Re: [boost] checked_delete.hpp fix

From: David Abrahams (dave_at_[hidden])
Date: 2003-02-25 19:42:11

Jason Shirk <jasonsh_at_[hidden]> writes:

> From: David Abrahams
>>> Hum, it looks like Microsoft took you up on it.
>>
>>Well, it was MS I was haranguing most-loudly about it.
>
> I don't recall a discussion on warnings about ADL, I'll need a
> refresher.  

I remember mentioning it to you in Herb's car, among other places.
Does that help?

It is an attempt to find unqualified calls to functions which really
should've been qualified (the author didn't intend ADL to be used).
The ones my heuristic will warn about are those which don't happen to
use ADL in the author's tests but might unintentionally use ADL in the
field with bad results.  The other part of the equation, of course, is
that we'll also want two #pragmas:

  One we can use to label a function as a customization point, and
  suppresses the warning for calls that resolve to this function.

  Another we can use to label a function call as a customization
  point, and suppresses the warning for that call.

Could be the same #pragma, I guess.

> I do recall a discussion we had on non-dependent names.

Hmm, I don't recall that one.  Care to refresh my memory? ;-)
Maybe they're the same discussion...

-- 
Dave Abrahams
Boost Consulting
www.boost-consulting.com