diff --git a/README.md b/README.md
index 89ac8d3..2281e35 100644
--- a/README.md
+++ b/README.md
@@ -1,2 +1,37 @@
 # tab2opf
 Remake of tab2opf dictionary builder for kindle
+
+Script to convert tab delimited dictionary files into opf file to run
+with kindlegen into a translation lookup dictionary for kindle.
+
+Based on the generally available tab2opf.py by Klokan Petr Přidal
+(www.klokan.cz) from 2007
+
+The script has been mostly rewritten and extended. The encoding
+convolutions have been removed, and the code migrated to python3
+
+The input form is:
+Word(s) \\t Definition
+
+By running tab2opf.py path/file.txt in the current directory file.opf
+and file*.html are created which can then be converted with kindlegen
+file.opf into file.mobi
+
+--source and --target options define which language to which we are
+translating.
+
+tab2opf has a -m option to load a module to load getkey and getdef
+functions and mapping dictionary from that namespace into the current
+one; if it doesn't find such defined member(s), no error is produced.
+
+getkey converts the term (the Word(s)) into a search key.  getdef
+converts Definition in some arbitrary way.  mapping is a dictionary
+mapping from char -> char to replace some set of characters in the
+input with characters in the output.
+
+Used and tested is the dictcc.py, which converts a dict.cc german ->
+english dictionary (http://www.dict.cc/?s=about%3Awordlist&l=e). The
+keys are the longest word in the Word(s) after removal of
+prepositional phrases and some pronouns (in getkey). The definitions
+remap the dict.cc pattern of "definition \\t part-of-speech" into
+(part-of-speech) definiton.