On Aug 22, 2013, at 6:49 PM, Gary Gregory wrote: > On Thu, Aug 22, 2013 at 6:45 PM, Scott Deboy <scott.de...@gmail.com> wrote: > See the Column casting in JDBCAppender Column casting in JDBCAppender > thread on the user list for context. > > I think it's time to discuss this issue on the dev list. There's no > need for us to say 'no' 20 different ways on the user list when the > reality is that the PGSQL enum issue can be easily handled if we > choose to do it. > > Remember, any of us can hack the appender to add this feature..I > personally don't find your list of reasons to not add dialect support > compelling. > > I tend to agree with Nick here. The level of complexity involved is not > trivial to address one bug in one JDBC driver. Adding a whole layer of > indirection in the form of an SPI and impl classes should not be taken > lightly.
Exactly. It would not be nearly as simple as the Dialects that Scott references from the legacy code. Each dialect would have to construct the prepared statement differently and bond the parameters appropriately. > A simpler customization might be to let the configuration hard-code the SQL > INSERT statement. Would that do it? Unfortunately, no. The problem isn't so much in the statement but in the binding of parameters. PGSQL enum parameters must be bound with setObject(i, value, Types.OTHER) instead of setString(). > Gary > > > Scott > > --------------------------------------------------------------------- > To unsubscribe, e-mail: log4j-dev-unsubscr...@logging.apache.org > For additional commands, e-mail: log4j-dev-h...@logging.apache.org > > > > > -- > E-Mail: garydgreg...@gmail.com | ggreg...@apache.org > Java Persistence with Hibernate, Second Edition > JUnit in Action, Second Edition > Spring Batch in Action > Blog: http://garygregory.wordpress.com > Home: http://garygregory.com/ > Tweet! http://twitter.com/GaryGregory
