Thoughts on 15 day CJ crypto

[Eric Hughes]

   From: Phil Karn <karn at unix.ka9q.ampr.org>

   Isn't it common practice to pad out a plaintext block with random
   garbage to the size of the modulus before you RSA-encrypt it?
   [...]
   Wouldn't this thwart the kind of attack you describe?

It would, but not having ever applied for a 15-day CJ, I can't speak
to the details of what the implementations actually do.  Perhaps they
permit random padding, perhaps not.  It's certainly possible that the
padding is required to be fixed; that certainly in the style of NSA
'requests' for 'features'.

Can anybody here shed some light on the subject?

Eric