Has anyone actually tried out Extreme Programming in all its glory? Is it worth the effort?
This entry was posted
on Monday, December 8th, 2003 at 12:00 pm and is filed under Uncategorized.
You can follow any responses to this entry through the RSS 2.0 feed.
You can leave a response, or trackback from your own site.
3 Responses to “Extreme Programming: worth the candle?”
Well I just wrote an article for Linux Format where a pro XP bod and anti XP bod thrash out the concepts. It seems that people are quite divided in terms of practical implementation of XP. There is also the issue that geeks don’t seem to like pair programming as it is…wait for it…social.
Personally, I think these programming paradigms are of mixed use. Stick to design patterns and leave it at that.
Pair programming rocks, plain and simple. Do you ever find yourself procrastinating when you should be writing code? Do you compulsively check email, weblogs and fail to get any work done? Do you get stuck on one annoying little bug and get put off for hours on end? I know I do, but when I’m pair programming I can’t procrastinate, I can’t waste time with email and blogs and no bug is too tough to resist two minds working on it at once.
Unit tests are fantastic as well, but so far we’ve only managed to use them for things that have an input and an output. Most web related code ends up talking to a database, and we still haven’t figured out how to write unit tests for database interactions. The single biggest benefit of a good test suite is you can mercilessly rewrite code safe in the knowledge that any new bugs will almost certainly be caught out by your tests.
Those are the main XP things I’ve been using and they are both well worth the effort. Collective code ownership is pretty cool as well, but I can’t really speak for the others. A 40 hour week would probably work fine if I didn’t spend time at work reading blogs and checking email ;)
Simon, by “unit tests”, do you mean plain and simple unit tests or the so called test driven programming? TDD (of which unit tests are of course an important part) is the thing closely married to XP, and is a whole another ball-game.
Well I just wrote an article for Linux Format where a pro XP bod and anti XP bod thrash out the concepts. It seems that people are quite divided in terms of practical implementation of XP. There is also the issue that geeks don’t seem to like pair programming as it is…wait for it…social.
Personally, I think these programming paradigms are of mixed use. Stick to design patterns and leave it at that.
31 hours later
Pair programming rocks, plain and simple. Do you ever find yourself procrastinating when you should be writing code? Do you compulsively check email, weblogs and fail to get any work done? Do you get stuck on one annoying little bug and get put off for hours on end? I know I do, but when I’m pair programming I can’t procrastinate, I can’t waste time with email and blogs and no bug is too tough to resist two minds working on it at once.
Unit tests are fantastic as well, but so far we’ve only managed to use them for things that have an input and an output. Most web related code ends up talking to a database, and we still haven’t figured out how to write unit tests for database interactions. The single biggest benefit of a good test suite is you can mercilessly rewrite code safe in the knowledge that any new bugs will almost certainly be caught out by your tests.
Those are the main XP things I’ve been using and they are both well worth the effort. Collective code ownership is pretty cool as well, but I can’t really speak for the others. A 40 hour week would probably work fine if I didn’t spend time at work reading blogs and checking email ;)
2 days later
Simon, by “unit tests”, do you mean plain and simple unit tests or the so called test driven programming? TDD (of which unit tests are of course an important part) is the thing closely married to XP, and is a whole another ball-game.
As per usual, what comes to XP in general and in specific, I recommend reading the reports by Karlsruhe Empirical Informatics Research group.
3 days later