—
Jest-in-Time compiler
New C++ papers available
http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2018/2018-02-pre-Jacksonville.htm …
What'll make it to C+20? 
Show this thread
-
I just presented “signed integers are two’s complement” http://wg21.link/p0907r0 to a joint session of C++ numerics and undefined behavior groups! Broad support, but they want to keep UB on overflow (and have separate type or operator for wrap). I’ll update the paper
Show this thread
Not only is the C++ standards committee happy to make signed integers two's complement in C++20 with http://wg21.link/p0907 , the *C standards committee* voted to do the same thing for the next version of C! 
I just have to write the wording. 
-
-
Does this mean conversions to signed types will be well-defined as modular reduction (e.g. (int)-1U==-1)?
-
For C++ yes, for C they asked for more targeted papers for ~each bullet point I had in C++. I think some of the effects of the C++ wording wasn’t clear to the C committee, and I’m not there in person to explain. So maybe? I’ll certainly try to get the same effects as C++.pic.twitter.com/N5lS9UkrWt
-
That being said, there are gotchas around enum when converting to/from int. Furthermore, C sometime differs from C++ in ways I’m not familiar with. I’ll lean on wording experts, hopefully
@zygoloid, to help get things harmonized between the two languages.
End of conversation
New conversation -
-
-
Does this mean that base 3 computers can't run C++20 code without emulating base 2 for signed integer types?
-
Which base 3 computer tries to run modern C or C++?
-
None, that I’m aware of. It has significant advantages over base-2, though, if (big huge if) we can figure out how to build them.
-
In that case, my proposed change breaks none of them
End of conversation
New conversation -
-
-
WHAT ? ?? ? ? ? ? ? WHat just happened to hell, it froze /o\
-
Neither is standard yet! Hell can still spark back up
-
next thing you'll know they'll add expression statements... and strong enum types.
-
We’ll truly know the apocalypse is at hand when default automatic variable initialization gets standardized
-
lol, it will never happen.
-
My new paper: “You can have any undefined value you want, as long as it’s 0”
-
Which zero tho?
More seriously… I intend to look into this as well. -
If you manage this, I will make it my life’s work to create a national holiday in your honour
End of conversation
New conversation -
-
-
There are a handful of UNISYS programmers gonna be sad about this.
-
No sadder than they already are. The unisys compiler was last updated in 2017 and still only supports C90. It emulates old hardware on x86 with attached FPGAs. I venture that no new code is being written, otherwise they’d use less expensive hardware.
End of conversation
New conversation -
-
-
That was not unintentional! The mailing deadline was the 2nd, so I held on...
End of conversation
New conversation -
-
Does this also mean that right shifting of signed integers will finally be formally defined to do 2s complement sign extension, rather than being 'implementation dependent'?
- 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.