Subj : Re: Software Job Market Myths
To   : comp.programming,comp.software-eng
From : CTips
Date : Thu Aug 18 2005 10:14 pm

Baxter wrote:
> "CTips" <ctips@bestweb.net> wrote in message
> news:11ga43l49me62d2@corp.supernews.com...
> 
>>Chris Sonnack wrote:
>>
>>>CTips writes:
>>>
>>>
>>>
>>>>>>Hint: you forgot a #include ...
>>>>>
>>>>>NFI why `const' is "wrong" :-)
>>>>
>>>>This code might be running on a freestanding implementation, in
>>>>which case the strcat/strcpy functions might be roll-your-owns,
>>>>in which case both the str being reserved and the addition of
>>>>const-ness are not correct.
>>>
>>>
>>>*Maybe* on the use of the str without the #include, but I really
>>>don't see why the const would be wrong in a RYO.
>>>
>>
>>Because the strcat/strcpy didn't use const's? You'd get warnings/errors
>>from the compiler.
> 
> 
> 
> Nope.  Compiler could care less if you supply a non-const for a const
> parameter.
> 
> For:
>    char *strcpy( char *strDestination, const char *strSource );
> 
> 
> strDestination  MUST be non- const, and space reserved.  strcpy WILL make
> changes to strDestination.
> 
> strSource  can be const or non-const - the compiler does not care.  "const
> char *strSource" means that strcpy will not make any changes to strSource.
> 
> 
> 

Ummm...we're talking about freestanding implementations [with a RYO 
string implementation] where the prototype for strcpy is:
char * strcpy(char *dst, char *src);
If you use:
foo( const char * x)
{
    ....
    strcpy(.., x);
}
you'll get warnings.

.