tag:blogger.com,1999:blog-9648422.post115072661093514764..comments2008-12-13T05:54:19.001+02:00Dan Buneahttp://www.blogger.com/profile/02360135573826474086noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-9648422.post-1151492697718509312006-06-28T14:04:00.000+03:002006-06-28T14:04:00.000+03:00Hi Stuart,<BR/> <BR/>I do agree that my experiment isn't scientific, and that my measurements are more or less ok. Since I only keep it as a blog entry, and not as a scrientific experiment, I do want the community , both those familiar and those not familiar with TDD, to tell me what needs to be adapted, redone etc and for that I have to thank you for your observations which are very good.<BR/> <BR/>I took 30 seconds for the simple function, because I said you would need more pairs of values to be inputed manually and the result checked (6 and 2, 1 and 2 and 8 and 0 which fails etc). For this particular function the time depends for instance whether the program needs to be restarted to check a new pair of values or not etc.<BR/><BR/>You are right that such things can backfire. What do you think the right times would be?<BR/><BR/>Thanks,Dan Buneahttp://www.blogger.com/profile/02360135573826474086noreply@blogger.comtag:blogger.com,1999:blog-9648422.post-1151483462946436682006-06-28T11:31:00.000+03:002006-06-28T11:31:00.000+03:00While I'm a big fan of TDD, I'm not sure you're measurements of time for this function are entirely scientific!<BR/><BR/>For example, why would it take me 30 seconds to test such a simple function? Plus I'd need to test it using a program (e.g. an interpreter) to call the method... hence this is running on the computer too... so why are the unit tests 'at least 10 times faster'. And when did 30 seconds divided by 10 equal 1 second?<BR/><BR/>I agree with what you're saying, but I think you should watch out for trying to 'prove' something using made up data. This will always be thrown back in your face.Stuarthttp://www.blogger.com/profile/05314855991059095188noreply@blogger.com