Steps to reproduce:

* Create a new spreadsheet.
* Enter "abc" into A1 and "ABC" into A2.
* Press the column label for column A to select the column.
* Go to "Data->Filter->Standard Filter" in the menus.
* Click "More" to open the extra options.
* Uncheck the "Range contains column labels" check box.
* Check the "No duplication" check box.
* Change the "Value" of the first filter criteria to "- not empty -".
* Press OK.

Expected result:

The duplication checking should work case-insensitively since the "Case
sensitive" was not checked, and hence "abc" and "ABC" should be treated as a
duplicate and only "abc" should be kept.

Actual result:

Both "abc" and "ABC" is kept, they are both treated as unique values.

Additional info:

I found this in 2.4.0 (Kubuntu package), but Richlv on IRC (#openoffice.org @
irc.freenode.net) could reproduce this on DEV300_m14 as well.

Could it be that the duplication checking was added after the case insensitivity
checking, and so the case insensitivity checking was left to only pertain to the
three "normal" filter criterias at the top of the Standard Filter dialog?