Skip to main content

Mayhem roundup

Posted by evanx on June 8, 2006 at 6:13 AM PDT


"All my life I've had one dream: to achieve my many goals" Homer Simpson

I've been meaning to start blogging on for a while. To promote myself and my project, and because i really love writing. I go into a whole new world, into a trance with a silly grin on my face. Same as coding.

empty_audience_200.jpg align=left hspace=16 />
This new era of an interactive web, this so-called Web 2.0 thing, is something special. In the past, you read an article, and then moved on. Now you can post a comment, and engage other readers, and the author. And blog yourself. This is something special. We can participate, and interact.
We can do what human beings were made to do!

Typically i mostly listen. This comes from when i was a stutterer at school. You were not gonna catch me opening my mouth if i could possibly help it! One day that all changed, and i starting giving big public talks at every opportunity. Anyway, that's probably why i like writing so much, because it's always been a safe, comfortable environment to express oneself without any risk of stuttering and stammering!

So i'm between contracts. Supposedly working on my project fulltime.
But then The Java Posse rode into my iTunes...

Explicit Reflection

"Internet! Is that thing still around? Oh, so they have internet on computers now!" Homer Simpson

posse3.jpg align=right vspace=16 hspace=16 />
When listening to Java Posse in March one morning, working on my project, I emailed in to "If I was the king of Java..." about "explicit reflection." It was something i had wished there was in the language, because it would have helped me.

I was so excited about my email being lasso'ed by The Posse, that i decided to write my first blog to expand upon the issue.

This started a theme pursued in subsequent articles, like Refactoring Translations and Bean Curds 1 and 2, that string literal references are unwelcome.

Netbeans_125.png align=left vspace=16 hspace=16 />

Netbeans Day (South Africa) Trilogy

"Remember as far as anyone knows, we're a nice normal family." Marge Simpson

As it happened, Netbeans Day was taking place that week in South Africa, with James Gosling in attendance.

During the presentation, I decided to take notes of the interesting points. Out of habit. Not because I ever go back and read my notes, but it just feels good to "Save As."

When i got home, I found myself writing a summary of the talks, and those notes came in handy!

href="">Part 1 summarised Geertjan Wielenga intro on Netbeans 5. Preaching to the converted in my case! But there were some very handy tips for me, eg. F2 for bookmarks, "Camel Case Completion" eg. WL for WindowListener, Alt Shift W to enclose with try/catch, and

Geertjan left a nice comment, and also Chuk Munn Lee (who also presented on the day). I was chuffed!

To quote myself from Part 2

Personally i love the idea of wizards, and sample projects, with integrated documentation, rather than traditional static documentation, like a PDF. I call this "active documentation." Netbeans 5.0 JEE Blueprints are a great example.

and regarding the RCP

I have a happy suspicion that we'll see some great Netbeans RCP apps (and IDE extensions) popping up in future. Personally I'd love to see a whole desktop suite built on RCP - mail client, tabbed browser, and file browser. Then we'll really have a consistent, integrated, and extensible Java desktop!

Part 3 summarises Sang Shin's talk on Netbeans for JEE. I dig his master tutorial index on :)

In part 3, I got gushy...

The golden age is dawning, with fantastic languages, libraries, tools, databases, servers, frameworks, components - all freely available, and cross-platform - with tutorials to rule them all - and with Netbeans, it's all in one little download.

Roman Strobl commented on Derby integration.
I was chuffed to find that anyone was reading my blog entry, but Roman Strobl! I have really appreciated his video tutorials, and blogs and podcasts, in the past, so that was great.

I ended the article wishing for a Netbeans weblogger plugin, and
Gregg Sporar
pointed me to

Geertjan left a comment that he was gonna link to my summaries of Netbeans Day (South Africa) from his blog. Wow, this blogging thing was really working for me! You start to feel like you are inside this whole community, and not outside looking in, and it feels great!

thinkfreeCalc175.png align=left vspace=16 hspace=16 />

Swing trumps... everything!

"Burns releases the hounds on every charity that comes to his door - Feed the Children, Save the Whales, even Release the Hounds."

I was writing a comment to Swing trumps Ajax.

I suggested that Sun will opensource Java because...

"Sun wants developers to be a field of Sunflowers, following the Sun."

Everyone has their favourite things, that they feel passionately about, and other people feel passionately about different things. I wanted to write passionately about my favourite things. At the same time, one doesn't want to offend and upset people. So it's a tight rope. I gotta respect other views that I might disagree strongly with. Not least because they might be right. Certainly they are right, in a situation and perspective other than my own.

I learnt a few tricks at navigating the tight rope. For example, I said, "I think that GNOME is gonna overtake KDE." GNOME people are happy because I'm saying GNOME is best, and KDE people are happy because I'm saying KDE is best. Whereas if i had said, "GNOME is better than KDE" or visa versa, well, the "comments" section would have got messy. Which fortunately it didn't! :)

Lemme paraphrase some thoughts presented there.

My dream for the future of computing, is stateless rich client applications using web data services, where we cache in on rich applications via the web, like Thunderbird, OpenOffice, and of course new Swing and .Net ones too, via WebStart, Google Pack, or whatever, and where your documents and settings follow you around on your GDrive, LiveDrive, Amazon S3, YahooDrive, SunGrid, or whatever.

Personally I think that Sun should reinvent itself as a Web 3.0 software service company in this way, and become the biggest
consumer of their own hardware.

Step one is to build SunGrid storage support into OpenOffice (and Firefox and Thunderbird). Why sell CPU time to a few organisations, when you can sell computing to the whole world?

The issue of Java/Swing performance came up in the comments, and i weighed in...

Swing_s.png align=right vspace=16 hspace=16 />

If i'm given the choice (and clients too) of delivering something in three months that's gonna need 1Gb of RAM to run, or the same thing in six months, at twice the cost, but that'll only need 512Mb to run, i think you can guess everyone's answer ;)

Considering that Java and C# are languages much like C/C++, any implementation performance issues are solvable engineering problems. This is evident in Hotspot and Mustang. Memory use will be higher (for garbage collection), but apps will be much easier to write and debug. Performance will be comparable, sometimes slightly better (eg. Glassfish's Grizzy using NIO), sometimes worse, than native C/C++ (or Python, or C#) implementations. When it's a lot worse, and that's a big problem, it becomes a priority to fix it, and it gets fixed. It just takes resources allocated - excuse the pun.

You point out that C# is newer and learnt from Java's mistakes - which ones bother you the most, and can't the community do something about them? An (dis)advantage of Java over C# is that there is a community and the JCP (of which Apache, IBM, Google, Oracle, et al, are no small part of) to evolve the language (and standards and libraries), rather than one company. Java 5 has evolved, in response to C# and thanks to the Java community - introducing enums, generics, varargs. Java 5 took away all my big itches.

The point is that every language (and tool and library) needs to evolve in response to new developments, and Java is doing that, and C# is doing that, and they are in healthy competition. In hindsight "mistakes" are made of course (even tho at the time there were the right decisions). When they are identified they should be fixed as soon as possible (eg. aggressive refactoring). Sometimes they can't be fixed right away because that will break backwards compatibility. So lets deprecate them and fix them in five years time. Yes, there are cases where in hindsight, we wish we had done things differently from word go, and would be better off if we had (eg. generics, altho at the time, that was too much effort, and other priorities were more important, probably). Which is why designs and APIs should be, first and foremost, blindingly simple and minimal. Then there's less opportunity to do wrong. And Java got that pretty much right.

Let's make Java (through the JCP), stand the test of time, and get better with age, so we don't have to throw everything out and start again with C#, or the next "best language." Lets make Dolphin etcetera, be the next "best languages."

But enough about you, let's talk about me!

"You couldn