I used groovy to build an internal app that worked with subversion, jira, and a local sqlite db this week and it was so much fun. It was so productive that I decided to put a web ui on top of the core app. Rather than use grails I used groovlets with the built in HTML builder. My favorite line of code had to be:
(Also posted as a comment on the blog:) Groovy would have won the "Java.next" title by now if they weren't so stubborn about being a dynamic language. If they invested half the effort they've put into "IDE-time" static type inference/type checking in Eclipse/IntelliJ (so that programmers can still get the warm fuzzies of auto completion) into building static type inference/checking into the language itself, I think they'd have seen much wider adoption.
This is basically what the guy behind groovy++ has done, but, from what I've heard, they basically told him several years ago that they had no interest in taking the language in a static direction.
This sucks because, in IMO, 90%+ of the calls in a Groovy program can be statically dispatched just like regular Java. Sure, duck typing is useful in a few places, but those few places shouldn't mean my entire program has to be dynamic.
Groovy has an awesome syntax, awesome compile-time AST transformations, etc.--but a dynamic language, in enterprise environments at least, will not be Java.next. Scala is maturing, and Groovy's window of opportunity is quickly closing (or already gone).
I've been building the web app I'm working on completely in Groovy, and running it on Google App Engine. I'm quite happy with my choice to switch over to Groovy from Java. There are 3 main features of Groovy that greatly speed up my productivity as compared to coding with Java.
1. Closures - items.find { it.name == 'groovy' } is much easier than looping.
2. Language support for collections - Maps can be declared like: [key:'value'], and lists can be declared like: [1,2,3].
3. Easier handling of nulls - The "Elvis Operator" can be used to specify defaults like: (item ?: "defaultvalue") and the safe navigation operator will return null rather than throwing a NullPointerException when doing something like this: (item?.address?.street).
As much as I've enjoyed writing Grails apps and Groovy scripts, what I don't like about groovy is that it comes with a large runtime jar and its binaries are not really usable from java projects without API uglyness.
I wish Java had been designed with the same multi-language philosophy that .NET has had from the get go. Binaries produced in any language in .NET don't give any clues to what the language the binary was implemented in. For example, if your writing something using VB.NET and working with a library that was implemented in C# there really isn't any evidence that this is the case and the signatures look the same.
But that's mostly because VB.Net and C# are isomorphic, it's the same language with different syntax.
I remember reading about the first versions of Eiffel on .Net. That gave the impression that unless you put a lot of effort into your language, it will probably be just a variant of C#/VB. The first versions of Eiffel for .Net lacked multi-inheritance and I think DBC was crippled somehow too. :-/
Actually, this isn't quite true. Close, but the languages do have a few differing features.
Your main point is a little questionable - sure, there are languages very similar to C# (Nemerle, J#, etc.), but then you've got F#, P# (Prolog), IronScheme, IronPython, IronRuby... I mean, the existence of the DLR itself kinda puts holes in your argument (for some of the above languages).
Claiming "yeah, but the F# team did everything" is somewhat true (as they did for the first release), but it's not a meaningful complaint. That's going to be the case for virtually any new paradigmatic language being put onto a general-purpose (i.e., procedural, object-oriented) programming framework. The Clojure team had to do everything to make it work on the JVM, too. And in current releases of .NET, more and more of that is being pulled into .NET core and being leveraged by C#.
Unrelated, but something is up with the encoding on your site (or the quotes used), and it makes it terribly hard to read and comprehend anything. (win7/ff4)
The article is using unicode quotes (\x92 and \x93.) You can just switch the character encoding used by Firefox in view > char encoding to UTF-8 and it renders fine.
Chrome reads it as UTF-8 without prompting, I'm curious why that isn't the same in Firefox.
It identifies itself as UTF-8 in the XML header, but is sent without the XHTML content-type so that's ignored and the encoding is ambiguous. (If it were sent as XHTML it wouldn't render at all, since it has an XML error.) The author should probably correct this.
Groovy's syntax is so close to Ruby's you can barely tell them apart sometimes, and its tooling as pretty good, certainly way better than what I believe is available for Go.
> There is also the fact that the JVM(currently) doesn’t have support for dynamic memory dispatching which is feature vital for dynamically typed languages to get a decent performance.
Can someone explain what "dynamic memory dispatching" is? Is it perhaps better known under a different name? (Google is no help.)
The company I work for tried incorporating Groovy into several of our projects, but we started getting random permgen JVM crashes as a result. I believe this is due to the large number of classes it creates behind the scenes. This left a bad taste in peoples' mouths and unfortunately groovy will now be a hard sell.
Edit: Yes, we increased the permgen size and whatnot, but the outages it caused just left people with an unfavorable view of groovy. Which, as I said, is unfortunate, because I think it's pretty cool.
My favorite approach to using groovy in projects is with the groovy compiler plugin for maven. Compile and bundle the resulting byte code. All you have to remember is too bundle the groovy runtime and remember how groovy interacts with Java natively.
The one thing that disturbs me about groovy is that overloaded methods seems to only be signature checked at runtime, which results in an exception when invoked with the wrong parameters - not a compiler error.
Are you asking why the article was about Groovy, or "why groovy?" in a broader sense?
On the 'broader sense' angle, Groovy is a good for certain people/teams who are looking to get more out of Java development without giving up too much of what they have invested in. Pretty much all Java code will run in Groovy unchanged, and you can update existing working Java code with some Groovy decorations in the same file without issue. Moving to JRuby or Jython, for example, require you to make larger changes to rethink entire classes in the new syntax.
Admittedly some people consider this a small point, but it's certainly nice for some people to be able to reuse existing Java syntax without needing to learn a lot of new syntax up front - with Groovy you can try out new syntax as you go, but always drop back to raw Java when you want (again, in the same file - you can use existing Java classes in JRuby and Jython IIRC).
* Groovy is easy to develop in - it's dynamically typed and far less verbose than Java
* It's mature and stable
* It's easy for Java programmers to learn
* It's easy for Python and Ruby developers to learn
* Grails, groovy.sql.Sql, XmlSlurper and a host of other Groovy libraries
* Documentation is good and support forums are very active
* Groovy is actively developed. Version 1.8 was just released. I haven't had the time to look at the new features but glancing at the changelog they look pretty good.
* And last but not least, Java shops with a heavy investment in Java and Java servers can easily integrate Groovy as opposed to rewrite things in Python and Ruby (For green field development I still prefer Ruby)
Groovy is different to Scala and Clojure, the other two languages in that blog series, in that Groovy uses the Java classes whenever possible, and isn't intended to run on the CLR. Scala and Clojure were both intended to be dual VM from the start. Rich Hickey says the 'C' in Clojure is for CLR and the 'j' is for JVM. Scala and Clojure have their own collection classes runnable on either VM, while Groovy is JVM-bound.
And of course Python and Ruby do it differently, having J- and Iron- implementations, each at different versions to the primary implement and each other.
table{ myMap.each{k,v-> tr{ td k; td v;}}}
that turned myMap into a HTML table.