Re: PQinitSSL broken in some use casesf

On Fri, Feb 13, 2009 at 4:53 PM, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
>> BTW, the bitmask isn't perfect either --- doesn't it just reintroduce
>> the problem already complained of with your idea for PQinitSSL? That
>> is, how does the client know whether the function recognized all the
>> bits it passed?
>
> Well, if we add the PQgetLibraryVersion() function I suggested
> upthread, then it can check that first. I find it difficult to
> believe that isn't a good idea independently of how we solve this
> particular problem.

I'd prefer PQversion() as a name. Also, it doesn't necessarily handle
the issue directly. For example, it doesn't tell you which bits are
valid...you have to guess. Also, do we really need a function for
this?

Is the generic init worth discussion or a non starter? I guess we
have a scheduling problem here...I think the ssl problem is serious
enough to warrant a fast-track into 8.4, but maybe it's 'too much' to
hash out a generic library initialization routine this late in the
cycle.

merlin