Requiring trap is attractive but problematic. Precludes almost all algebraic transformations at compile time. Analogous to fenv for floats.
Yes, experts do this because they think they're smarter than they are. You can be smart enough once or ten times, but not N times as N→∞.
-
-
But it's still conscious breaking of simple rules that they (we, me included) should be following.
-
Your musl example was bad because that was one of the few pieces of imported code, but the IPv6 & DNS parsing vulns were real examples.
-
The latter were just me being overly clever & wrong about assuming loop invariants rather than using easily-provably-correct conditions.
-
Future direction for musl is to reduce cleverness, optimize for simplicity whenever there's not a demonstrable need for cleverness.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.