Alvaro Herrera Munoz wrote:
>On Tue, Sep 09, 2003 at 04:53:06PM +0200, Gaetano Mendola wrote:
>
>
>>"Andrew Dunstan" <andrew@dunslane.net> wrote:
>>
>>
>>>The particular assertion that was proposed doesn't strike me as terribly
>>>useful - It should be checked at the point of call rather than inside
>>>pstrdup, I should have thought.
>>>
>>>
>>Are you going to trust the client of that function ?
>>
>>
>
>Yes, because it can only used in backend code and C functions, which
>can be written only by a trusted user ('cause C is an untrusted language).
>
>(I might be wrong...)
>
>
>
Besides that, trust isn't the issue, but rather what useful information
can be gathered. How useful is it to know "someone called pstrdup() with
a null pointer"? Not very, IMNSHO.
cheers
andrew