March 30, 2006
Listened to the third episode of RedMonk Radio on my way in today. My favorite bit:
Coté: There's the evolution of, y'know, way back when the military radar system drove technology, right, and then there was NASA, and then it went into private enterprise, and now it's kind of the consumer space that seems to drive everything.
James: Oh, Coté, you're my man, I'm so glad you're saying that. Let's create games platforms to drive technology instead of war to drive technology. I am all about that. Consumer-driven innovation is the better kind. Let's stop funding all the military stuff and fund fun that will drive technology.
Right on.
March 28, 2006
Rubyists talk about methods calls as sending message to objects, a notion that I understand is inherited from Smalltalk. It always seemed a little strange, but I could accept it as an odd quirk ("Bah! I understand method calls!") and move on to writing code. This is unusual behavior for me, since normally I try to understand a language on its own terms. Now, though, I think I finally grok all this talk of messages, thanks to A Little Ruby, A Lot of Objects, an incomplete book in the style of The Little Schemer. And as it turns out, terminology matters.
class Dog
def speak
"Woof!"
end
end
fido = Dog.new
Calling the speak method on fido (fido.speak) is really sending a message to the fido object. When fido receives the "speak" message, it knows to invoke the Dog class's speak method. It answers, or returns, with the String "Woof!"
Yeah, that doesn't seem too special, does it?
But here's the thing. In Ruby, everything's an object. 2 is an object of type Integer. 2 + 5 in Ruby is really saying, "Send the message "+" to the object 2, along with the parameter 5." That is, 2 + 5 is convenient shorthand, syntactic sugar for 2.send("+",5) (which is, by the way, valid Ruby). 2 receives the message and invokes its + method, passing 5 as a parameter. It responds with 7.
Okay, this is still not seeming very special. Doesn't this just muck up the discussion? Sure, it's neat that 2 is an object, but what do you gain from all this talk of sending messages?
I'll tell you where it finally clicked for me: polymorphism. I'm not sure why, but for some reason polymorphism is often a tricky concept for beginning programmers. Couched in terms of sending messages to objects, it's crystal clear (at least to me :). You can send the same message to different objects, each of which will invoke the appropriate method of its class.
class Cat
def speak
"Meow!"
end
end
garfield = Cat.new
With both fido.speak and garfield.speak, we're sending the message "speak" to the object. fido receives the message and invokes its speak method, and garfield invokes its own speak.
+ does different things for Strings and Integers. This is commonly called operator overloading.
2 + 5 == 7
"Abe " + "Lincoln" == "Abe Lincoln"
The message is the same: "+" — but the methods are different. Ah ha!
And now I finally understand duck typing in relation to polymorphism. It should have been obvious earlier, but I never thought of it in that way (probably because I figured that polymorphism wasn't that tricky, so I didn't think about it). Dave Thomas's explanation of duck typing is fairly clear. In Ruby, types are defined by what objects can do rather than their class.
When I write
fred << "dog"in Ruby, I don't care whether fred is a String, a File, or an Array
What matters instead is whether fred can respond to the message "<<".
(If you're a programmer who only knows Java or C#, by now you're probably freaking out. That's okay. It gets easier.)
So talking about method calls in terms of sending messages to objects starts to make sense. I get a better (or at least different) sense of what's happening behind the scenes, and what otherwise can be abstract concepts become a bit more concrete and useful. I'm probably still missing something, but I feel closer.
If you missed the implicit recommendation of A Little Ruby, A Lot of Objects, let me make it clear now: it is a fine read, highly recommended. And now I want to go read The Little Schemer.
March 22, 2006
Kiara's getting her Master of Fine Arts in Creative Writing from Hamline this spring, and will be reading from her theses this Friday. She blogs about that over at Wordspinning.
Yes, theses. Maybe it's just considered one thesis, but she's done thesis-length works in both poetry and prose. That's Kiara for you: why do anything half-assed when you can do twice the amount of top-notch work?
It sounds like she'll be reading "Mary's Baby," which is a fine, fine short story embedded in her novel. I'm excited to see how she reads that, as the narrative voices in the story are so strong and so distinct.
March 20, 2006
The JavaScript shell is nice, but if you're using OS X there's a better option. On a tip from Todd Ditchendorf, I'm now running both a JavaScript and a Ruby interpreter in Dashboard, just a keystroke away.
It's easy. Todd gives directions for installing WidgetTerm, which is key to the whole thing. Then drop a jar from the Rhino project in your Java classpath, set an alias to a JavaScript shell, fire up a WidgetTerm instance, and you're done. It sounds more complicated than it is.
I need to pay more attention to Rhino. It caught my attention when I read this discussion of choosing a Java scripting language and saw how Rhino stacked up against JRuby, Groovy, Jython, and others. Here's an obvious application.
March 17, 2006
No Fluff Just Stuff is in town this weekend and I'm missing it. Sigh. I have to make a more concerted effort to attend this sort of event.
March 15, 2006
My RSS feed is on Feedburner now, so I am redirecting requests for the RSS there. If you use an aggregator, it should follow the redirect, but you might want to update to the new URL anyway.
Why Feedburner? Two reasons to start:
It will be nice to see some stats on the feed use, too, something I occasionally generate with a series of Perl scripts but don't follow closely.
March 14, 2006
"So," Tim asked, "still down on Java?"
When we had last spoken, I was reading Beyond Java, so that was an understandable question. I've had the conversation an awful lot lately and people keep being surprised by whatI say, so I figure it's about time I write something here to explain my position.
Am I down on Java? Not per se. Java is not a language I generally enjoy working with, but I write web apps for a living. I do not believe that Java is an ideal choice for the work I do — I don't even think it's a good one. There are rays of hope: RIFE looks to be especially promising (hell, even exciting!) for making Java web apps far less painful than the norm. But by and large, the way I believe most people write Java web apps makes it a very unpleasant platform to work with. Unpleasant in the sense of slow and ineffective, with far too much code.
I've also been turned off by unbearably slow development cycles, which I cannot lay entirely at Java's feet. Other factors like politics and methodology contribute. Still, in most people's hands it takes longer to write something in Java, becaCan a team of experienced, solid developers work fast and well with Java? Of course. But most development teams are not filled with that kind of programmer.
If it works for you, great. I don't think it's worth anyone's while to get into pointless language debates. But personally, I would rather use something other than Java for web apps, especially the whole JEE stack. My experience with J2EE has been barely manageable complexity in both development and operations management, which stands a good chance of making apps more difficult to maintain and secure. It is a good fit for some applications, when JEE serves to address and alleviate complexity in the problem space by providing simple solutions. Assuming that happens. :) I just don't see web app control flow and presentation as being a hugely complex problem until Java enters the picture.
So what's the alternative? Oh, there are plenty. Could or should Java play a role in our development? Of course. There's often a reluctance to involve more than one language in an application stack to keep things "simple," which I think is short-sighted. I often like to use more than one language for complex apps because it forces me to keep the design and APIs clean, a collection of loosely coupled services. Sometimes I like to use a full-stack framework, too. It depends. Use the best tools for the job. Leverage the strengths of the languages in your arsenal. Java fits well on the far back end, integrating legacy systems and managing interplanetary navigation. Closer to the front end of a web app, I'd use something a bit more suited to web development, like PHP. In the end, I think it would be faster to develop, easier and cheaper to maintain, and make developers happier. I try to be pragmatic and happy in my work.
March 14, 2006
On the latest Pro::PHP Podcast, Marcus Whitney interviewed Microsoft's Brian Goldfarb and Joe Stagner about PHP on Windows.
PHP developers may have already heard this interview (if not, why not? It's a good podcast!). I wanted to be sure to mention it here because just as I think it's useful and important for PHP programmers to separate the language from the LAMP platform and get away from knee-jerk reactions against anything having to do with "M$", I believe that .NET developers should hear discussion from within Microsoft of non-Microsoft languages on their platform and how that might be a Good Thing. Good stuff.
I haven't talked or written much about it, but I like a lot of what I see in .NET. Microsoft has done a far, far better job than Sun (or Oracle, or IBM...) at marketing to developers. I'm still not about to take much time to play around with .NET, since I don't have a Windows box at home and have better things to do with my time at work, but it is compelling. I'll probably watch some of the webcasts at learn2asp.net, and who knows? Maybe I'll get Mono up and running on my Mac, although I think it's too early for .NET 2.0 on Mono at this point, right?
Of course, first comes more Ruby, more PHP 5, then Smalltalk, Lisp, Haskell, Erlang… oh, and raising two kids. Do you see my problem? :)
Go listen to the interview.
March 13, 2006
I spend far too much time explaining to my coworkers that no, Internet Explorer is not behaving correctly: that caching behavior is broken, that box model is not standard, that JavaScript behavior is unusual. Sometimes it's not even that IE is wrong, it just chooses a different path than every other damn browser we test. Yet the expectation I confront is that if IE behaves a certain way, it is right and everything else is broken, so we should just tell people to use IE. Aargh! And best yet, it doesn't matter that IE on my computer doesn't behave the way it does on theirs, because IE on their computer does what they expect so I must have a nonstandard configuration. Grrr…
So it is with great relish that I read Amy Hoy on troubleshooting Internet Explorer. I'm pinning that to my wall, you can be sure.
March 10, 2006
I've had Selenium on my radar for a while. Automating web app tests in a browser? Sounds great. FIT-style test definitions? Fantastic. Run it from JUnit? Wow! But I was always "too busy" and never got anywhere with it until I discovered Selenium IDE, a Firefox extension that can create and run Selenium tests. I watched this screencast, installed Selenium IDE, and was up and running in no time. I highly recommend it.
It immediately proved useful, as I discovered an intermittent bug in one of our web apps. Selenium IDE recorded my session on the web app, including a check for certain text that indicated the bug. I saved the test definition as an HTML file and attached it to the bug report. After I walked the developer through installing Selenium IDE and loading my test, she could run through it several times to finally trigger the bug. Bingo! I don't think I'll need more than that to sell the tool to the rest of my team.
Now I need to dig deeper and see what more Selenium can do for me. Also, take a look at Selenium on Rails.
March 09, 2006
I've been using RadRails for Ruby on Rails development at work. It's a Rails IDE built on Eclipse, available as a standalone (which they recommend) or as a plugin. We already use Eclipse for Java development, so it's an easy step for me to make. And it does make a few things easier than, say, JEdit. ActiveState's Komodo now sports Ruby and Rails features, but its project file handling annoyed me so I quickly gave it up. TextMate is definitely my editor of choice on the Mac, but on Windows I'll happily stick with RadRails for Ruby and Rails projects.
They've just posted a screencast that demonstrates RadRails features. It does about what you'd expect with regard to editing: RDT for Ruby editing, and an RHTML editor that, while not perfect (no tag completion yet), gets the job done and has enough syntax highlighting to make my life easier. If all I wanted was a text editor, though, I'd stick with a text editor and avoid the Eclipse overhead. What keeps me with RadRails is the easy stopping and starting of a WEBrick web server, generators, tailing the development log, and the simple database browser. This is all available through a terminal window or other apps, yes, but sometimes it's nice to have everything in one place. The raison d'être of an IDE.
Strangely, this isn't how I work on a Mac, where I feel more at home with small, well-defined apps that each do one thing well. On Windows, though, I'm more at home with an IDE. Odd.
Ah, who am I kidding. I ought to just cave and start using Emacs.
One more nicety: the developers podcast their releases, discussing what's new in the release and what's coming up. I enjoy these. They're easier to digest than just scanning through a changelog.
March 08, 2006
"Do you really read all those goddamned blogs?!"
My brother asked me that after he saw a couple hundred feeds in my blogroll. "Sure," I wrote back, "it's not as bad as it seems."
That was 200 feeds ago.
Now that I've tipped over the 400 mark, I've had to make some changes. It happened when Alec was born and I took three weeks leave from work. I decided to stay away from the computer for a while, checking email only occasionally and not opening my feedreader at all.
It felt good. Really good.
I fell back on my old habits, though, for several weeks, feeling like I was more and more behind the times, until I caught Merlin Mann's interview on Inside the Net. Listening to Merlin talk about email management, I was feeling pretty good about my inbox. I'm brutal with it. I open email just a few times a day, and clear out the inbox completely. To-do items end up in Remember The Milk, some email gets filed for reference, and an awful lot gets deleted. It feels great.
Then I opened up Feedlounge, which I was test-driving as my new web-based aggregator, and was crushed by the 3000+ unread items. Then and there I decided to simplify. I cancelled Feedlounge and resolved to stop using web-based aggregators entirely. Now I read only at home on my Mac, checking in on just a handful of core feeds every day or two. I still have the full feed list in NetNewsWire and scan through it occasionally, but feel no compunction whatsoever about marking everything read. Very liberating.
How very sad that this feels liberating. But it does.
And yes, NetNewsWire. My friend Jim suggested Vienna, which I used for a while but had to abandon because it was choking on one of my core feeds (Stephen O'Grady's). NetNewsWire does what I need, perhaps because it's very good at what it does.
