Subject | Re: [Firebird-Java] Re: Jdbc 2.0 Patches |
---|---|
Author | David Warnock |
Post date | 2002-05-21T08:12:32Z |
David,
dbms (we use postgresql and mysql as well as firebird). While xa support
is not available for them (nor available in several frameworks -
turbine, melati, jrf for example) then we need different coding
techniques to use xa when using firebird. That is
a) a pain foi programmers
b) a pain for dbms portability
packaging clearer as to where the various functionality is. That makes
the code much more approachable.
developing applications for any length of time will already have
connection pooling. We cannot just drop 4 years of good solid code that
has been working totally reliably and is nicely maintainable to jump
onto jboss/jca.
David
> IMO most of the excessive complexity of the current driver comes from theBut these are precisely my points. Our code has to work with other free
> shenanigans necessary to have xa support.
>
> Again IMO the principal selling point of firebird and this driver is that
> it is as far as I know the only free open source driver/db with xa support.
dbms (we use postgresql and mysql as well as firebird). While xa support
is not available for them (nor available in several frameworks -
turbine, melati, jrf for example) then we need different coding
techniques to use xa when using firebird. That is
a) a pain foi programmers
b) a pain for dbms portability
> Rather than creating 2 or more java drivers I'd rather see the effort putI don't think anyone is suggesting 2 drivers, merely trying to make the
> into adding into firebird 2 the ability to do work on any transaction on
> any (physical) connection.
packaging clearer as to where the various functionality is. That makes
the code much more approachable.
> Remember, among the goodies the jca support gives you essentially for freeOnly if you can use it within your application. Anyone who has been
> is a DataSource with built in connection pooling.
developing applications for any length of time will already have
connection pooling. We cannot just drop 4 years of good solid code that
has been working totally reliably and is nicely maintainable to jump
onto jboss/jca.
David