The next iteration of scope

[ketmar via Digitalmars-d]

On Thu, 02 Apr 2015 16:04:59 +0200, Jacob Carlborg wrote:

> On 2015-04-02 14:02, ketmar wrote:
> 
>> .di files has no function bodies, yet they still works. so compiler
>> *can*
>> do full processing, but it is not required, and if compiler does that,
>> this restriction can be removed in the future. but with automatic
>> inference for functions we are stuck with full semantic analysis.
> 
> I talked about when the body is available. In my experience most
> projects do not use .di files.

i just used .di an example that demonstrates that there is no need to do 
full-blown semantic analysis even for functions with bodies. this process 
(semantic analysis) can be both slow and memory consuming, so building 
big application with many modules will be unnecessary slow and will 
require more resources. and we already has a kind of "mark" that told 
compiler to do attribute inference: empty template arg specifier.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: not available
URL: <http://lists.puremagic.com/pipermail/digitalmars-d/attachments/20150402/40acd214/attachment.sig>