on Linux and would have who knows what unpleasant effects on other platforms.
If you need another include file for Windows, then add it; don't go
messing with the semantics of every other port's include files.

window.

commmit then do a cvs update again.

        Win32 port is now called 'win32' rather than 'win'
        add -lwsock32 on Win32
        make gethostname() be only used when kerberos4 is enabled
        use /port/getopt.c
        new /port/opendir.c routines
        disable GUC unix_socket_group on Win32
        convert some keywords.c symbols to KEYWORD_P to prevent conflict
        create new FCNTL_NONBLOCK macro to turn off socket blocking
        create new /include/port.h file that has /port prototypes, move
          out of c.h
        new /include/port/win32_include dir to hold missing include files
        work around ERROR being defined in Win32 includes

class when lc_collate is not C.

force an initdb for this.

	- Accept CPP defines for type definitions.
	- Do not parse system include files automatically for Informix mode

Includes fixes from Joe Conway.

only remnant of this failed experiment is that the server will take
SET AUTOCOMMIT TO ON.  Still TODO: provide some client-side autocommit
logic in libpq.

the type OID and typmod of the underlying base type.  Per discussions
a few weeks ago with Andreas Pflug and others.  Note that this behavioral
change affects both old- and new-protocol clients.

This makes no difference for existing uses, but allows SelectSortFunction()
and pred_test_simple_clause() to use indexscans instead of seqscans to
locate entries for a particular operator in pg_amop.  Better yet, they can
use the SearchSysCacheList() API to cache the search results.

but that was enough tedium for one day.  Along the way, move the few
support routines for types xid and cid into a more logical place.

efficient insertion of large bytea values through the BIND interface.

dropped.  The simplest fix for INSERT/UPDATE cases turns out to be for
preptlist.c to insert NULLs of a known-good type (I used INT4) rather
than making them match the deleted column's type.  Since the representation
of NULL is actually datatype-independent, this should work fine.
I also re-reverted the patch to disable the use_physical_tlist optimization
in the presence of dropped columns.  It still doesn't look worth the
trouble to be smarter, if there are no other bugs to fix.
Added a regression test to catch future problems in this area.

more cases than I thought, so ExecTypeFromTL() will have to be fixed
anyway.

where the table contains dropped columns.  If the columns are dropped,
then their types may be gone as well, which causes ExecTypeFromTL() to
fail if the dropped columns appear in a plan node's tlist.  This could
be worked around but I don't think the optimization is valuable enough
to be worth the trouble.

detected during buffer dump to be labeled with the buffer location.
For example, if a page LSN is clobbered, we now produce something like
ERROR:  XLogFlush: request 2C000000/8468EC8 is not satisfied --- flushed only
to 0/8468EF0
CONTEXT:  writing block 0 of relation 428946/566240
whereas before there was no convenient way to find out which page had
been trashed.

instead of calling it a 'server process'.

dead xlog segments are not considered part of a critical section.  It is
not necessary to force a database-wide panic if we get a failure in these
operations.  Per recent trouble reports.

Cleanup new crypt() source code.

for rint prototype.

the folly of not passing element type to typsend/typreceive, so fix that.

so that COPY BINARY regression test passes.

fastpath.c had better check for strict functions.

messages.  Binary I/O is now up and working, but only for a small set
of datatypes (integers, text, bytea).

testing purposes.

Keep SRA copyright on file because it contains BSD license clause.