• Don <nospam nospam.com> Nov 15 2009
• Ellery Newcomer <ellery-newcomer utulsa.edu> Nov 15 2009
• #ponce <aliloko gmail.com> Nov 15 2009
• Andrei Alexandrescu <SeeWebsiteForEmail erdani.org> Nov 15 2009
• Frank Benoit <keinfarbton googlemail.com> Nov 15 2009
• Jesse Phillips <jessekphillips gmail.com> Nov 15 2009
• =?ISO-8859-1?Q?=c1lvaro_Castro-Castilla?= <alvcastro yahoo.es> Nov 16 2009
• Trass3r <mrmocool gmx.de> Nov 16 2009
• Bernard Helyer <b.helyer gmail.com> Nov 22 2009

Don <nospam nospam.com> writes:

Before TDPL is published, I think we should finally get rid of C-style 
declarations. Supporting two styles of declarations just causes 
confusion, and AFAIK it doesn't have any benefits -- for example, it 
doesn't significantly help in porting code from C. D declarations are 
just better. And there are bugs such as this one, which seem to require 
abandoning C declarations:

bug #2392 Parsing ambiguity between function pointer declaration and 
function call

At the very least, can we make sure that TDPL doesn't commit us to 
retaining them forever?

Ellery Newcomer <ellery-newcomer utulsa.edu> writes:

Don wrote:
 Before TDPL is published, I think we should finally get rid of C-style
 declarations. Supporting two styles of declarations just causes
 confusion, and AFAIK it doesn't have any benefits -- for example, it
 doesn't significantly help in porting code from C. D declarations are
 just better. And there are bugs such as this one, which seem to require
 abandoning C declarations:
 
 bug #2392 Parsing ambiguity between function pointer declaration and
 function call
 
 At the very least, can we make sure that TDPL doesn't commit us to
 retaining them forever?
 

#ponce <aliloko gmail.com> writes:

Don Wrote:

 Before TDPL is published, I think we should finally get rid of C-style 
 declarations. Supporting two styles of declarations just causes 
 confusion, and AFAIK it doesn't have any benefits -- for example, it 
 doesn't significantly help in porting code from C. D declarations are 
 just better. And there are bugs such as this one, which seem to require 
 abandoning C declarations:
 
 bug #2392 Parsing ambiguity between function pointer declaration and 
 function call
 
 At the very least, can we make sure that TDPL doesn't commit us to 
 retaining them forever?
 


Yes please. There is no real use for:
int myvar[4];

instead of:
int[4] myvar;

and porting from C requires some rewriting anyway.

Andrei Alexandrescu <SeeWebsiteForEmail erdani.org> writes:

Don wrote:
 Before TDPL is published, I think we should finally get rid of C-style 
 declarations. Supporting two styles of declarations just causes 
 confusion, and AFAIK it doesn't have any benefits -- for example, it 
 doesn't significantly help in porting code from C. D declarations are 
 just better. And there are bugs such as this one, which seem to require 
 abandoning C declarations:
 
 bug #2392 Parsing ambiguity between function pointer declaration and 
 function call
 
 At the very least, can we make sure that TDPL doesn't commit us to 
 retaining them forever?


I agree. Walter and I agreed that TDPL will not even mention the C-style 
declarations for functions and arrays. There are a couple of other older 
features (such as "length" in array index expressions) that, again in 
agreement with Walter, TDPL does not mention and are on the road to 
deprecation.

Andrei

Frank Benoit <keinfarbton googlemail.com> writes:

Andrei Alexandrescu schrieb:
 ...(such as "length" in array index expressions) ...


finally yay!

Jesse Phillips <jessekphillips gmail.com> writes:

Thanks Don for keeping this page up to date:


http://www.prowiki.org/wiki4d/wiki.cgi?LanguageDevel

=?ISO-8859-1?Q?=c1lvaro_Castro-Castilla?= <alvcastro yahoo.es> writes:

Don Wrote:

 Before TDPL is published, I think we should finally get rid of C-style 
 declarations. Supporting two styles of declarations just causes 
 confusion, and AFAIK it doesn't have any benefits -- for example, it 
 doesn't significantly help in porting code from C. D declarations are 
 just better. And there are bugs such as this one, which seem to require 
 abandoning C declarations:
 
 bug #2392 Parsing ambiguity between function pointer declaration and 
 function call
 
 At the very least, can we make sure that TDPL doesn't commit us to 
 retaining them forever?
 


++votes;

Trass3r <mrmocool gmx.de> writes:

Don schrieb:
 Before TDPL is published, I think we should finally get rid of C-style 
 declarations. Supporting two styles of declarations just causes 
 confusion, and AFAIK it doesn't have any benefits -- for example, it 
 doesn't significantly help in porting code from C. D declarations are 
 just better.

Bernard Helyer <b.helyer gmail.com> writes:

On 16/11/09 04:27, Don wrote:
 Before TDPL is published, I think we should finally get rid of C-style
 declarations.


Yes please.