A proposal for `system' imports

[Raphael Collet]

> Whatever you do, unqualified names will be a problem. This probelm is
> well known and is well documented. We should not waste time on
> replacing one solution with drawbacks by another one.
> 
> Moreover, the solution proposed by Denys has no advantage whatsoever
> to me. 
> 
> The current solution allows easier first-time exposure to modules, and
> this is in my viewpoint a very strong argument.

IMHO this is a dangerous argument.  BASIC was considered to allow easier
first-time exposure to programming ;-)

The current solution looks like a magic thing at first sight.  And it
rapidly sucks when you use it for real.  Suppose I write a functor
Foo.ozf as part of an application, and that another Foo.ozf is added to
the future standard library.  I know that I could spend hours before
finding where such a bug comes from.  I already did...

> So I am very much against this, in addition please consider that
> changing everything (module manager, linker, all the docs, many
> programs) will inevitebly introduce bugs.

You're right.  But the current solution will eventually introduce bugs,
as the library grows up.


--
raph
-
Please send submissions to hackers at mozart-oz.org
and administriva mail to hackers-request at mozart-oz.org.
The Mozart Oz web site is at http://www.mozart-oz.org/.