News of the day (oh well, may be yesterday) is that the newest upcoming programming language to take on Java's enterprise market share could be Ceylon. A big thank to my good friend Damith for putting me on this news.
Yep, you heard it right. (Check the date of this post if you don't believe me. See, it's not April 1st!) The very first thought to spring out screaming is why that particular choice of name ? Well, I think it's a easy guess (and I'm taking a personal risk of being wrong here).
Folks at Sun Microsystems used to drink lot of coffee while developing the new programming language they intended to take on C++, when they were faced with the task of choosing a name for it, they choose the place where their favorite beverage is coming from. Java. So I bet good folks at RedHat are drinking tea and buckets of it! May be it is the ultra healthy Green Tea. Well, good for them! at least they have got their beverage right, Ceylon Tea!
Back to the language itself : Some have rushed to call Ceylon the (would be) Java Killer. Although the creators of Ceylon themselves are more cautious and pragmatic. Oh, I forgot to give the original article yet. Here it is.
According to the Gavin King (as mentioned in the article above) lead of the Ceylon team at RedHat, New language will have following goals : It would run on the JVM and feature static typing, automatic memory management, and built-in modularity.
Following current problems in Java were also given. Tied too much with XML, absence of a good way to design a user interface, A lack of language-level modularity, issues in Java meta-programming, Java SE (Standard Edition) SDK lacks modernization and is bloated etc.
Back to Ceylon vs Java. Now then here are my interpretations :
1) Language is being made by a very capable team at RedHat who are already "unapologetic fans of Java" So it's a safe bet (unless they make some really bad design decision upfront) that the new language will be a better one than Java. When Subversion team wanted to build on existing CVS and make a better one out of it, they did it, for instance. These folks know the strengths and weaknesses of Java so this part is a safe bet I guess.
2) The point above does not in any way imply Ceylon will beat Java at the enterprise software market. Being just a better language is not enough for that! Developers should have a very compelling reason to switch over to a new language. I think Ceylon will not beat Java in a hurry and become the top language in that sector in any time soon. There is simply too much inertia for Java right now. Unless.... well, lets get back to that in little later.
3) I'm not sure how important it is to have a "good way of designing a user interface". While it is certainly a good thing to have, I think the days of thick clients are on decline. To be the best future language, this user interface design part (assuming they mean thick client user interfaces, like in Java Swing) will not contribute that much. However it will be hugely important to have a light weight framework (MVC preferably) to develop web applications, may be that should be very close or built in to the language components it self, or may be the language should make it really easy for others to develop such frameworks. Web applications are going to be the future, current Java frameworks (struts, JSF etc) suck.
4) Having a JVM and automatic memory management is a must these days nothing surprising there.
5) To beat Java at their own game, one need some revolutionary new features, aspects or paradigms. Let's see whether the new language has got any. May be they have few cooler stuff hidden on their sleeves but if they have I didn't hear King mentioned any yet. Let's see what we mean by revolutionary changes.. How did Java beat C++ ?
Java had cool features C++ could only dream of. Java introduced (or got it right and made it popular) memory management, a huge issue in C++, a JVM a very novel concept which gave them the theme "write once runs everywhere" (or "write once debug everywhere" depending on which camp you are in), garbage collection etc These were Huge improvements in those days that C++ folks have migrated en mass.
What would we like to have in a future language which would embarrass Java ? Here is my List, for more see in here :
1) Dynamic language features (such as in LISP, Ruby) lot of meta programming.
2) More support for scripting. macros etc
3) Do away with Type Safety and statically bounded types (this would be a tough one for a enterprise language)
4) Inherent support for web applications. (Java introduced JSP, where to go from there?) and AJAX in particular.
Success of Ceylon being a Java Killer will depend on those factors (and also probably few others I have missed) I think. So far I haven't heard much on those lines.
Ceylon would be a better language but will it be better enough to convert Java followers, that's the question!
However, there is no doubt that the King and his team at RedHat have got the most important thing absolutely right. The name!! that will certainly give the new language some wind initially. (partisan thinking here, take no notice! :-) )
Going to have my cuppa now.. and I like the green type.
(Image source http:/www.decidetostayfit.com)