🚀 go-pugleaf

Andrew Koenig <ark@research.att.com> wrote:
> > why aren't you agitating for a name-mangling standard in C?
> > Name mangling is a potential problem there too.

To which Peter van der Linden replied:
> Name-mangling in C++ was a grotesque hack (like having the compiler
> implemented as a C-preprocessor).  It provided a compiler 4 months sooner,
> and set the language back five years.

If you're going to flame implementations that hacked their way
through pernicious problems, you might well point out the problem
they hacked their way through: the linkers under UNIX (and most
everywhere else) were (and are) too feeble to handle typesafe
linkage.

I'd tend to say that the C++ language designers did us all a favor
by going ahead with language features that stressed the environment,
rather than one that subjugated its goals to the limitations of
the (external) tools at hand.

Rather than grotesque, I'd tend to describe the CFRONT name-
mangling 'hack' as an elegant, though complex, solution to a
difficult problem.

> Let's try and face up to the deficiencies of C++.

In this case, don't you mean CFRONT's use of UNIX (et al) linkers?

How would you have maintained C-style linkage and C++ typesafe
linkage in the same object module without changing the linker?

Cheers,

Larry Baker