>Last weekend I started working on it, and have done much of the work
>already. I've finished coding all the interfaces, and will start
>working on the few classes and exceptions later this week. I see the
>documentation evolving over time, as there is a ton of it to do.
I guess Brian doesn't get to do java.sql!
>Should I continue working?
Yes. I would strongly prefer that you do the documentation at the same time
as the code, however. It is never easy to go back and add the docs later.
The documentation is actually the bulk of the work here as you are aware.
When do you think you will be completely done?
>Also, as has been noted on the list, java.sql.* is largely interfaces
>and exceptions. Doing JDBC will require drivers. Is there interest
>in developing drivers within this group? I know there are some
>databases like mySQL that have drivers already; we could make more of
>an impact in this area. Perhaps starting with a JDBC-ODBC bridge?
>Then moving on to databases with published interfaces (I work mainly
>w/ Sybase, so know it best.)
Right now, I don't think it is necessary to develop any drivers. If that is
what you really want to work on, then by all means do so. If you do though,
I highly recommend that you write drivers for free software databases such
as Postgress SQL. Free drivers for non-free databases are of secondary
importance (at least to me). Also, the FSF is planning on distributing our
code, and they tend to take a dim view of code that explicitly supports
proprietary software. Perhaps drivers would best be distributed separately
anyway, since people might want to use them with packages besides ours (such
as Sun's JDK or Kaffe). If you want a mySQL driver, you don't want to have
to download and build GNU Classpath in order to get it.
BTW: Sybase already has a 100% Pure Java JDBC driver. It is called
jConnect. It is proprietary, but then again, so is Sybase!
--
Aaron M. Renn ([EMAIL PROTECTED]) http://www.urbanophile.com/arenn/
