On 12/08/2014 08:50 PM, Doug wrote:
I don't think everything you said is true. Some of the Linux distros have paid programmers, and I'm sure that they must do some testing before they put an app out. Even programmers who donate their time must do some testing too.No programmer I ever knew would just write code and ship! And in a long engineering career, I've known quite a few!
Indeed, and I've managed many. Programmers test their own code to do what they think they have written. Sometimes they find that what they wrote isn't what they thought they wrote, but on the whole programmers are useless at doing meaningful testing of their own code. They are too attached to it to see it for what it really is. I knew one programmer who wrote terrible code -- he enjoyed debugging it! So I took him off programming and had him debug everyone else's code. I insisted that if they wrote such great code he wouldn’t find any bugs and so if he did they had to pay him $0.50 per bug. Their egos caused them to agree t that. Fools! He made about $100 the first week, about $50 the second and settled down to about $10 each week for the rest of the project. Read that: THE REST OF THE PROJECT. These programmers who "put their money where their ego was" still managed to produce about 20 bugs per week! (that was about 1 bug per day per person.) I recall reading that IBM reckoned that there were about 200 bugs in each release of MVS. And each "fix" of one generation introduced more, even as it fixed those, so the number remained about constant. We see that certainly in Microsoft releases. And you'd think there is a learning process, but no, the way the 'system' works the experienced programmers get promoted. I did. The programmers' attachment to their own code that blinds them to its defects, both in the design and in the coding, is well known and long established.. You can fin it in Weinberg's "The Psychology of Computer Programming" from 1971 and he references some studies on this going back to the early 1960s. Most references on programming teams make it very clear that the people who write the code and the people who test is should NOT be the same. Some recommend, as I found, separate and competitive teams. Its not that programmers don't test their own work, its that they are blind to many of the mistakes they make. Heck so are writers, that why they have copy editors. Now if only my spelling checker were smart enough ... -- /"\ \ / ASCII Ribbon Campaign X Against HTML Mail / \ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org