It has been fun to see the open source model working with FlexMonkey.  Stu put our testing tool out there for the world only a month or two ago, and we are already seeing significant contributions.  The latest contribution is from Marvin Froeder (VELO), who has helped us update the framework to work with Flex-mojos (the popular build tool).

I have been hearing really good things about Mate for a few months now.  I have been rebelling against dragging additional frameworks into Flex just for MVC for a while now (Flex and MVC /  Cairngorm).  I finally got a chance to spend some time coding with Mate on a flight back from San Fransico this weekend.  I think it looks promising.  I am still not sure the leverage is enormous, but the cost seems low.  It appears to provide a reasonable structure for Separation of Concerns and managing events, without requiring much additional code.  Hopefully, I will have some more time to do something more complex soon with it, but I think it may be the type of framework I have been wanting.

James and I have posted our latest video, Performance Pitfalls of Flex’s ArrayCollection.  It is a follow-up to my recent post on the problems we encourtered with ArrayCollection.

Enjoy!

Chet Haase, an engineer on the Adobe Flex Platform Team, shared a few thoughts this week on how JavaFX and Adobe Flex compare.  Chet brings a unique perspective, as he was a member of Sun’s JavaFX team before joining Abobe.  You can read his comments in my InfoQ post.

I was chatting at the office last week about code quality, and thought I would jot down some of the things we talked about.  I think it is fair to say that it is difficult to know when code is “good” and when it is “bad.”  There are of course obvious times when the coder was incompetent, or cut corners with little explanation.

Ignoring those types of issues, judging code is difficult because we do not understand the reality of when the code was written.  Was the code written with an extreme deadline?  Was it built with a clear expectation of being prototype code and then put into production by some aggressive manager?  Was the architecture changed multiple times without time to rewrite / refactor?

In addition to not understanding the actual circumstances around the time of original development, developing perfect code is expensive and there is a point where we all need to ship our imperfect code and move on.  Only a really rich, somewhat misguided, client would pay for the real amount of time to develop perfect code.  And of course, we have to understand that much of our feelings toward code come back to style and preference, and not everyone has the same approach to these items.

So, let’s all try and buck the tradition of bashing the last guy’s code, and when there are real issues that need to be addressed then please approach them gracefully.  Remember, someone will be following you soon too!

Corner-Case-Driven-Development is an exciting version of development where all extreme corner cases are highlighted each step of the way.  It is especially important to interject a corner-case once someone starts to talk about the fundamental features of the system, as getting distracted by the core things, like what a user wants to accomplish, is a no-no in the corner-case methodology.  When embracing corner-case-driven-development, a team can expect a number of exciting things to happen, including: difficulty developing and maintaining the code, un-met deadlines, and generally low team morale.

Read about more exciting methodologies here: http://www.scottberkun.com/blog/2007/asshole-driven-development/.

I am listening to the latest Java Posse podcast, a session from this years round-up titled, Don’t Repeat Yourself. In the discussion, Tor Norbye made the statement that, “programming is all trade-offs.”  This is a great point.  For what ever my strengths and weaknesses are as a developer, I have long believed that one of the main values I bring to each of my projects is a natural cost benefit analysis mechanism.  It seems that most anything is truly possible in software these days.  So, understanding the costs of each decision is one of the most important parts of building quality software.

Well, I have been pushing him to get in the Blogging game for a while, as he brings a unique take to all things software.  He is our CEO that codes.  I don’t know what that says for his CEO skills, but for us developers it makes for a great boss.  Anyhow, here is the official introduction to Mr. Stu Stern’s blog, Big Gorilla.

In Thomas Friedman’s latest article, he suggest we have Steve Jobs run the car companies for a year, as part of a bail out deal.  After living in an IT industry built on inovation for the last decade, it is hard to understand how our automakers have failed so clearly in this respect.  And, it is even harder to see them heading towards a taxpayer bailout when it is sink or swim for the rest of us.  How do you ever get better if you keep getting bailed out?  Its like the kid given too much early in life… it is almost impossible for them to ever break the cycle.  I hope its different with our automakers.  I hope Mr. Jobs is offered the job and accepts.  I doubt he will fail to innovate!

Well, this seems like a stupid statement, but for me, “Flex is Java!”  We have heard a lot the last year about Java the platform versus Java the programming language, with the push to add scripting languages to the JVM and add extensions beyond the core programming language.

When you look at Flex you find that Data Services is built to run in a Servlet container.  The IDE is an Eclipse plug-in.  The only thing that really breaks from Java is that the client runtime is not a JVM, but Java has basically failed to ever deliver a viable client side runtime for browser-based applications.  So, why not consider Flex / Flash for this option?

That is actually where my statement comes from, as I think Sun should have started claiming, “Flex is Java” in 2007 as a marketing campaign, rather than starting their client side initiative, JavaFX.  Even if JavaFX ends up becoming a viable technical option, how will Sun ever profit from it?  I know-I know, I read Jonathan Schwartz’s blog and hear they are making money on Java, but in all seriousness… why doesn’t Sun promote things like Flex as successes of the Java platform? Is it just another case of “not built here syndrome?”