I have a proposition.

Some programmers (including me) dont want to organize their programs into classes/structs because they are worried that by doing so would mean that every global function would have a hidden extra parameter, which in a large program could pose a potential slow down (wether this is true or not is besides the point).

So it means that their programs become big messes of global variables / functions, and everything can be accessed from anywhere without any qualification.

So I propose that D be included with a keyword that would make any global in a specific module require qualification before it can be used.

For example if I had a global variable in module.d called x, any file in my program could access x just by doing:

int test = x;

-and it would be very unclear where x comes from. With this proposed keyword the programmer would be forced to type:

int test = module.x;

Which is lot clearer. I know that D allows you to qualify the name already, but more often than not people wont bother, and this proposed keyword ("qualify"?) would help maintain readability in large programs where OOP is not used.

"imr1984" <imr1984_member@pathlink.com> wrote in message news:d3om7g$lsf$1@digitaldaemon.com...
> With this proposed keyword the
> programmer would be forced to type:
>
> int test = module.x;
>
> Which is lot clearer. I know that D allows you to qualify the name
> already, but
> more often than not people wont bother, and this proposed keyword
> ("qualify"?)
> would help maintain readability in large programs where OOP is not used.

Ahh, yet another use for explicit namespaces.. ;)

Are namespaces planned for D?

In article <d3p7p3$163t$1@digitaldaemon.com>, Jarrett Billingsley says...
>
>"imr1984" <imr1984_member@pathlink.com> wrote in message news:d3om7g$lsf$1@digitaldaemon.com...
>> With this proposed keyword the
>> programmer would be forced to type:
>>
>> int test = module.x;
>>
>> Which is lot clearer. I know that D allows you to qualify the name
>> already, but
>> more often than not people wont bother, and this proposed keyword
>> ("qualify"?)
>> would help maintain readability in large programs where OOP is not used.
>
>Ahh, yet another use for explicit namespaces.. ;)
>
>

"imr1984" <imr1984_member@pathlink.com> wrote in message news:d3pj10$1ep0$1@digitaldaemon.com...
> Are namespaces planned for D?

I hope so.  They were something I suggested a few weeks ago.  Walter never replied, but that usually doesn't mean much.

In any case, explicit namespaces could solve your problem and mine..