I would agree with your points. Test suites are often buggy but it's the constant comparisons that matter. If it takes a huge manual effort to run a test the tested system will improve slowly if at all.
The whole TDD movement gets this and programmers using TDD spend as much time writing, revising and running testing code as they do system code. This is a welcome development. Cheers Sent from my iPhone > On Apr 12, 2015, at 3:34 PM, Raul Miller <[email protected]> wrote: > > I would add, though, that: > > 1) test suites can also be buggy, > > 2) some requirements are impossible to fulfill (literally), and > > 3) "features" often are not, and > > 4) specialization is for insects. > > Oh, and this really belongs in the chat forum. > > Thanks, > > -- > Raul > > >> On Sun, Apr 12, 2015 at 4:18 PM, John Baker <[email protected]> wrote: >> I am sure everyone in this forum has experienced at least one software >> death march during their career. Still it's important to warn the younger >> generation of the ever present dark side. >> >> >> https://bakerjd99.wordpress.com/2015/04/12/cutting-the-stinking-tauntaun-and-other-adventures-in-software-archeology/ >> >> -- >> John D. Baker >> [email protected] >> ---------------------------------------------------------------------- >> For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
