Just like most other programming projects, it easy to midjudge the time it takes to finish a web development project.

For one thing, when things get going, you always want to add little things here and there — even if you are not in charge of making the decisions.

For a second reason, testing can take longer because you have to test it for various environments (I.E. on Windows, Firefox on Windows or Macintosh or Linux, Safari on a Macintosh, or various other combinations of browsers and operating systems).

Probably the last and most important thing is that it takes practice and experience to estimate.

I’m getting pretty good at estimations. I pick a number that I think I could do it in if nothing went wrong. Then I double that number. And if the project is any size at all, I double it yet again. If it involves Flash or Javascript, it might be a good thing to double it again.

The interesting thing is that I rarely end up overestimating a project, even after all the doubling. The good thing is that I rarely go over by much, either.

• Perl 6 sounds promising by David Hickenbotham
• PHP Article Commentary by David Hickenbotham
• Lavaball introduces Design, Development & SEO Tips

This entry was posted on Thursday, June 9th, 2005 at 9:03 pm and is filed under Development Tips. You can follow any responses to this entry through the RSS 2.0 feed. Responses are currently closed, but you can trackback from your own site.