Switching to the use of PostgreSQL's COPY command and Oracle's SQL*Loaderto load the data for this package. Much, much faster than a bazillionINSERT statements.
- Added site_node::rename proc- Enhanced get_children so it can return a specific element from the site_node array structure, e.g. node_id, instead of the url.
Added a hack to delete permissions for the application group, so we can delete the group when deleting the subsite instance. We really ought to make permissions cascade. but that's a much bigger fix.
Moved csv files to match naming scheme that makes it easy to load in bothPG and Oracle - dd-table-name.csv where dd is the sequence number, usedto control the order in which table data is loaded.
1. removed some errant log statemets previously inserted when debugging2. sql/common/*sql files weren't being displayed by the APM, are now3. added index on start_date in timespan-create after user reported very slow calendar queries in .LRN
