https://bugs.exim.org/show_bug.cgi?id=2285

--- Comment #5 from Markus Elfring <[email protected]> ---
(In reply to Philip Hazel from comment #4)
>> * Since which version of the application programming interface was this
>>   detail valid?
>
> I cannot remember, but probably since the very beginning.

This information is interesting. Does it eventually qualify for backporting
in API descriptions?


> The C standard "free()" function  has this behaviour, and I was probably 
> copying it.

This aspect can be nice, can't it?

How much do you try to avoid duplicate (or redunant) checks in related
software?


>> * Is there a need to mark such resource release functions in a special way
>>   so they could be grouped?
>
> Sorry, I don't really understand what are suggesting here.

The tolerance for null pointers in specific parameters could be expressed for
each function.
I would find it helpful to list these functions also in a corresponding
documentation section.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-- 
## List details at https://lists.exim.org/mailman/listinfo/pcre-dev 

Reply via email to