[ https://issues.jboss.org/browse/TEIID-2554?page=com.atlassian.jira.plugin... ] Steven Hawkins edited comment on TEIID-2554 at 6/14/13 1:12 PM: ---------------------------------------------------------------- Merge doesn't buy you that much for materialization updates. New rows you can already process using the existing merge syntax. Unless you are comparing all columns (some of which may be none comparable), there is no graceful way of detecting updates. You'd have to have a notion of an update timestamp/increment, which you can process efficiently without the merge in any case. I'd have to give this pretty low priority/leave it unbucketted given the lack of standardization (although it is part of ansi, just not widely followed) and the amount of pieces needed to implement it. Also under the covers the current merge handling is nothing more than a compensating procedure, there's still no notion of pushdown merge support (somewhat due to the previous statement). was (Author: shawkins): Merge doesn't buy you that much for materialization updates. New rows you can already process using the existing merge syntax. Unless you are comparing all rows, there is no graceful way of detecting updates. You'd have to have a notion of an update timestamp/increment, which you can process efficiently without the merge in any case. I'd have to give this pretty low priority/leave it unbucketted given the lack of standardization (although it is part of ansi, just not widely followed) and the amount of pieces needed to implement it. Also under the covers the current merge handling is nothing more than a compensating procedure, there's still no notion of pushdown merge support (somewhat due to the previous statement). -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira