In message <mailman.778.1289325458.2218.python-l...@python.org>, Terry Reedy wrote:
> I've been wondering why C programmers keep writing code susceptible to > buffer overruns ;=). I am continually disappointed with the ‘do as I say, not as I do” attitude among people offering up sample code. I remember writing to one contributor to a programming magazine many years ago, about his admonition not to simply copy-and-paste sample code, to point out that he had done exactly that (complete with bug). The main sticking point seems to be error checking; for some reason, writers of sample code seem to believe the code is “clearer” if you leave this out, when in fact all real-world applications of the illustrated techniques would include the error checks. So what do they end up illustrating? Nothing relevant to the real world. Is this why so many real-world programs fail to check for errors in important places, and segfault or behave unpredictably instead of gracefully reporting a problem? -- http://mail.python.org/mailman/listinfo/python-list
