Thursday, September 15, 2005

Is Ruby Better Than…?

As I speak about Ruby with folks around the world, a question comes up with depressing regularity: “is Ruby better than Xyz?”, where Xyz is the questioner’s current language du jour.


The simple answer is “no.”


But then let’s ask the same question differently. Is Xyz better that Ruby?


The simple answer is “no.”


The longer answer is that the question falls into the same category as “have you stopped beating your wife?”—it makes an implicit and (hopefully) incorrect assumption. In this case, it assumes that it’s possible to compare two languages and come up with some kind of binary A trumps B evaluation.


Is your chisel better than my hammer? If I’m forming dovetail joints, yes. If I’m nailing two-by-fours, no.


Is my Ruby better than your Java (or C#, or …)? It depends.


So, over the next few weeks I’ll try to post some answers. I’ll look at areas of Ruby which seem to be hot buttons for folks when making language comparisons.


Let’s start with a big one:


Performance


Ruby is probably slower than Java or C# in most real-world situations. Does that worry me? Not normally.


Think about your average web application. An incoming request appears on an Ethernet interface. The packets make their way up through the IP and TCP stacks, then get queued up as a data buffer. The web server receives the data, decodes some of the content and passes it off to your application server. There the information is massaged into some kind of request object, and your application is invoked. Your code hits the database a couple of times, probably using an OR mapping layer such as Hibernate to sanitize the access. It then constructs response data before calling some kind of templating library to format a response. The response heads back through the application server, and possibly through the web server, before being deconstructed into packets and sent over the network.


Whew! There’s a bunch of processing going on there. Millions and millions of machine instructions being executed. There’s latency while packets are queued, processes are scheduled, threads are dispatched, and disk heads seek to the right track.


How much of this processing is done in code that you write? How much of the total time spend handling the request is spend executing the actual instructions corresponding to your code?


Zip.


I’m guessing maybe 5% on a really bad day, with a complex application. Let’s face it: you average commercial application isn’t burning CPU cycles solving NP-complete problems. We typically write code that moves chunks of data about and adds up a couple of numbers.


In these scenarios, is it worth worrying about the relative performance of the language used to do the moving and adding? Not in my book.


Instead, I’d rather write in a language that let’s me focus on the application, and which lets me express myself clearly and effectively. And, if I can do those two things, I believe that sometimes I’ll be able to write code which is cleverer than something I’d write in a lower-level language. A better algorithm will easily gain back any marginal performance hit I take for using a slower language.


Justin Ghetland experienced this recently on a Rails project. Having coded the same application twice, once in Java and once using Ruby on Rails, he was surprised to discover that the Rails application outperformed the Java one. Why? Justin believes it’s because Rails does smarter caching. The Rails framework contains some very high-level abstractions, and that allows the folks writing the framework to be smart about what they do. They accepted the linear hit of writing in a language that executes more slowly because they got a non-linear increase in speed from being able to write better code.


Clearly there’ll be times where you need to squeeze the most out of your CPU, where your application itself is the bottleneck and it’s CPU bound. In these cases, Ruby might be a bad choice. You probably want to look at a high-performance language such as OCAML. But even then, the choice isn’t always clear. Ruby with a good parallizing matrix library would probably beat Java or C# code which inlines the same operations.


So, is Ruby fast enough for your application?


I don’t know. But I do know that I wouldn’t assume that I can’t use Ruby for performance reasons. There are plenty of sites out there pumping high transaction volumes through a Ruby application. The real answer, as always, is it depends. If you’re concerned about performance as you start to develop a new application, it doesn’t matter if you’re writing Ruby, Java, C#, or assembler. It’s prudent to spend a small amount of time doing some rough tests on your proposed architecture before you start laying down the code.

No comments:

Post a Comment