Comments on: Focusing on prevention rather than cure http://melalouise.net/2005/02/focusing-on-prevention-rather-than-cure/ Aware parenting, sewing and life in Amsterdam Sat, 08 Jul 2006 05:04:24 +0000 hourly 1 http://wordpress.org/?v=abc By: blondie http://melalouise.net/2005/02/focusing-on-prevention-rather-than-cure/comment-page-1/#comment-35 blondie Fri, 18 Feb 2005 09:43:51 +0000 http://melalouise.net/?p=51#comment-35 I'm staggered at the reliance some organizations have on testing teams. To me this suggests a clear disregard for the benefits developer testing bring, which from a business perspective is an easily accessible, automated, efficient and repeatable process that can provide as comprehensive coverage as the quality gauge suggests (a reference to the good ol' days of brother Gibson's project dashboard). There's one other important aspect of developer tests that the management frequently overlook - no human element is necessary. To me, that's a massive point. We all know humans are error prone - you can bank on testers not testing the same thing every time. It comes down to trust; I trust statistics above a head testers stating 'Yes, we believe the system is ready for production'. I want to see "nothin' but green" with execution times within performance boundaries, then I'll make a decision. Testers are still important, manual testing does play an important part in QA, but there should be much less of it than what I've seen, which at it's worst has been a 3:2 ratio of developer to tester in an organization that seemingly valued testers above developers. Indeed, testers were working more frequently to the wee hours of the morning than developers - Madness! I say look at those principles you talk of, cull numbers down to say 5:1 and get those developers to pump out tests before coding! I’m staggered at the reliance some organizations have on testing teams. To me this suggests a clear disregard for the benefits developer testing bring, which from a business perspective is an easily accessible, automated, efficient and repeatable process that can provide as comprehensive coverage as the quality gauge suggests (a reference to the good ol’ days of brother Gibson’s project dashboard). There’s one other important aspect of developer tests that the management frequently overlook – no human element is necessary. To me, that’s a massive point. We all know humans are error prone – you can bank on testers not testing the same thing every time. It comes down to trust; I trust statistics above a head testers stating ‘Yes, we believe the system is ready for production’. I want to see “nothin’ but green” with execution times within performance boundaries, then I’ll make a decision.

Testers are still important, manual testing does play an important part in QA, but there should be much less of it than what I’ve seen, which at it’s worst has been a 3:2 ratio of developer to tester in an organization that seemingly valued testers above developers. Indeed, testers were working more frequently to the wee hours of the morning than developers – Madness! I say look at those principles you talk of, cull numbers down to say 5:1 and get those developers to pump out tests before coding!

]]>