but why doesnt the code work you used?

I don't know

lol

lol

} catch(...) { cout << current_exception().__cxa_exception_type()->name() << endl; } This is implementation-specific but works 😐

its horrible

just go for clang please

I agree with rule.

lol

looks like clang has several linking issues in msys. I should try the Windows version.

also libc++ seems to be bugged in the msys version

try msvc

only in the msys version?

Hm... the TDM MinGW64 version of g++ looks to be working fine

I need my code to be platform independent (at most)

then dont use exceptions

aren't c++ exceptions portable?

I was catching an exception instead of an exception*

Oh, it's not going to be that portable

huh?

my targets are desktop systems, mostly gnu/linux and windows. FreeBSD and OSX are planned but I'm not sure about that.

right

go for WxWidgets

void code_that_throws_logic_error() { throw new logic_error( "Cannot instantiate a Nut with no content"); } int main(int argc, char **argv) { try { code_that_throws_logic_error(); } catch(exception e) { cout << e.what() << endl; } return 0; }

The exception is a match if [...] - T is a reference to an unambiguous public base class of E - [...] Try catching via catch (const std::exception& e) (not throwing new)

try { thing_that_throws(); } catch(logic_error *e) { cout << e->what() << endl; } this works.

remove the new as well

what new?

#include <iostream> #include <exception> void code_that_throws_logic_error() { throw std::logic_error( "Cannot instantiate a Nut with no content"); } int main(int argc, char **argv) { using std::cout; using std::endl; try { code_that_throws_logic_error(); } catch(const std::exception &e) { cout << e.what() << endl; } return 0; }

gdb exits with "Signal received: ? (Unknown signal)"

So, this is just a type mistake.

gdb basically quit because neither you nor gdb did not catch the exception

why wx, is there a particular reason? I know it's full-featured anyway

What else to use?

what new?

I don't get people accusing compiler of bad code-gen if they don't even know what's wrong in the C++ code ... feels bad..

I don't get people accusing compiler of bad code-gen if they don't even know what's wrong in the C++ code ... feels bad..

I didn't asssume anything, I actually asked just why it didn't work. There's to say, I didn't specify how I threw the exception. My bad.

You did throw new ... I would stronglly suggest not to do it. In this case you are responsible for the exception allocation and deallocation and it's very hard to tell when it's safe to relrease the memory. Rather throw exception and catch by const reference

Nothing wrong :) This message was not intended for you

Nothing wrong :) This message was not intended for you

Uhm, makes sense to me. Is there a case where you'd rather throw new instead?

Uhm, makes sense to me. Is there a case where you'd rather throw new instead?

You know, if my code works but doesnt anymore when i use O3, i'm pretty sure that we can blame the compiler ;)

no.. because this applies to both GCC and Clang. So chances are your code is wrong

it's always easier to blame the compiler

new and delete are bad habits in Cpp