On Saturday August 04 2018 12:49:53 David Faure wrote: >Exactly, so a wrong assert leads to an unwanted crash (in a debug build). > >> - non-debug builds will simply ignore the test > >Sure, as always. Doesn't mean an assert shouldn't be correct :-)
I can't judge here, but IMHO asserts are often used as a sort of lazy, quick and dirty way to (not...) deal with situations that rarely occur. And then you forget about them, until someone does run into the situation with a release build which starts crashing in a seemingly unrelated location... >> (and apparently, crash on launch). > >I don't think so, the crash was in debug mode (= the assert). I don't think so either, but I can't say I have *never* seen DrKonqi crash. Oh well, sorry for the noise... R.
