Re: CVS commit: src

2010-12-25 Thread Jukka Ruohonen
On Sat, Dec 25, 2010 at 04:42:30AM +, David Holland wrote:
 On Fri, Dec 17, 2010 at 10:38:21PM +0200, Jukka Ruohonen wrote:
   A related note: should we provide typeof(3) in the restricted namespace
   instead, i.e. as __typeof(3)?
 
 People are more likely to find the page if it's installed as
 typeof(3), I think, since that's how they'll think of it, and if
 necessary the page can include discussion of when it's an available
 symbol.

No, I mean shouldn't this be defined conditionally in cdefs(3)? But as typeof(3)
can not be replaced by another compiler-agnostic construct (?), probably not.

- Jukka.


Re: CVS commit: src

2010-12-25 Thread David Holland
On Sat, Dec 25, 2010 at 09:51:00PM +0200, Jukka Ruohonen wrote:
A related note: should we provide typeof(3) in the restricted namespace
instead, i.e. as __typeof(3)?
   
   People are more likely to find the page if it's installed as
   typeof(3), I think, since that's how they'll think of it, and if
   necessary the page can include discussion of when it's an available
   symbol.
  
  No, I mean shouldn't this be defined conditionally in cdefs(3)? But
  as typeof(3) can not be replaced by another compiler-agnostic
  construct (?), probably not.

Oh, I see. It might be desirable but as it's a language extension I'm
not sure we get that degree of control.

However, since AFAICR typeof is fairly likely to appear in c1x,
probably many compilers will eventually supply it.

-- 
David A. Holland
dholl...@netbsd.org