Tuesday, December 28, 2010

My list of 10 most influential programming books

Here is a list of top ten programming books that influenced my software development life, in the order of publishing date:

Peopleware: Productive Projects and Teams
Lots of companies still try to reinvent the wheel when it comes to managing software teams. This book defines all the best practices. The authors even went as far as recommending the cubicle size.

2. Philip & Alex's Guide to Web Publishing
This book was on the list of recommended readings for the ArsDigita bootcamp. I was for the first time learning how to program a dynamic web site.

3. Refactoring: Improving the Design of Existing Code
I would not go as far as
Steve Yegge and say that after reading the book I’m awash with a horrid cold feeling, as if I’ve just learned I’ve been coming to work for years with my pants down around my ankles. However, it did make a big impact. For example, after reading this book I started putting less emphasis on comments and more on smaller methods with descriptive names. It also made me start actually using the refactoring features in my IDE.

4. The Pragmatic Programmer: From Journeyman to Master
A true classic. Any further comment would be superfluous.

5. Effective Java
If there is a reference on correct usage of the Java language and its API, then it is this book. It could have easily been called Java Tips and Tricks or Java Gotchas. One of my favorite items:
Always override hashCode when you override equals.

6. J2EE development without EJB
It was pretty interesting to read this book after working for a company that had an EJB developer job title and realize the whole world doesn't evolve around EJBs.

7. Head First Design Patterns
I had tried reading the
Gang of Four book several times, without much success. I found it great as a reference book, but not quite suitable for learning design patterns. Then came the Head First Series one and finally shed some light.

8. Java concurrency in practice
Another classic. I ultimately understood the Java Memory Model after reading this book.

9. Programming in Scala
Of course, a Scala book had to make this list, and this is still the most comprehensive one.

10. The Passionate Programmer: Creating a Remarkable Career in Software Development
I read the first edition of this book(
My Job Went to India (And All I Got Was This Lousy Book): 52 Ways to Save Your Job) because I was intrigued by its title, even though the author states it painted the wrong picture. The book is so motivational that I promised myself I will read it every 6 months. In addition to that, the idea for this post came from this very book.

Sunday, October 11, 2009

Organizing Boston Area Scala Enthusiasts

Photo by Pear Biter

I like the concept of a local user group. It serves multiple purposes. As I embarked on learning Scala recently I wanted to join a local Scala user group. To my surprise, there weren't any in Boston area. However, there were already such groups in San Francisco, Chicago, New York and Boulder, just to name a few. I found it shameful.

After all, I thought, we are the 2nd ranked startup hub and overall one of the biggest software hubs in the U.S. Therefore, there have got to be local software professionals interested in all the new stuff. But then, even the local Groovy/Grails group got formed rather late (May 2009). I couldn't explain why Boston seems to lag when it comes to forming groups around latest technologies.

Instead of sitting in despair, I decided to take some initiative and see if it was just simply lack of interest or something else. I was happy to find out it was the latter. Turns out, there were couple of people thinking about starting the group for a while, but they needed that extra little push to actively start working on it. As a result, Paul Chiusano, Chris Herron  and I officially formed the group on October 1st, 2009. Ten days later we had 28 members - not bad at all. Even more exciting, I found out there are at least couple of local companies developing software in Scala, an encouraging sign.

So, if you're from the area and want to share your Scala enthusiasm with others, please join us.

Thursday, July 23, 2009

Second JVM language

So far I've been nearly oblivious to the learn-one-new-programming-language-every-year motto from the Pragmatic Programmer book. I did play to some extent with Groovy and Ruby, but that was only enough to get me through Grails & Rails. With constant releases of so many new technologies and frameworks in the Java world there is hardly any time left for anything else to focus on.

Still, my recent attendance of the Java One conference made me realize it's high time to start applying that advice. There are multiple reasons which all contributed to my decision. First is that I want to learn functional programming. Second is that it seems very attractive and concise to write code in a language that supports closures. And last, but not least, is that I would like to be able to leverage other concurrency paradigms than the Shared State offered by Java.

With all this in mind, the choice of the language was pretty simple: Scala. It satisfies all of the requirements, plus it's not a big leap from Java at all.

When you decide to learn a new programming language you are to some extent making a bet, as you hope that the newly learned language will be in demand some day. From that perspective it is encouraging to learn that James Gosling would use Scala as a second language on JVM and that speculations started around Scala replacing Java in a long run. On the other hand, if you take a look at the TIOBE index, at the time of writing this post Scala holds 32. position with only 0.228% of worldwide programmers actually using it. And if you search for Scala jobs currently you would find only a few across the whole US. However, I believe the bet I am making in this case is rather safe because even if Scala ends up being just a fad, I should benefit immensely from getting exposed to new programming and concurrency paradigms.

So, for a start, I got the "Programming in Scala" book and started a journey. Stay tuned...

Sunday, June 21, 2009

JavaOne 2009 impressions

This year was the first time I attended the JavaOne conference. Previously I would focus on local events, such as NFJS and SD Best Practices. This time though I followed advice of several JavaOne alumni who convinced me it's quite an experience and worth the trip.

It was not only my first time at the conference, but also my first visit to San Francisco and West Coast. Coming from Boston and heading to California I expected San Francisco to be much warmer in June. Instead, with the exception of couple of sunny intervals, I had to wear long sleeves at all times. For some reason it felt chilly even in conference rooms.

The conference overall was very well organized. However, food, especially lunch, was a slight disappointment. Although I haven't attended conferences of this size before, I was expecting some more choices and warm food instead of lunch packages. On the other hand, the fact we were given a hand sanitizer during registration, as well as multiple of them positioned throughout the conference area was a pleasant surprise, especially since swine flu was still a bit of a concern.

This year Sun was really pushing hard for JavaFX technology. It was well covered during general
sessions and had quite a few specific sessions dedicated to it. Moreover, its demos looked cool and it finally makes the "Java = everywhere" mantra true. The prevailing feeling I received from general sessions was that Sun was trying to assure us all that Java as a language is far from getting retired. In order to boost our confidence in Java's strength, the sessions were full of showcases of successful companies which implemented their whole technology stack in Java. Word "innovation" got mentioned a lot, but even new features in SE 7 and EE 6 just look like a result of catching up with other frameworks from Java world. Especially with closures out of the picture for version 7, Java will still be much more verbose and less expressive than majority of the alternative languages targeted for JVM. Technical sessions which compared Java to those alternative languages by presenting language specific solutions to same problems, were an actual proof of this.

The most innovative piece of technology I learned about at the conference was
. Monty Taylor gave quite an entertaining talk about Drizzle's history, objectives and features. It sounds so compelling that MySQL will probably in future be used just for legacy reasons, at least in the web application arena. It will also be interesting to see the direction of this project after Oracle acquisition of Sun is completed.

Speaking of technical sessions, I truly enjoyed presentations from regular JavaOne Rock Stars:
Soctt Davis's talk on Resource Oriented Architectures, Brian Goetz's and Cliff Click's talk about VM modern architectures, Rod Johnson's talk about new features in Spring 3.0 and Josh Block's talk on Effective Java. In addition, I was pleasantly surprised and really learned something out of presentations from Bill Venners on Scala, Jonas Boner on JVM Alternative Concurrency Paradigms and Alex Miller on Java Concurrency Gotchas. I highly recommend seeing all of those.

I promissed myself this was going to be the last conference where I would select sessions
primarily based on the topic. Namely, in several sessions into which I was greatly interested the presentations weren't so good and they barely increased my knowledge on the topic, if at all. I can't comprehend how some presenters can just read through slides or do an immediate deep dive without first giving an introductory overview. Jared Richardson, for example, in his Career 2.0 presentation talks about this and nicely gives tips on preparing a successful presentation. Furthermore, some speakers apparently find it sufficient to just place code snippets on slides and expect the audience to quickly digest it. All it takes to see how to properly do it is to attend one of the Josh Block's talks and pay attention how he slowly explains every line of code, no matter how trivial it looks like. So, next time, I am firmly determined I will build my schedule entirely based on speaker reputation. I wish I knew about the JavaOne Rock Star Wall of Fame before the conference. Even if I end up at let's say a UI session (which is not my area of expertise), I am sure I will learn something new since an interesting speaker will most probably attract my interest to the subject.

I was also kind of expecting more prominent presence from Google. I would have rather seen them
talking in general sessions than Microsoft or IBM. Of course, they had their own developer conference (Google I/O) just prior to JavaOne, but they are also one of the biggest innovators in Java space. The other problem is the fact they are still fairly secretive. I did attend one of their BOFs and right in the beginning the answers from Google team started along the lines of "We can't really talk about that".

Overall, I find the biggest advantage of JavaOne over smaller conferences the ability to hear more first class speakers who are leaders in their field (language founders, framework authors, etc...). And finally, attending developer conferences in general brings many benefits: you get to see the latest trends, take a peek into future technologies and discover some exciting existing ones for the first time...

Sunday, April 12, 2009

Computer setup

After thinking a while what to write in my first blog post I decided to talk about my computer setup at work. I often discuss this topic with other software developers, so I thought it would be nice to have my own configuration described on the web.

My setup consists of two computers, a desktop and a laptop. Apparently this is a standard combination in some software companies. The desktop is running Linux and that's where I do most of my development work. It typically has only a few applications open at any time: IDE (IntelliJ IDEA, for those curious enough), terminal and PDF viewer. I try to run as many processes as possible within the IDE, including an application server and use terminal mostly for remote sessions.

The laptop, on the other hand, runs Windows XP. To set things straight, I am not a huge Windows fan, mainly because of the poor command line. However, I really like some of Microsoft’s software, such as Outlook, which I find to be the best e-mail client out there (after experimenting with numerous other clients) and Office 2007 (especially Word and Visio). Additionally, I keep Firefox open on Windows as well, as it runs faster than on Linux.

So, now that I've covered which app runs on which machine, here is the best part: I only use a single keyboard and mouse and there is no KVM switch. This is thanks to an awesome piece of software called Synergy. It works great with the Windows-Linux combination and as of now the Mac OS support is not yet fully complete. This is a partial answer for all those who wonder why I use Linux as a desktop OS, when there is OS X. The other reason, of course, is that a Linux box is much cheaper than a comparable Mac one. Don't get me wrong, I use Mac OS at home and believe there is no better desktop OS, but in my case where the majority of my work is done inside the IDE, it doesn’t make that big of a difference. Again, the lack of proper Synergy port is a deciding factor here.

And finally we come to the last piece of the puzzle: monitors. There are total of four: the desktop is hooked up to a widescreen 22" and a 19" one. The laptop also has a 19" external monitor attached to it, in addition to its own widescreen 15.4" display.

These four screens is something for which I get most of the remarks. Some are nice comments and some are teases (“So, what’s the deal? Do you get paid by a monitor?”). My wife thinks I am a snob because of them (“How can the rest of us perform all our work with only one or two?”). Some people ask me why bother with multiple monitors when large ones became moderately affordable recently. Although I've never had a chance to work with a truly large monitor before, I think the biggest advantage of working with multiple regular ones is that you are free of window management. Namely, all the apps on any of my monitors are usually maximized and I never re-position them on the screen or move them around.

Now, let me provide some more details regarding specific application to screen allocation. The smallest screen - the laptop display, has Outlook open so that I could glance at any moment over new email (even though I would personally like to check my email less often, our development team is distributed and email is unfortunately used as a primary communication tool). Second screen is reserved for Firefox, since web is used as a reference in everyday work. Furthermore, the external widescreen monitor is a best fit for the IDE and the last fourth one is interchangeably used for terminal windows and documentation viewing.

Overall, I believe the above described setup allows me least amount of window switching and thus highest degree of productivity. To be able to work on the code, browse through APIs and read documentation at the same time as well as have visibility into e-mail inbox is my dream come true. Here is a photo, to illustrate how it all looks like: