Re: [HACKERS] NEXTSTEP porting problems

>> A NEXTSTEP3.3 user reported some porting problems.
>>
>> 1. #if FALSE problem
>>
>> For example in src/include/utils/int8.h:
>>
>> #if FALSE
>> extern int64 *int28 (int16 val);
>> extern int16 int82(int64 * val);
>>
>> #endif
>>
>> Unfortunately in NEXTSTEP FALSE has been already defined as:
>>
>> #define FALSE ((boolean_t) 0)
>>
>> What about using #if 0 or #if PG_FALSE or whatever instead of #if
>> FALSE?
>>
>
>Done, by you, I think.

Yes. Marc has applied my patch.

>> 2. Datum problem
>>
>> NEXTSTEP has its own "Datum" type and of course it coflicts with
>> PostgreSQL's Datum. Possible solution might be put below into c.h:
>>
>> #ifdef NeXT
>> #undef Datum
>> #define Datum PG_Datum
>> #define DatumPtr PG_DatumPtr
>> #endif
>>
>>
>> Comments?
>
>Is Datum a #define on NextStep. Can we just #undef it?

I will ask the NextStep user.
--
Tatsuo Ishii