This isn't really nonconformance (was: Re: (list a1 a2 ... aN) proposed implementation)

From: Daryle Walker (darylew_at_[hidden])
Date: 2001-12-19 19:21:14

on 12/17/01 6:07 PM, Peter Dimov at pdimov_at_[hidden] wrote:

> From: "Mark Rodgers" <mark.rodgers_at_[hidden]>
>> From: "Peter Dimov" <pdimov_at_[hidden]>

[The non-conformance is about Apple using "nil" as a macro in their Mac OS
system headers.  Some members of Boost wish that "nil" was available so they
can name default types with it instead of "null_type."]

>>> I'd address this kind of nonconformance by providing a macro BOOST_MPL_NIL
>>> that expands to the right thing.
>>> 
>>> There's no need to enshrine compiler/platform-specific bugs in
>>> specifications.
>> 
>> There's no need to uglify things with macros by choosing a name that we
>> know in advance is going to be problematic.  null_type is now established
>> in the tuples libraries - why not just reuse it?
> 
> Only the users of nonconforming compilers/platforms will need to use the
> macro. Nothing is uglified.
> 
> Deciding between "works now" and "is the right thing" is difficult, but we
> need to take into account that names we choose may well last for five-ten
> years or more, long after the specific nonconformances are addressed by the
> vendors.

You shouldn't call this non-conforming.  The word "nil" isn't mentioned in
the C++ standard.  We have no higher right to use it than anyone else.  It
may be annoying, but Apple doesn't have to have any mandate to get rid of
it.

-- 
Daryle Walker
Mac, Internet, and Video Game Junkie
darylew AT mac DOT com