Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

> I assume that my CPU works

This is not a blind assumption that you make, but one based on lots of evidence (for example you are aware of the testing procedures and QA processes of the CPU manufacture, and you are aware that millions of people and companies use the same CPU model as you and if there were bugs they would have detected them).

If you were to buy some CPU that someone shady made themselves in their basement, them you would probably not make the assumption that the CPU works.

A lot of times programmers make assumptions without properly verifying that the assumption is true, and that is when problems happen. for example, when working with a library function, the programmer might test that it works with certain types of inputs and then just assume that it also works with other types, without actually checking the documentation or verifying



TBH, I got bitten by this assumption. Once, in the course of many decades. A single flipped bit - in a place where it never could happen in the program. After checking everything up and down the stack, this was our conclusion: "the issued command was invalid, thus caught and aborted; the higher-level operation was rolled back, retried, and succeeded; impact: some insignificant delay; root cause: probably cosmic rays; probability of repeating: infinitesimal. Result: no point in guarding against that." (Note that this was in normal athmospheric conditions, not in deep space, hence the probability.)




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: