
- integrate gettext lint tools (http://mega.ist.utl.pt/~pmmm/gettext-lint/)

High priority:

- save catalog manager content to a file, so that it can be put into
  CVS for some project

- inserting %s, %d, <html_tags> into translation with a shortcut 

- Add parsing of msgfmt's error output (e.g. msgid "foo\n" msgstr "foo" will 
  produce fatal error about \n mismatch)

- make list control customizable: optionally display only "original string"
  column (suggested by Mladen Mintakovic)

- sortable list (by columns) both in catalog manager and the main frame

- use virtual wxListCtrl
- plural handling (in wxWin, too), see info:/gettext/PO Files
- remember the exact way msgid/msgstr was stored (where the boundaries between 
  multiple lines are) and save it like that unless modified.

- TMX import/export: http://www.opentag.com/tmx.htm

- loading of .mo files

- make it possible to edit (or at least export & import) translation memory.
  Reason: there may be wrong translations in it.

Lower priority:

- (??) Some means of automatically submitting updated translations: either via
       a web or xml-rpc application or by directly accessing CVS...

- spell-checker interface (ispell, aspell)   [Vadim]
- automatic translation via web services     [Vadim]
  (Babelfish and Perl/Python script to parse it; maybe Google SOAP)

- statistics (# of words, # of characters)
- TM should split input string into sentences before querying db
- OnResize should move the sash so that its distance from bottom edge stays same
- shared databases for multiple users?
- reading MO files directly
- configurable file extension of .mo file (.gmo)
- when saving po files, write more references on one line
- It should be possible to sort for original and tranlation strings on the main
  window. This would make it easier to find similar translations, to see them
  immediately.
- Search for comment could help to find some strings faster. 
