Friday, June 12, 2009

Fan vs. Scala: My background

Well, my previous post on why I prefer Fan to Scala got some attention. It also got a few votes up and down at DZone. Also, I'm afraid I wasn't sufficiently clear in my post. I too rarely am. In any case, it seemed like it might be worth extending my discussion of the points I listed.

I ought to give some personal history first, though. I love Eclipse and what other such modern IDEs do for Java. I can navigate and fairly well tame code bases made of 1000s of files. I also love that Java performance is sometimes near C, except that I can just download and run jars. Getting stinking complex open-source C or C++ code to build correctly can really make me mad. (The download and run mentality of Windows or Mac really beats Linux in this respect, too, though maybe if Ubuntu becomes more ubiquitous, people can just target that and make life easier. Official packages are always so far behind for the few apps and libs that I really want fresh.) But back to Java, I really don't like it as a language. Checked exceptions, needing to repeat oneself, lack of closures, and so on really don't make my day. And I like to learn other languages, new and old.

I occasionally noticed Scala over the past few years. It seemed very promising. When I saw it starting to get attention a couple of years ago, I figured I'd learn it more and help be part of the momentum to see if it could overturn Java. Among other things, I submitted a game to the Java 4K game competition (jar size, not source size!) written in Scala. Scala doesn't shrink as nicely as Java, but I got it mostly working and in the right size. And I kept learning the language. I gave a brown bag presentation at my job on Scala about a year ago.

And just about the same day as my brown bag presentation, I saw posts from Cedric Beust and Stephen Colebourne about this other new language, Fan. I got to the site, looked around it, and I immediately switched my interests. I thought, hey, this is even closer to my own preferences, and it doesn't smell like you need a PhD to understand it (despite the fact that I'm now working on getting a PhD). I don't prefer every decision that had or has been made, but it's just so much closer to what I want than anything else out there with any momentum. It's statically typed with important limitations to structure programs, but it still has the feel of a scripting language. Scala might be convenient (usually), but it doesn't have that same relaxed air. And it doesn't even make some of the same guarantees you get from Fan.

So history aside, I'll try to delve deeper into my list of good things about Fan in the near future. And I'll keep the "vs. Scala" perspective, as well as "vs. Java" and maybe a few others. Nothing personal against Scala. If it's good for you, then good for you. I'm just afraid that for many folks looking at Scala, Fan is a better choice, but it's not high enough up the radar yet. Or the differences might not be clear enough. That's why I'm giving this focus.

4 comments:

  1. My down vote on DZone for your post was entirely because I think arguing that one language is better than another by bashing that other language is a very bad way to argue. The fact that I also didn't agree with your reasons is not really the point.

    ReplyDelete
  2. Yes, feel free to vote me down for any reason you see fit, and thanks for the info in this case. Maybe if I continue the series, I should spend the time to be more even-handed, even if I cover the same points that I personally like about Fan.

    ReplyDelete
  3. Hi,

    You maybe be right but sometimes purity has many drawbacks in the real world :)

    Also large adoption of a language is based on legacy of its creators and popular support. In this batlle, Scala is clearly a winner, don't you think ?

    Now I intend to look at Fan because of your post. I'm just beginning to learn scala ( see http://reboltutorial.com/scala ) and may in parallel look also at fan.

    ReplyDelete
  4. Scala does have more following right now, and I agree that popularity matters. (With enough people working on something, things just happen "automatically".) I'm just working a bit to push Fan into focus, too, since I like it better overall.

    ReplyDelete