Just ship it!
© Copyright Frank D. Kanu 2000-2008
Recently I witnessed a discussion about testing. “How much should we test before we consider to market our software?” said the one manager to the other. “Bah, just ship it! Software is never ready and always bugs!” said the other.
That reminded me of a software company whose project manager shipped an untested new version destroying live data. At the big meeting following the disaster the project manager expressed his gratitude that he was not fired.
But even more amazing was the opinion of the CEO comparing the writing of software with the writing of a proposal and then concluding:
Do you think that anybody would tolerate a similar attitude from their doctor: “The operation was a great success - patient dead.“?
Still too many times testing is seen as the stepchild of programming. When the deadlines are approached and the project is late—what gets cut first? Yep, you guessed it: Testing!
During my second term at university I had the pleasure of meeting a professor who truly believed that every single software program needed the mathematical proof that it actually fulfilled its desired objectives.
While I do not know if he still thinks that way I believe that most software in our days gets shipped way too fast. The best project I have ever worked on was delivered bug-free. After eight man months of development it went through two man years of testing. That is a 1:3 ratio.
Most projects do not even get 3:1!
Wouldn’t you agree that more testing costs you way less in the long run?
Technorati (All Links are external): attitude behinds best project bugs disaster few minutes gratitude i do not know just ship it mathematical proof new version programmers project manager ratio software company software program testing thinks yep business it management
Recently I witnessed a discussion about testing. “How much should we test before we consider to market our software?” said the one manager to the other. “Bah, just ship it! Software is never ready and always bugs!” said the other.
That reminded me of a software company whose project manager shipped an untested new version destroying live data. At the big meeting following the disaster the project manager expressed his gratitude that he was not fired.
But even more amazing was the opinion of the CEO comparing the writing of software with the writing of a proposal and then concluding:
“You programmers are not less intelligent then my sales people. So get off your behinds and test the software better! It only takes a few minutes to look and see what’s wrong.“
Do you think that anybody would tolerate a similar attitude from their doctor: “The operation was a great success - patient dead.“?
Still too many times testing is seen as the stepchild of programming. When the deadlines are approached and the project is late—what gets cut first? Yep, you guessed it: Testing!
During my second term at university I had the pleasure of meeting a professor who truly believed that every single software program needed the mathematical proof that it actually fulfilled its desired objectives.
While I do not know if he still thinks that way I believe that most software in our days gets shipped way too fast. The best project I have ever worked on was delivered bug-free. After eight man months of development it went through two man years of testing. That is a 1:3 ratio.
Most projects do not even get 3:1!
Wouldn’t you agree that more testing costs you way less in the long run?
Tags:
attitude behinds best project bugs disaster few minutes gratitude i do not know just ship it mathematical proof new version programmers project manager ratio software company software program testing thinks yepTechnorati (All Links are external): attitude behinds best project bugs disaster few minutes gratitude i do not know just ship it mathematical proof new version programmers project manager ratio software company software program testing thinks yep business it management
This blog-entry is protected by a digital fingerprint:785273ed81985582c8a1be62f78c9459
