually.
In the 21-st Century, the legions of low skilled people rushed into software development and the more coding generated and tested automatically the better.
During Elad’s presentation I’ve learned a new term for something that we all experience in the corporate world: Asshole-Driven Development. Here’s the slide on the subject:

The technical part of the day was closed by another panel. This time it was on the business of software – how to estimate development cost, how to talk to clients, how to run teams etc. This topic drew a lot of interest among the developers, but to me, most of the answers were sugarcoated. The panelists were saying the right things, but to my taste, it was not open enough. There is a lot more BS in this area than it was presented. I liked the comments made by RJ Owen (he was on the panel) – RJ has a good sense of humor.
A couple of takeaways from this session are:
1. Development can be good, fast, and cheap – pick two.
2. When you are developing a project, it’s like working in the construction business. The difference is that instead of wood and bricks you have to use live kittens – they move, play, fight…
The level of the attendees is different. For example, one person told me that he wrote a large Flex application – 8000 lines of code in one file. But he understands that this is not right. In today’s podcast The Flex Show one girl said, “We use a lot of Flex, especially for our front end.” Nice!
Organizers of the conference marked each presentation by a level of complexity: 100, 200, and 300. Mine was marked as 300, but one guy stopped by saying the he loved it even though it was of a level 400. He also said that during the preso I “sold” a lot of our copies of our upcoming book. I was not presenting to sell books, but hey O’Reilly, have you heard this?
At the end of the day I went for dinner to P.F.Chang’s with Shashank, Tom, Jeff, and John and group of other people. Five Adobe evangelists were eating there already. We’ve asked the waiter to give us a table far from them assuming they might need privacy to discuss some secret things, like the date of Flex 4 release. He-he, if they’d read my yesterday’s blog they’d know that I calculated that date already.
For me, it was the second day in a row at P.F.Chang’s. I was the only one who have been there already. At this restaurant, they bring you the bill and fortune cookies at the same time, I guess, to ease the pain. Shashank’s cookie had something about multi-touch, which is a good sign since he’s presenting on Wed on this subject.
Wednesday is the closing day of the conference.
Another day, another dinner at P.F. Chang’s.
Yours,
Yakov Fain