On Jun 17, 2013, at 6:28 PM, Brad Roberts <braddr@puremagic.com> wrote:

> On 6/17/13 11:58 AM, Sean Kelly wrote:
>> On Jun 17, 2013, at 11:47 AM, "H. S. Teoh" <hsteoh@quickfur.ath.cx> wrote:
>>> 
>>> Do linkers actually support 8-bit symbol names? Or do these have to be translated into ASCII somehow?
>> 
>> Good question.  It looks like the linker on OSX does:
>> 
>> 	public	_D3abc1A18さいごの果実MFiZv
>> 	public	_D3abc1A4ªåMFiZv
>> 
>> The object file linked just fine.  I haven't tried OPTLINK on Win32 though.
>> 
> 
> Don't symbol names from dmd/win32 get compressed if they're too long, resulting in essentially arbitrary random binary data being used as symbol names?  Assuming my memory on that is correct then it's already demonstrated that optlink doesn't care what the data is.

Yes.  So it isn't always possible to fully demangle really long symbol names.  This is not terribly difficult to hit using templates, especially if they take string arguments.

On Thursday, 30 May 2013 at 11:29:47 UTC, Manu wrote:
> Have you ever worked on code written by people who barely speak English?

I did. It's better than having a mixture of languages like here: http://code.google.com/p/trileri/source/browse/trunk/tr/yazi.d
assert(length == dizgi.length); - in one expression!
@property Yazı küçüğü() const - property? const? küçüğü?

BTW I don't speak English myself, and D code doesn't comprise English either. How well do you have to know English to use one word to name a variable "player"? And I believe everyone who learned math know latin alphabet.

Unicode identifiers allow for typos, which can't be detected visually. For example greek and cyrillic alphabets have letters indistinguishable from ASCII so they can sneak into ASCII text and you won't see it. You can also have more fun with heuristic language switchers.

Try to find a problem in this code:
------
class c
{
	void Сlose(){}
}

int main()
{
	c obj=new c;
	obj.Close();
	return 0;
}
------

That's an actual issue I had with C# in industrial code. And I believe noone checked phobos for such errors.

I was taught BASIC at school and had no idea I should complain about latin alphabet even though I didn't learn English back then.

On Tuesday, 28 May 2013 at 00:11:18 UTC, Walter Bright wrote:
> Every time I've been to a programming shop in a foreign country, the developers speak english at work and code in english. Of course, that doesn't mean that everyone does, but as far as I can tell the overwhelming bulk is done in english.
>

OOo codebase is historically mostly in german. They try to reduce the amunt of german in the codebase with each new version.

Some massive codebases are non english.

> Naturally, full Unicode needs to be in strings and comments, but symbol names? I don't see the point nor the utilty of it. Supporting such is just pointless complexity to the language.

I know this is a crazy idea, but someone told be once that most people on this planet aren't living in english speaking countries. Insane isn't it ?