Tom Lane wrote:
> Joe Conway <mail@joeconway.com> writes:
>>Tom Lane wrote:
>>
>>>Also, if we do this we probably ought to remove the special-purpose
>>>hack for preload_libraries to specify an init function --- it should
>>>just happen by default. Any objections to simplifying that?
>
>>The original idea of using the init function with preload_libraries was
>>to eliminate library startup that was expensive and only needed once.
>>Specifically in the case of libR (and presumably other libraries as
>>well), the init time was much greater than the actual library load time.
>>If it is removed from preload_libraries, then we'll pay that price for
>>every backend startup, no?
>
> No, my thought is that you'd rename PL/R's init function to PG_init, and
> then it'd get called automagically without needing to assume that the DBA
> remembers to specify it in preload_libraries. If there's a reason *not*
> to do that then it'd be a strike against this whole proposal, methinks.
Oh, well that sounds perfect to me. At least in the case of a procedural
language handler you can easily initialize and cache anything that must
be done per-backend anyway. It's the "expensive but must be done at
least once stuff" that was a problem. As long as that happens, I'm
happy. And if we eliminate a dba dependency, so much the better.
Joe