The Source for Java Technology Collaboration
User: Password:



John O'Conner's Blog

December 2005 Archives


Contributing to Mustang: A submitted bug fix

Posted by joconner on December 22, 2005 at 03:58 PM | Permalink | Comments (8)

After so long anticipating this moment, after the deed is done, I feel like it was too easy. I'm definitely experiencing an anticlimax.

I submitted the following from the "Submit a fix" button located on the collaborator page. Remember, you have to follow the steps to get contributor/collaborator status. Here's my submission for the javadoc bug for JTextComponent's getDisabledTextColor method:

Source: j2se/src/share/classes/javax/swing/text/JTextComponent.java
Release: Mustang b64
Bugid: 4994762
Diff:
--- JTextComponent.java 2005-12-22 14:42:53.542200000 -0800
+++ JTextComponent.java.new     2005-12-22 14:42:33.012680000 -0800
@@ -1300,7 +1300,7 @@

     /**
      * Fetches the current color used to render the
-     * selected text.
+     * disabled text.
      *
      * @return the color
      */

The fix, although small, for bug id #4994762 is on its way to Sun engineers. I can see it now, a small record in a database somewhere just waiting for someone to find it tomorrow morning. Or maybe...I can only hope...the bug fix submission triggered loud clanging bells that sent several engineers scurrying back to their offices after playing Playstation 2 games in a shared lounge area. Hey, a guy can dream, right?

There is actually a reason why I've picked this seemingly insignificant bug. Here's my thought process:

  1. Any really big, important bugs will be fixed by engineers at Sun...or at one of their licensees like IBM.
  2. An invitation to the world to contribute is going to potentially get lots of responses, probably a lot of little fixes that just happen to be important to someone.
  3. There must be a cutoff somewhere, right? Where's the lower limit on what they'll accept? Will it be my insignificant fix, even though it is potentially really important to me?

In the end, I wanted to find something small and easy, and I did. However, the big conclusion I'm coming to is this: whatever you fix will be important to you...will your contribution be important enough to Sun?

Now that is the purpose of this experiment...to test this process, to determine whether it works, and to find out whether a normal, everyday developer can get something fixed if he's willing to fix it himself? Stick around...this blog thread isn't over. I just submitted the fix. Now, let's find out what happens to it.



Contributing to Mustang: Low hanging fruit

Posted by joconner on December 22, 2005 at 02:01 AM | Permalink | Comments (8)

After a brief rest from this public Mustang community experiment, I'm back. And I've found the perfect bug...it's easy to fix, very low hanging fruit that's just right for a new contributor to Mustang. What is it? It's bug/change request id #4994762, and I'm all over it.

Bug #4994762 is a javadoc bug, priority 4 (low). Not too important as far as bugs go, but I'll have this fixed in no time. Why did I pick this? Well, it was easy to fix, and the type of bug isn't particularly relevant to this experiment. The goal is to test the contribution process, not my bug fixing skills. Besides, this is exactly the kind of bug that beginning contributors are encouraged to tackle. Once I get some respect around here, I'll move on to something bigger and better...a priority 1 or 2 bug no doubt. Until then, what's my bug? JTextComponent's javadoc for getDisabledTextColor is incorrect.

The incorrect docs say this:

Fetches the current color used to render the selected text.

And it should say this:

Fetches the current color used to render the disabled text.

OK, I've got my bug; I've got the fix. Now what? Join me tomorrow after I find out how to submit this thing. Until then, you can find your own bug to fix from the online bug database.



Character Conversions from Browser to Database

Posted by joconner on December 20, 2005 at 04:07 PM | Permalink | Comments (1)

It's relatively easy to enter multilingual text into a rich client, Swing application. Doing so in a browser-based, web application isn't always as easy. If you're interested in how to avoid data loss as characters move from browsers to database, you may want to take a look at this recent Java EE tech tip: Charset Conversions from Browser to Database.

As always, your feedback and suggestions for new articles and material are always welcome.





Powered by
Movable Type 3.01D
 Feed java.net RSS Feeds