[04:58] <RAOF> Why does cross-thread error propagation need to be so hard?
[05:52] <duflu> RAOF: Why does it exist? :)
[05:52] <RAOF> Because we're not using a language with a sensible runtime :P
[05:52] <duflu> Oh look, C++ isn't bad enough. Let's add a couple hundred more pages to the spec and make it worse
[06:05] <RAOF> Obviously you wouldn't do that.
[06:06] <RAOF> There are already two or three terrible languages in C++; no need to add more :)
[06:06] <RAOF> You might do something like Rust, for example.
[06:15] <duflu> I noticed C++11(?) recently broke even basic C code: "string"STRING
[06:15] <duflu> doesn't work any more
[07:34] <anpok> duflu: missing a literal operator?
[07:34] <duflu> anpok: No I mean "hello""world" is now invalid in C++. You have to insert a space "hello" "world"
[07:35] <duflu> Which is annoying if one of your literals is a macro and your mature stable code worked perfectly until C++11
[07:40] <anpok> interesting
[07:40] <anpok> they had to change the pre processor to get that
[07:40] <anpok> http://xkcd.com/1172/
[07:47] <duflu> anpok: I don't think it required a preprocessor change. Only a language change... ?
[07:48] <anpok> they allow "foo""bar", but not define BLA "bar"  "foo"BLA
[07:48] <anpok> i guess  "foo"BLA is a token at cpp stage
[07:48] <anpok> and not two tokens
[07:49] <anpok> otherwise they would have to deal a lot with whitespace ugly trickery within the grammar