[GRASS5] (no subject)

Glynn Clements glynn.clements at virgin.net
Sun Nov 17 09:07:21 EST 2002


Alex Shevlakov wrote:

> there have been a certain number of messages like
> this(:(
> due to the fact that two needed include files are
> not in one catalog, you have to configure
> with option
> --with-postgres-includes='/usr/local/pgsql/include 
> /usr/local/pgsql/include/internal'.

I'm becoming more convinced by the day that g.column.pg should have
the use of VARHDRSZ replaced by the literal constant 4, thereby
eliminating the dependency upon postgres[_fe].h (which are internal
headers; you shouldn't need to use them unless you're writing server
extensions).

It's not as if the value varies by platform (it's defined as
"sizeof(int32)"), nor is it likely to change (more than 4Gb for a
single field of a single record? I don't think so). And if it did
change, there would be no guarantee that the value which was used by
the server matched the value which was compiled into the client.

-- 
Glynn Clements <glynn.clements at virgin.net>




More information about the grass-dev mailing list