Re: Compile failure in CVS - JDBC3

Haakon,

First a little background/history

when jdk1.1 came out java.sql.* was implemented in the
org.postgresql.jdbc1 package. Probably not packaged this way at the
time, but it is before my time so I am guessing.

Then jdk1.2 came out and this required a seperation of the packages;
primarily because of additional Types in java.sql.Types. They couldn't
be referenced in the jdbc1 package because they didn't exist in it.

So it looks like the orignal people who did the jdbc2 version simply
copied the jdbc1 stuff, repackaged it into jdbc2 and added the methods
etc. Along they way they started using some jdbc2 specific methods such
as the Collections package. The challenge was that there was alot of
duplicate code shared between the packages, which meant that an update
in one required an update in the other ....this was tolerable with only
two packages to deal with, now we have jdbc3 and presumable we can
expect 4, 5, .... etc.

So the rewrite solves the problem of duplicate code shared between the
two versions of jdbc1,jdbc2, and now jdbc3.

We will have to implement the new methods in the jdbc3 spec.

Dave

On Mon, 2002-07-29 at 05:38, Håkon Hansen wrote:
> Monday 29 July 2002 11:17, skrev Dave Cramer:
> > Bruce,
> >
> > Try getting a fresh copy, Barry did a very large rewrite to prepare for
> > jdbc3 and removed alot of files. I can compile clean.
>
> What does this rewrite do (or solve)? I thought JDBC3, which is suppoded to be
> included in J2SE 1.4, would be supported by the pgjdbc2.jar driver. At least,
> the comment next to the download link says that JDK 1.4 is supported ...
>
> (And Dave: Might this have something to do with my connection problem?)
>
>
> Yours,
>
> Haakon Hansen
> Norway
>
>