PosgreSQL hackers,
Here's an initial bit of my attempt at cleaning up the the timestamp datatype.
I have gone through the backend and made a couple small changes to stop using
the HAVE_INT64_TIMESTAMP define to select a type in code by creating typedefs
in a header and using the typedef in the code. I think this small bit is ready
for inclusion for this small bit, but I have a couple questions for further
work.
1) Is there a reason that header information is duplicated between normal
posgresql include and ecpg includes instead of defining the info in one place
and #including it into the files that need it?
2) Would it be reasonable to change timestamp.h into a file that includes other
files that define the specific parts depending on HAVE_INT64_TIMESTAMP instead
of testing for HAVE_INT64_TIMESTAMP many times throughout timestamp.h? I think
this might more cleanly separate the logic for the different timestamp types.
Thanks,
wt
