Make bad code good

Refactor broken Java code for fun and profit

1 2 3 Page 3
Page 3 of 3

Moreover, a reverse-engineering tool would be useful for determining the architecture of the system. As implied above, I don't think the architecture of the system is anything to be concerned about when you start. After all, there's not much you can do about it. However, when you have the small details under control, you'll almost certainly need automated support to figure out the existing architecture. Look for the buzzword "round-trip engineering" on the Web.

Note: I won't make any specific recommendations for tools, as I have not used most of the sophisticated ones. I'm usually happy with find, grep, and a Python interpreter. I've cobbled together utilities to do most of the things I need.

Conclusion

If you're just starting out with bad code, you've got a long way to go. But don't fret, when the going gets rough, the tough get their chance to shine. You can write good code with a smug sense of superiority and a more than average amount of job security. Moreover, you can learn from the mistakes of others (but blame them when you mess up). You can run your unit tests and say, "It works now." And you can write on your resumé, "I can make bad code good!"

That's got to be worth paying for.

Acknowledgements

I would like to thank all the bad programmers in the world, without whom life would present no challenges. I would also like to thank all the good programmers in the world, without whom sympathetic shoulders to cry upon would not be found. Working on bad code can be frustrating, and kindred spirits will keep you sane.

Further, I would like to acknowledge the contribution of the Extreme Programming inventors, and all contributors on the WikiWikiWeb, without whom the evils of Big Design Up Front and the value of refactoring and unit tests would not be duly recognized.

I would like to thank Ian Clatworthy, Darryl Collins, Christian Larney, Simon McMullen, and Thomas Maslen for their comments on the first draft; they have made this a better article. I would also like to thank David O'Brien for his comments on my English -- it's amazing what you can learn from a professional.

Dr. John Farrell has been a Java programmer since the days when even programmers thought it would be mostly used for writing fancy Webpages. He currently maintains a distributed object server infrastructure that stretches from JDBC through CORBA to just below the GUI layer. It was formerly bad code, but he made it good. Before being a Java programmer, John completed a PhD in functional programming, which gave him an appreciation of elegant languages that run slowly; and worked as a C programmer, during which time he gained an appreciation of automatic garbage collection. He currently spreads the word that Java runs fast enough for real applications, if done right; and that good object-oriented design almost always represents the right way. On the second Monday of every month, he is an active heckler at the Australian Java User Group meetings in Brisbane, Australia.

This story, "Make bad code good" was originally published by JavaWorld.

Copyright © 2001 IDG Communications, Inc.

1 2 3 Page 3
Page 3 of 3