Zenbits are posts which include a variety of interesting subjects that I’d like to talk about briefly, without writing a post for each of them.
Merb: A few days ago Merb 1.0 was released. Congratulations to Ezra Zygmuntowicz on this important milestone, the Merb community and Engine Yard (who finances the project). Merb 1.0 wasn’t even out yet when some people had already started commenting on the fracturing of the Ruby community that this new framework might bring with this, and the impact that this high visibility “competitor” might have on Rails. I believe that having more than one widely adopted web framework will only benefit the Ruby community. Furthermore, it’s important to remember that this is not a zero-sum game. Ruby programmers are perfectly capable of learning two frameworks and using one or the other, depending on the project at hand. This is particularly true if we consider that Merb, for all of its advantages – and disadvantages – when compared to Rails, is not totally different from its forerunner. If you are an expert Rails programmer, you should be able to become proficient in Merb in very little time. To help with this process, the Merb community needs to concentrate on the documentation now, given that the API is finally stable.
Rails Myths: David Heinemeier Hansson began a series of posts about Rails Myths. I like the idea of seeing common myths addressed straight from the horse’s mouth. Over the past two years, Rails has received quite a bit of backslash and old fashion FUD, so it’s important to set the record straight, whether the myths are entirely fabricated or if there is some element of truth to them. Whether you agree with David or not, it’s also nice to hear two sides of the same story. In fact, at the beginning of my book I debunk a few myths, just to set the record straight regarding what some readers may have heard surrounding the framework. It was a fun part to write.
My Book: Speaking of my book, Ruby on Rails for Microsoft Developers, I’m getting closer to the finish line. I’m about to complete Chapter 9 (out of eleven chapters). The initial schedule I was provided with has been extended slightly so that there will be sufficient time to properly review the content and ensure that it’s up to date with the final release of Rails 2.2. Some people wondered what the “Microsoft Developers” part means. Is it for people that work at Microsoft? Is it for .NET programmers? Is it for people who develop on Windows?
The truth is that “Microsoft Developers” is probably just a marketing term that Wrox selected as a catch-all for of the aforementioned categories of programmers. As an author I’m trying to serve all of them well, by providing a guide that sneaks in much of the Rails culture and softens the migration path by using an Operating System, and to a certain extent, tools that they’re already familiar with. In my opinion one of the major obstacles when switching to, or trying, Rails when coming from the Microsoft world, is the culture shock. The documentation and most books assume that you are familiar with *nix systems and tools, and this can be frustrating for those who are forced not only to learn a new language and framework, but also an entirely new set of tools. As it’s targeted at Microsoft developers, the book obviously makes quite a few references and comparisons to the .NET world, where they fit. This is done so that the many .NET programmers amongst the group of so called “Microsoft Developers” will find the book particularly useful. Yet the book remains generic enough so that it can be used by any programmer (particularly Windows users), even those without any knowledge of the Microsoft .NET Framework or ASP.NET.
Python books: While on the subject of books, I wanted to mention that the final version of the Pylons book is available online. Despite the much less fancy UI, the book pretty much does what the Django Book did in the past. And both are available in print as well (The Definitive Guide to Django: Web Development Done Right and The Definitive Guide to Pylons). Pylons is a Python web framework that can be viewed as a Ruby on Rails clone, in a far greater way than Django could ever be considered.
Another thing I want to mention is that I received a copy of Expert Python Programming. I haven’t gotten to far into it yet, but from what I’ve seen so far, things look good. I hope to be able to read it through, over a weekend in the near future and then provide a proper review. Stay tuned.
DB2: This interview shows a few good reasons why even smaller and medium sized companies are increasingly adopting DB2. And while the video doesn’t mention it, IBM is coming out with an updated version of DB2 Express-C 9.5. This new version, 9.5.2 or 9.5 FixPack 2, is going to introduce exciting new features, including an engine for full text search.
The Great Ruby Shootout These days you hear a lot of talk about parallel programming. Intel promotes it and despite their bias, it’s plausible that parallel programming will become important as the CPU market heads towards an increasingly larger number of cores, as opposed to focusing on the frequency of said CPUs. In the world of Ruby, this translates into multiprocessing, as opposed to multithreading due to the infamous GIL (Global Interpreter Lock). This means that Ruby will most likely approach the problem similarly to how Python 2.6 did with the multiprocessing module, which is a process-based interface. The obvious exceptions are JRuby and IronRuby, which establish a 1 to 1 relationship between green threads and OS threads.
For the shootout, it would be interesting to see some multithreaded code, so as to get a better sense of how well JRuby and IronRuby compare to MRI and 1.9, when more cores are available. In fact, the long-promised shootout will be performed on a quad-core machine with 8GB of RAM. If Charles Nutter, John Lam, or any of their team members would like to contribute some programs that are able to take advantage of “native” multithreading, I’d be very happy to include them in the Ruby Benchmark Suite, to be used for my shootout.
The repository requires some love and refactoring, since it needs to be split in two types of benchmarks. The simpler one will evaluate the execution time minus the startup time, while the more advanced benchmark will also exclude the time required for parsing and loading modules, classes and methods in the AST. It would also be nice to test each program with variable input sizes and report these results accordingly. Right now I’m very busy with the book, but as I become more available, I’ll start working on this.
Finally, I want to point out a very interesting article about performance and UIs. Slow is indeed a very relative concept, and it’s important to understand how to analyze and respond to the user requirements when it comes to the responsiveness of an application as a user interacts with it.
Hardware: I finally bought a Trackball made by Logitech and the Microsoft Ergonomic Keyboard (Microsoft makes great hardware). I don’t have wrist problems, but I’d like to see how these two affect my extensive computer usage. I plan to report my experience as soon as I’ve had a chance to use these input devices for a while, since I know this is a topic that interests lots programmers (many of whom end up being victims of RSI, and some of the IRS :-P). I also bought a bad-ass color laser printer which is quite handy when you’re a programmer and you are writing a book. I’ll let you know how it goes. What I didn’t buy, but still think is awesome, is the Flip minoHD. It’s the equivalent of an iPod for the world of camcorders. $235 for a camcorder that’s so perfectly compact, and yet that can record in HD, is a pretty sweet deal. I’m considering it for Christmas, assuming it reaches Canada by then.