🚀 go-pugleaf

Andrew Koenig (ark@research.att.com) wrote:
> In article <45efb3$u7@stc06.ctd.ornl.gov> kennel@msr.epm.ornl.gov writes:

> > That's a good reason.  Did the people who put forth the request for proposals
> > mention "ability to perform safety checks if desired" as an area that was
> > suitable to address?

> The committee doesn't operate in terms of `requests for proposals'
> in the sense I think you mean. 

So who sets the agenda of its scope?  And what defines proposals?

>  But in general, safety is not very
> high on the lists of concerns of the committee members.  Performance
> and conceptual compatibility with C are much higher, for example.

> After all, if you want semantically safe languages that are fundamentally
> different from C, you don't have to look very far.

Perhaps the ISO standard should require that all conforming 
implementations' marketing documents include these two paragraphs.

1/2 :-) 

My personal opinion: sometimes C++ has suffered when it made some
provision for "conceptual compatibility with C" when amplifying and
continuing such behavior turns out (IMO) to be a mistake in context of
C++.  For instance, automatic conversions.  I would have preferred that
there be no implicit conversions, and the ones in C be allowed to remain
as a 'deprecated' feature. 

When it cameto tightening up static type checking and other similar issues
there didn't seem to be much hesitation by the committee toward promoting
safety. 

cheers
Matt

> -- 
> 				--Andrew Koenig
> 				  ark@research.att.com