Re: [boost] Re: [filesystem] Operations predicates

From: Peter Dimov (pdimov_at_[hidden])
Date: 2005-05-04 05:16:15

Beman Dawes wrote:
> "Peter Dimov" <pdimov_at_[hidden]> wrote in message
> news:005701c55008$0eacc010$6401a8c0_at_pdimov2...
>> Rob Stewart wrote:
>>> I don't like the status()/symlink_status() split.  How about
>>> overloading like this:
>>>
>>>   struct follow_symlink_t { };
>>>   extern const follow_symlink_t follow_symlink;
>>>
>>>   status_flags status(path const &);
>>>   status_flags status(path const &, follow_symlink_t);
>>
>> FWIW, I'm in favor of the current design.
>
> Rationale?

I don't view overloading for overloading's sake as improvement; there's 
nothing wrong with giving different names to different behaviors.

The practice of overloading on behavior seems inspired by new(nothrow), but 
in that case we simply don't have the option of providing a 
differently-named function.

One practical argument for not introducing overloading is that it's harder 
to use boost::bind on an overloaded function.