Re: Problem with compilator DEV C++

<jcoffin@[EMAIL PROTECTED]
> wrote in message
news:MPG.225401fa9eb637d7989c32@[EMAIL PROTECTED]
> In article <64t180F2dcf9oU1@[EMAIL PROTECTED]
>, r124c4u102@[EMAIL PROTECTED]
> says...
>
> [ ... ]
>
>> > That is NOT a using directive but a perfectly (to me) acceptable
using
>> > declaration.
>>
>> Granted, it's a declaration and not a directive.
>>
>> Now let's see if I have this straight.  Let's say, just for the sake of
>> argument, that stealing is a bad thing.  But it is "perfectly
acceptable"
>> if
>> you only steal *one* thing at a time?  Is that the rule?.
>
> A poor analogy. Likening it to stealing, implies that either a using
> directive or a using declaration is fundamentally wrong.
>
> That's not the case -- they're included in the language because there
> are times that (albeit open to some disagreement) it's perfectly
> reasonable to put them to use.

Yes, my point is that it is fundamentally wrong to use either one *in a 
header file*. Putting
a using declaration in a header file means that anyone who uses that file
must inspect it to see what is there and then *remember* each one that was
used.  In a pretty trivial program I would use at least 20 of the names in
std::.  I do not think I should impose such a burden (remembering my
personal list) on my fellow programmers just so I can avoid typing a few
keystrokes.   The advantage of the using thingies that I see, is avoiding 
pollution of the source code with noise words. The typing labor can be 
minimized by the editor or other macro key aids.

If someone has a one man shop he can - pretty obviously - do whatever he
pleases, no one cares.

A widely distributed package that allowed using declarations in its' 
associated  header file sounds like a Really Bad Idea to me.  I think I
have 
said pretty much all I want to say on this subject.