Railo: What do You Give a Sh*t About?

Stu Halloway made a really interesting comment at a "Languages on the JVM" BoF I attended at No Fluff Just Stuff in Boston the other week. His argument was that when choosing a language you shouldn't look at the language features, libraries or frameworks currently available as those change way too quickly to be meaningful. Rather you should ask "what does this community give a sh*t about" as that is a much better predictor of whether it'll continue to meet your needs over time. The question I want to ask is, what do we give a sh*t about as CFML developers, and is that a sustainable position?

There are many sub-groups within the CFML community - I don't think Chris Scott, the author of ColdSpring cares about the same things as a graphic designer looking for a simple way to put stuff from an Access database onto their website using CFML. But as a whole, I'd argue that CFML developers have always cared about making hard things easy - with elegant, simple tags allowing someone without much of a formal programming background to build fairly complex web applications.

By Comparison . . .
To take a stab at some other popular languages, I'd argue that PHP developers as a whole care about "hacking it together", using pre-built packages and the absolute minimum of spit and bailing wire to get an application that their client will pay the $500 they committed to. That's not to belittle the work of the OO, TDD and best practices communities that have grown up within PHP. There are many excellent PHP developers and many scalable well architected PHP web apps - some of which would put many Java apps to shame. But by sheer volume of developers, I'd argue that PHP is about getting it working more than making it maintainable.

.net language developers? I'd say they care about following the Microsoft road map. Using a full stack of Microsoft tools that'll integrate and work well together and leveraging solutions like sharepoint to provide mid priced offerings to mid sized companies and departments in larger organizations. Personally I like to hang out with the crazies who are trying to figure out how to integrate svn or git with Visual Studio and who are driving Microsoft to take TDD, CI and other best practices from the broader programming community seriously. But they're not representative of what the majority of .net developers care about.

Scala and Clojure developers probably care about concurrency (although they could just like functional programming or lisp specifically or just be into the "flavor of the month" languages). Groovy programmers generally care about writing Java applications using Java libraries, deployment tools and monitoring/instrumentation solutions - but doing so more quickly than they could if they had to write everything in Java. JRuby devs generally care about Ruby (and/or Rails) and want to be able to get it inside the enterprise by being able to deploy their apps as a standard Java WAR file.

Back to CFML . . .
So, back to CFML. I'd argue the community generally focuses on "making hard things easy". This was born in the era of the "amateur web developer" where *nobody* really knew how to build web applications and we were all learning as we went - computer science background or not. Nobody wanted to open and close file streams or database connections and simple tags like cffile and cfquery allowed us to write powerful applications quickly.

However, as I'll argue in my next posting, I believe that the website development world is changing from a place where we want to focus on "making hard things easy" to a world where we want to focus on "making experienced devs more productive" - and that's a slot that ColdFusion now has more competition in.