Tom Lane wrote:
> Bruce Momjian <[EMAIL PROTECTED]> writes:
> > I thought about that but I figured the less we change the tz library,
> > the easier it will be to replace later.
> 
> I have come to the conclusion that trying to minimize changes to the tz
> library is a bad tradeoff.  It's not going to be a minimally sized diff
> for long.  For instance, do you propose to exclude the tz code from all
> future pgindent runs?  If you look at the changes we made in the regex
> library to make it minimally compliant with our own coding standards,
> they were pretty extensive.  The same is probably going to happen to the
> tz code.  For one thing, it is definitely not going to have K&R-style
> function declarations for long --- those are trouble waiting to happen.
> 
> For another thing, it's not going to try to emulate the C library's API
> for long, because getting out from under that brain-dead API is exactly
> the reason for wanting to have our own timezone code.  (I suspect the
> only part of this code we really want in the long run is zic.c.)
> 
> In any case we should certainly not fix portability failures in the tz
> code by breaking our other code, or even just risking breaking it.

I wasn't aware we were going to be gutting it.  I am still trying to
just get it to work on Unix.

-- 
  Bruce Momjian                        |  http://candle.pha.pa.us
  [EMAIL PROTECTED]               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073

---------------------------(end of broadcast)---------------------------
TIP 9: the planner will ignore your desire to choose an index scan if your
      joining column's datatypes do not match

Reply via email to