I wrote:
> The TEA-N version I post on my web page is checked against the reference
> implementation (you can find a link to it at my page:
> http://members.tripod.com/~mdileo/) for various values of the text and
> keys chosen to avoid problems related to palindrome keys.

Since I got some emails about the correctness of my code, I have put a
table containing some of the values I used to check my TEA and TEA-N
code on my CryptoPIC page at http://members.tripod.com/~mdileo/. After
the Andrew Warren warning I double checked my tea11.zip and tea13.zip
code against the contents of the reference sheet I made finding no
discrepancies. FYI the sheet is an output form the execution of the
reference TEA (and TEA-N) 'C' implementation on a PowerPC machine.
Please note that my code use the LSB-first convention for the 32 bits
values used by the algorithms.

Ciao
  Marco



----
Marco DI LEO                  email: m.dileo@sistinf.it
Sistemi Informativi S.p.A.    tel:   +39 6 50292 300
V. Elio Vittorini, 129        fax:   +39 6 5015991
I-00144 Roma
Italy