• Oliver Hoog (11/11) Aug 08 2009 Since one of D's goals is dropping the bad things from C/C++, I wonder
• Adam D. Ruppe (7/10) Aug 08 2009 Pragmatism includes sticking to established habits in humans. This is on...
• Rainer Deyke (10/14) Aug 08 2009 I don't think D needs two inline comment forms, but /+...+/ is no less
• language_fan (3/6) Aug 10 2009 We could also overload .. to define a vector dot product, and *** for
• Daniel Keep (3/14) Aug 08 2009 It's not ambiguous. Tokens are matched with maximal munch meaning that
• Stewart Gordon (5/7) Aug 09 2009 Actually, it isn't always maximal munch - for example, 2..3 tokenises as...

Oliver Hoog <kingboscop nospam.gmail.com> writes:

Since one of D's goals is dropping the bad things from C/C++, I wonder 
why the ambiguous syntax for multi-line comments is still allowed:
---
	auto p = new int;
	*p = 10;
	auto j = 100/*p;    <- error
---
There is already the unambiguous /+...+/ syntax, so why not drop the 
/*...*/ syntax? Compatibility? Convenience? I think D should be more 
pragmatic.
What do you think?

"Adam D. Ruppe" <destructionator gmail.com> writes:

On Sun, Aug 09, 2009 at 02:25:04AM +0200, Oliver Hoog wrote:
 There is already the unambiguous /+...+/ syntax, so why not drop the 
 /*...*/ syntax? Compatibility? Convenience?

Both.

 I think D should be more pragmatic.

Pragmatism includes sticking to established habits in humans. This is one of
those habits.

-- 
Adam D. Ruppe
http://arsdnet.net

Rainer Deyke <rainerd eldwood.com> writes:

Oliver Hoog wrote:
 There is already the unambiguous /+...+/ syntax, so why not drop the
 /*...*/ syntax? Compatibility? Convenience? I think D should be more
 pragmatic.
 What do you think?

I don't think D needs two inline comment forms, but /+...+/ is no less
ambiguous than /*...*/.  Don't forget about the unary '+' operator:

auto p = 1;
auto j = 100/+p;

(Also, "multi-line"?  I only use /*...*/ for inline comments that end
before the end of the line.  I use multiple instances of '//' if I want
a comment that extends beyond the end of a line.)


-- 
Rainer Deyke - rainerd eldwood.com

language_fan <foo bar.com.invalid> writes:

Sat, 08 Aug 2009 19:26:48 -0600, Rainer Deyke thusly wrote:

 
 I don't think D needs two inline comment forms, but /+...+/ is no less
 ambiguous than /*...*/.  Don't forget about the unary '+' operator:

We could also overload .. to define a vector dot product, and *** for 
cross product.

Daniel Keep <daniel.keep.lists gmail.com> writes:

Oliver Hoog wrote:
 Since one of D's goals is dropping the bad things from C/C++, I wonder
 why the ambiguous syntax for multi-line comments is still allowed:
 ---
     auto p = new int;
     *p = 10;
     auto j = 100/*p;    <- error
 ---
 There is already the unambiguous /+...+/ syntax, so why not drop the
 /*...*/ syntax? Compatibility? Convenience? I think D should be more
 pragmatic.
 What do you think?

It's not ambiguous.  Tokens are matched with maximal munch meaning that
/* is ALWAYS the opening of a multi-line comment.

Stewart Gordon <smjg_1998 yahoo.com> writes:

Daniel Keep wrote:
<snip>
 It's not ambiguous.  Tokens are matched with maximal munch meaning that
 /* is ALWAYS the opening of a multi-line comment.

Actually, it isn't always maximal munch - for example, 2..3 tokenises as 
2 .. 3 rather than 2. .3.  But you're right about comment syntax.

Stewart.