Namespacing for modules (silent overwriting in cases of name collision?)

[Jose E. Marchesi]

> Hello All,
>
> Been experimenting more and having waaaaaay too much fun doing so.  I
> am looking forward to being able to program more and more in this
> language.  (The fft library is coming along nicely!)
>
> In my recent experimentation, I created two different modules that had
> similarly named pub procedures.  Everything compiled and linked just
> fine.  However, of the two similarly named exported procedures, it was
> the one from the "accessed module" listed second that would be run in
> the executable code.
>
> This is understandable.  However, it does raise a few questions:
>
> 1) Is there some way to obtain "namespacing" of modules that I am
> simply unaware of?
>
> 2) If this not possible presently, might this become possible at some
> point in time? (Or is this just inherently completely impossible?)
>
> 3) Should this name collision be a compilation/linker "error", or a
> "warning message"?

I am a little surprised.

If you have modules A and B and both modules publicize the same
identifier, if you do:


   access A, B (...)  or  access A, B def ... fed

Then you should get a compile error message.

But if you do:

   access A access B (...)


Then the definition in B will ghost the definition in A.

Still you should be able to get a warning in that case if you use
-Whidden-declarations=all.

Is that not the case?


>
> Regards,
> James
>
> PS Please disregard if this is bothersome question.  I can respect that.