While the TIOBE index is a useful metric of the popularity of *searching the web* for a particular programming language, it fundamentally cannot measure the true popularity of actually *using* a given language. Forget measuring lines of code in each language (a lousy metric anyway, given the variability in verbosity between languages), they don’t even claim to measure the number of projects that use a given language, the number of active developers for a given language, etc.

This comment isn’t written in defense of Objective-C, it’s written as a criticism of poorly-applied statistics. You simply can’t claim to infer something from a measurement of another completely separate thing, no matter how related they may appear. The most you can say is that language X is more popular to search for than language Y. Without even the context of whether the search originated from a developer or not, how can you claim anything more? What if someone were to claim that languages which rank at the top of the list are there because they’re inherently more confusing than lower-ranked languages? They’d have as much “proof” as this post presents.

IMHO, analyzing whether Apple made a mistake by choosing a programming language with miniscule market share is essentially laughable. Any programmer worth his salt will analyze a language based on its strengths and weaknesses, and its fitness for a particular purpose. Objective-C is extremely dynamic and well-suited for extremely rapid and flexible design, both of applications and UIs, especially when coupled with Interface Builder. It wasn’t the only choice for OS X and iPhone, but it was a great one.

I think time has already proven that developers are overwhelmingly willing to learn Objective-C in order to program for the iPhone. Especially for mobile devices, the language choice pales in comparison with the quality of the SDK and development tool chain. Apple did its utmost to preserve seamless consistency of its tools, and waited to release a solid SDK, not one that was half-baked. In that respect, I’d say they hit it out of the park.

Aside from some exceptions, such as games, it’s not obvious that many consumers are really prepared to pay for any software that runs on their desktops.

As for JavaFX phones. I agree that JavaFX phones that are simply poor clones of iPhone are unlikely to do that well. A big problem handset makers have is that, aside from Apple (and maybe RIM, sometimes), they appear to find it difficult to design attractive handsets. The truth is that many handsets on the market today are less attractive to consumers than the handsets that were on the market two years ago.

All this means…

- We’ll have to wait and see for the details for Project Vector.

- We’ll have to wait and see whether JavaFX gets into any mobile handsets that consumers actually want to buy

Check out the demos, they’re amazing.

Java sucks, and is trivial on the web. Ok. I missed that memo. Whenever I don’t install the Java plugin on a newly set-up browser, I can’t help notice that half the friggin’ sites on the we don’t work right. That’s right, sites *work* now, they don’t just sit there. And no big enterprises use Java, no sir. I bet Oracle uses an ObjC app-server/middleware suite to serve up high-availability ObjC apps over TCP/IP. Why, I bet you could use the world’s most introverted language to program an awesome platform agnostic online version of….I dunno…iPhoto or something. Sweet.

And Flash, well it has its liabilities, but it does come in useful on sites like YouTube, MSNBC, CNN, Facebook and MySpace…but your right…nobody goes there.

I think the sensible thing to do is, go buy a $600 phone that will try to lock me into the special asylum reserved for MacHeads even *more* or I can’t program it for shit, and which doesn’t support any technology not blessed by SJ– and then post on blogs about what an uberUser I am ’cause my effing phone costs half as much as my already over-priced white paper-weight of a pointless PC does.

Wait…not completely pointless. It came with iPhoto. Wow. You won’t see a killer app like that on any other platform.

Here are some points to consider.

#1. developers are often biased when it comes to technology they already know and are comfortable with.

If you asked Apple developers, who had lots of experience with Objective-C, whether Objective-C was the right choice for iPhone app development, what would you expect them to say?

Ditto for Java developers working with Java frameworks, open-source libraries, Eclipse, etc.

Ditto for Microsoft developers…

You get my point. Most people tend to defend what they know and like, it’s human nature.

#2. I’m pretty surprised the extent to which people poo-poo Java and Java developers.

There are a few subtopics here (A and B).

#A.
Yes, it’s true Java hasn’t succeeded in terms of client-side application development. No need to beat that dead horse.

That isn’t the full story though.

What language has arguably changed the face of UI development forever for the better? It wasn’t Java, and it wasn’t Objective-C either. It was HTML.

The web took front-end design out of the hands of the developers and put in hands of designers, thanks to easy-to-use presentation languages such as HTML. Designers made web apps usable, and this in turn has changed user expectations in terms of application usability, regardless of whether those apps are web-based or not.

So who cares if Java isn’t good for client apps?

What do people typically use Java for? They use it for everything *but* writing the UI itself. Middleware, business logic, backend code, MVC frameworks using HTML/CSS as the UI part, etc

#B.
We could argue the pros and cons of Java as a language. But let’s step back and look bigger picture than just the language itself.

Java, maybe moreso than any other language in the past decade, has been involved with changes in *how* we develop, esp in terms of setting trends, paradigms, and tools.

Consider these examples:
- Java was arguably the 1st language to make automated unit testing very popular, heck, even expected, as a part of software development, because of the widespread use of the open-source framework JUnit. I’m a big advocate of unit testing, the payoff (code quality, usability, and ability to refactor) is huge
- a lot of popular frameworks were initially implemented in Java, then were ported to other languages. Not the other way around
- Java frameworks were among the 1st to widely either adopt innovative programming concepts and/or help make them popular. E.g. dependency injection (Spring), convention over configuration (Maven, and notably Ruby on Rails which is obviously not Java-based), generating API documentation from code comments, etc
- Ant (and now Maven) was a huge leap forward in terms of compiling/building/packaging code. I remember the days of C/C++ makefile writing and maintenance, painful to say the least
- there are a crapload of open-source Java libraries available, many of them well worth using. There is simply critical mass
- in my experience leading development teams, Java is a excellent language from a TCO perspective. Easy to learn, easy to troubleshoot/debug, tons of open-source libraries available, etc

Are all the people who developed with Java just morons? Did they all write crappy code? I doubt it Java is not the silver bullet for all app development (no language is), but it has brought a lot of *business* value to technology and helped to change how developers “do” software.