Jan Wieck <JanWieck(at)Yahoo(dot)com> writes:
> On 10/27/2006 3:47 PM, Tom Lane wrote:
>> Is it a problem? If you really want the platform qsort you can #undef
>> qsort, but I don't entirely see why you would.
> It forces client programs to link against libpgport, which they didn't
> have to before.
Client programs that aren't part of Postgres have no business including
postgres_fe.h, and in any case the inclusion of port.h forces you to buy
into libpgport, no? This is hardly the only substitute function we have.
regards, tom lane
In response to
pgsql-hackers by date
|Next:||From: Simon Riggs||Date: 2006-10-27 21:19:54|
|Subject: Re: [HACKERS] WAL logging freezing|
|Previous:||From: Richard Troy||Date: 2006-10-27 19:57:34|
|Subject: Re: Replication documentation addition|