Re: statically linked PL's

Alex Pilosov wrote:
> I was thinking that a solution to problems with all dynamically linked PLs
> (such as perl and python, which must be compiled as shared library to be
> usable), is to link them statically into postgres.
>
> Anyone disagrees? If not, I'll write some patches, probably by hacking
> dynloader to understand when things were already linked in...

No need to change the dynamic linker. The PL's handler
function must be defined in language 'internal' instead of
'C' then and this must be done at initdb time. Just the
CREATE LANGUAGE would then control if a DB has it or not.

Since it should IMHO be a compile time config option, we'd
need config dependant entries in .bki files for initdb, and
createlang/droplang must check if there is an 'internal'
handler in pg_proc before mucking with it.

Jan

--

#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me. #
#================================================== JanWieck(at)Yahoo(dot)com #

_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com