Brian Ryner <[EMAIL PROTECTED]> writes:
> Ok, here is the debug output.
Thanks.
It seems that my first guess was wrong... The interesting part of the
log is
: 0x00000230: 4616622c54b80603b640340000002839
^^^^^^^^
: 0x00000240: 01e466a46e8906d0d83b5c02ad0598eb
: 0x00000250: c6c1166a92896ccb45486f09fe083052
: 0x00000260: cd39dd34ca5115
:
: handle_connection: Received packet of type 31 (SSH_MSG_KEXDH_REPLY)
The part marked with ^ is the length field for the signature blob.
Thus, the length field is definitely included. So it seems there's
either a bug in lsh's ssh2 workaround, or some more general
incompatibility in the signature process or in the hashing of the DH
values.
To resolve this, I'm afraid reading the ssh2 source code is a must.
BTW, I'm about to go on vacation. I may be reading my mail
occasionally, but I probably won't do any serious hacking.
I'm planning to get to the IETF meeting in Oslo which starts on
Sunday. And also to the Chaos Computer Club Camp in Berlin August 6-8.
Is anybody else going there? It would be great to see some of you
in RL.
I'll be back August 9.
Best regards,
/Niels
