- Give the main locationSet based on the csvFile a sortingLocationAttributeId that makes sense (e.g. Name or logical river order). This will be the first locationSet found by FEWS, and therefore will dictate the sorting order used in the Filter (aka DataViewer, configured in Filters.xml). This is because by default locations sets use the sorting of the parent location set.
- Give all locationSets derived from the main locationSet a sortingOrder to overrule the default by name sorting order.
- Create separate locationSets for displaysGroups and Exports when the order in these two is different.
Don't define numeric attributes as text attributes in your configuration. Don't use leading zeros for numeric attributes in your csv. Leading zeros and/or text attributes will significantly increase the memory usage and start-up time of FEWS.
- (General recommendation) After making these kind of changes to config, if you experience unexpected behaviour in Delft-FEWS after a refresh, restart Delft-FEWS to see if this behaviour is persistent.
- In the filters locations with the same sorting mechanism (same attribute or by name) are grouped together (the order of grouping can depend on Java version)
Option is available since 2014.01