Details
P4 Description
In my company, for our .Net applications, we are fond of the powerful devexpress grid. Such a lacking feature in the java ecosystem has been a recurrent argument against introducing java on the client side, in spite of having all of our backends written in java, and some challenges in terms of .Net/Java interop.
For that reason, and because I believe java deserves a decent advanced grid, I am suggesting the development of 2 features that are lacking in the TableView : grouping and sorting.
Grouping works by selecting column headers and dropping them into the group zone. As a result, the table morphs into something like a tree with subtables for each node. It still maintains a global behavior, specifically when it comes to sorting: sorting by 1 or multiple columns is still possible and it applies a sort on all subtables. Another aspect of it is that there is only 1 row of headers.
Filtering allows to define a filter per column. The filter expression can use operators such as starts with, contains, like, ... Additionnally filters can be independently activated or deactivated.
These 2 features (added to performance) become critical when it comes to data analysis with large sets of data, as it is usual in financial services. The lack of it, unfortunately, is a showstopper for certain guis, and then it gets harder to convince people to still do the "less demanding guis" in java.
see also the wpf grid http://www.devexpress.com/Products/NET/Controls/WPF/Grid/sort-group.xml?
For that reason, and because I believe java deserves a decent advanced grid, I am suggesting the development of 2 features that are lacking in the TableView : grouping and sorting.
Grouping works by selecting column headers and dropping them into the group zone. As a result, the table morphs into something like a tree with subtables for each node. It still maintains a global behavior, specifically when it comes to sorting: sorting by 1 or multiple columns is still possible and it applies a sort on all subtables. Another aspect of it is that there is only 1 row of headers.
Filtering allows to define a filter per column. The filter expression can use operators such as starts with, contains, like, ... Additionnally filters can be independently activated or deactivated.
These 2 features (added to performance) become critical when it comes to data analysis with large sets of data, as it is usual in financial services. The lack of it, unfortunately, is a showstopper for certain guis, and then it gets harder to convince people to still do the "less demanding guis" in java.
see also the wpf grid http://www.devexpress.com/Products/NET/Controls/WPF/Grid/sort-group.xml?
