🚀 go-pugleaf

In article <134754@lll-winken.LLNL.GOV>, fletch@framsparc.ocf.llnl.gov (John Fletcher) writes:
|> In article <1992Aug24.071236.3172@monu1.cc.monash.edu.au> amesml@monu1.cc.monash.edu.au (Mark L. Ames) writes:
|> >There is an excellent DES implementation available by ftp from:
|> >ghost.unimi.it in /pub/crypt
|> >nic.funet.fi   in /pub/unix/security/crypt/libdes
|> >ftp.psy.uq.oz.au in /pub/DES
|> 
|> We obtained a copy of this package and tried it.  It works fine on a Vax
|> (and was fast), but it improperly encrypts on a Sun3.  Examination of
|> the source reveals that only little endian computers are provided for;
|> this is strange since the documentation refers to various big endian
|> computers.  We also note that 64-bit computers, whether big or little
|> endian, would give problems.  Are we confused?  Do we have the wrong
|> version?  What's the story?
As author of the package, I think you may have an old version.
The original package written quite some time ago was developed on vax and
MIPS processors (which were running in little-endian mode).
I currently use the package on Sparc, sun3's (68020) (big endian)
and MSDOS boxes (little endian).  I have most of the library working on
a 64bit machine (an ETA10) but most of the problems are being caused
by its C compiler defining long's and int's as 8 bytes and shorts as 4 bytes.
I have not finished the mods for this machine due to work commitments and
the fact I do not have access to the machine (debugging by email :-) but if
your C compiler defines longs as 4 bytes the code should work regardless of
endian (all the DES parts of the library appear to work on the ETA10).

The version in pub/DES ftp.psy.uq.oz.au (130.102.32.1) is the most recent
version.  If it still gives some problems, let me know the output of destest
and what compiler etc. you are using on the sun3.

eric (who only has 32bit machines to play on :-)

-- 
Eric Young | Systems programmer - Psychology Dept. Queensland Uni.
aarnet     | eay@psych.psy.uq.oz.au