results

All of these comments are on examples-03. If there's a more recent
version, my apologies: I haven't seen it.

(1) The dumpasn1 output and base64'd content for AliceRSASignByCarl
don't match. similarly BobRSASignByCarl.

(2) In the dumpasn1 for 5.6, the octet strings at 1328 and 1431 should be
expanded, just so that the reader can run dumpasn1 on the decoded files
and get the same results as in the document.

(3) In 6.1, 6.4 and 6.4, the keyEncryptionAlgorithm should be
id-alg-ESDH, not dh-public-number.

(4) The MailListRc2.bin when base64 decoded is 17bytes long, and has
0d0a instead of 0a

(5) The RC2 Key Wrapping examples are still those that I found fault
with in Oslo, and Alexey Shamov similarly after that. My proposed
replacement (from July):


All the same up to "Pre Encrypt #1" (LCEKPADICV), then TEMP1 = 

035e 972a b15c c4c9 c4a0 3dba a35a 2166
67e4 3ebc a267 46ae 8608 dbc8 9e64 ca29

TEMP3 = 

29ca 649e c8db 0886 ae46 67a2 bc3e e467
6621 5aa3 ba3d a0c4 c9c4 5cb1 2a97 5e03
f797 9eb2 5900 d9c7

And the wrapped key is:
f4d8 021c 1ea4 63d2 17a9 eb69 29ff a577 
36d3 e203 86c9 0993 835b 4be4 ad8d 8a1b
c63b 25de 2bf7 7993


The good news is, with a hack to fix (3) and a hack to fix a bug of
ours, I can decrypt 6.1.

Andrew.