Re: Inefficient handling of LO-restore + Patch

On Wed, Apr 24, 2002 at 06:13:28PM -0400, Tom Lane wrote:
> 
> Is that sufficient?  The clients probably are not affected by quite as
> many config options as the server, but they still have a nontrivial
> list.  (Multibyte, SSL, Kerberos come to mind at once.)  I'd not like
> to see us assume that a one-line output format will do the job.
> 
> A way to interrogate the libpq being used by psql might be good too.

I like the way perl handles this, for example

perl -MExtUtils::Embed -e ccopts

for options to cc used when compiling(-I and stuff) and

perl -MExtUtils::Embed -e ldopts

for options to ld used when compiling(-L and stuff).
I think it would be really nice if we could have psql to ask its libpq to
spew out something similiar. Then you could do stuff like

cc -o ex ex.c `psql -ccopts -ldopts` 

and not having to worry about where the libraries are.

-- Magnus