DES.26.cs was not posted to the Squeak list, just to the off-list group
working on consolidating the plugins.  It is very possible that there is a
problem with that particular change set, and I think I also had a problem
with the simulator code on test1 (needs an 'asByteArray' message before
sending it off).

The case I used in test1 came from Schneier.  If someone has other test
cases and can validate them, I'd be more than happy to include them.

Anyway, I'll see if I can verify your problems.

Regards -

Duane

>Hi!
>
>I was testing the code that was posted by Duane Maxwell on 8 Feb 2000
>called DES.26.cs (I already deleted the original email, sorry).
>
>I didn't build a pluggable primitive but want to use the Smalltalk code.  I
>think, the test method DES>>primPluginAvailable is wrong.  Shouldn't it
>fail with "^ false" ?  After changing that, I could "cook" a key and
>transform a block of bytes although it took me a moment to figure out that
>#transform: will mutate the passed byteArray.  I'd prefer a method that
>returns the transformed object.
>
>It seem also that DES class>>test2 doesn't work (Character doesn't
>understands <<) because the test arguments are string and not byte arrays.
>
>Finally and most important, I think the algorithm doesn't work.  At least
>for me it didn't work unless I modified the ByteBit variable to contain
>
>#(1 2 4 8 16 32 64 128 )
>
>instead of
>
>#(128 64 32 16 8 4 2 1 )
>
>I found this by comparing the Smalltalk algorithm with a working Java
>implementation.  Can anybody confirm that this is a problem?
>
>
>bye
>--
>Stefan Matthias Aust  //  Bevor wir fallen, fallen wir lieber auf.


===================================================
Duane Maxwell          dmaxwell (at) entrypoint.com
CTO                       http://www.entrypoint.com
EntryPoint, Inc.    (858)348-3040  FAX(858)348-3100

Information contained herein is my personal opinion
     and not necessarily that of EntryPoint.
===================================================