Short answer: Yes, I am, including without testing. For several years I ran a challenge for exactly this question. I have a test harness and test suite, and wo

Short answer: Yes, I am, including without testing. For several years I ran a challenge for exactly this question. I have a test harness and test suite, and would run people's programs through it. Roughly 1 in 30 or 40 of those submitted without testing (self-certified) would pass all the tests. Those who tested before submission passed about 1 time in 3. The vitriol heaped on me was quite unbelievable, and in one case led to concerns for my physical safety. People who had decided to try the challenge then rounded on me and said it was unfair, unreasonable and unrealistic. I didn't affect the results - the vast, vast majority of people simply could not write a binary search that passed my test suite. FWIW, my first attempt had exactly one problem, to the best of my knowledge. I decided to derive my code formally from the specification. The mistake concerned computer arithmetic, not the algorithm, and was the one lunk to from the article. But my background is unusual.