“Ceylon is a programming language for writing large programs in a team environment. The language is elegant, highly readable, extremely typesafe, and makes it easy to get things done. And it’s easy to learn for programmers who are familiar with mainstream languages used in business computing. Ceylon has a full-featured Eclipse-based development environment, allowing developers to take best advantage of the powerful static type system. Programs written in Ceylon execute on any JVM.”
I knew nothing about Ceylon before reading their page. They seem to have a novel and productive mission. A very interesting niche product.
Only problem is that everybody and his dog want to do a new more productive JVM language.(Scala, Clojure, Groovy, Xtend, Fantom, $RandomIsland(Ceylon,Kotlin etc))
They will all have one thing in common: “Not enough traction”
I would like to see the majority of Eclipse Foundation members agreeing on one blessed statically typed alternative to Java.
You may be more right than you know. I find myself guilty of not using the best languages for the job simply because they’re not popular enough. In order for a platform to thrive, there need to be others in the ecosystem taking up commitments to push it forwards. Without sufficient backing, regardless of merit, it can never reach a high potential.
The truth of the matter is that these languages will sink or float based on their abilities to market themselves to big players who have the momentum to affect change.
Personally I think we already have enough languages targeting the JVM.
Somehow the latest JVM contenders, Ceylon, Kotlin and Xtext seem to show a “gold rush” trend where everyone tries to produce the next Java.
I think that effort could be better spent creating languages with native compilation by default, or offering better ahead of time compilation tools for the JVM and CLR environments.
I agree 100%.
I think people should, take a look at improving other Virtual machines. While I’m impressed at the variance in languages offered on VM that was originally designed with one language in mind, I like the idea of competition of ideas on all levels of computing from the processor all the way up to media players and note apps.
Could some one show parrot some love?
http://parrot.org/
I agree. Parrot is a great project, that just lacks some publicity. It certainly has everything needed to be a major VM, well everything besides publicity. It’s way too much the Perl 6 VM, even though that’s not true as there are already tons of programming languages (partly) implemented in Parrot.
The problem is that there are JVM and .NET/Mono, which are pretty much killing every other project, not intentionally, but they still do.
I think Parrot has also a slight problem of presenting itself, which is _very_ sad, because it leads to people think that it isn’t professional, which it isn’t but in the way like Debian or CentOS aren’t professional projects, meaning there isn’t a huge company behind like in the cases of .NET, Mono and JVM. However, the quality is extremely high. This can be seen by building and installing Parrot anywhere. It beats pretty much anything I’ve ever seen anywhere when it comes to this. Portability is usually something very important for a VM and maybe Parrot’s killer feature. The others don’t even come close, especially not out of the box.
@moondevil: Take a look at Crack at http://code.google.com/p/crack-language/ which uses LLVM for JIT and AOT compilation.
It’s portable to platforms that the LLVM runtime is ported to, with a “bitcode” intermediate format that’s analogous to JVM bytecode, although at a lower level.
One advantage it has over VM-based languages is that there’s no barrier between it and native libraries. And of course it benefits from all the LLVM work on optimization etc. from lots of people.
Edited 2011-11-28 16:47 UTC
Thanks for the hint, but I fear the language designers have chosen a poor name for it.
Nod.
Names can be changed much faster than languages/runtimes can be built though, so I’m not letting that bother be too much.
(full disclosure: I’ve contributed to the language)
The spec looks very nice. Testing it’s quite hard because you have to get everything from Git and build it. In case anyone else wants to try it, I’ve just published 0install feeds for the compiler, parser, runtime and sample program.
To try it (replace “apt-get” with “yum” on Fedora, etc):
$ sudo apt-get install git zeroinstall-injector
$ git clone git://zero-install.git.sourceforge.net/gitroot/zero-install/hello-ceyl on
$ cd hello-ceylon
$ make
[…]
Hello Bob
Seems quite immature, though. e.g. the “nonempty” test in src/helloworld.ceylon always passes, so you’ll get a NullPointerException if you try to print args.first, which was the first thing I tried.
[ note: if you get KeyError: ‘run’, this is due to a bug in 0install 1.2 (fixed in later versions); just run “make” a second time and it will work ]
Edited 2011-11-26 16:54 UTC
Post looks like advertise. It seems to me that somebody pays for such posts. I haven’t hear about this language before <a href=”http://www.essaypedia.com/“>essaypedia
Now, aren’t you something – a spambot grumbling about paid ~spam.