sdyer@bbnccv.UUCP (Steve Dyer) (06/05/85)
AT&T produced a new encrypt() algorithm for their "international version"
of System V which provides encryption only. Does anyone know what
algorithm is used in this new scheme, in comparison to the almost-but-not-
quite DES algorithm used in earlier versions of UNIX? The issue here is
that many of our customer sites have deemed the older DES-based password
encryption "adequate", but are unprepared to accept the new AT&T algorithm
without some background on what it is and how it differs from the old one.
Is the new algorithm still based on DES?
--
/Steve Dyer
{decvax,linus,ima,ihnp4}!bbncca!sdyer
sdyer@bbnccv.ARPAmcb@ihlpm.UUCP (m. baker) (06/06/85)
> AT&T produced a new encrypt() algorithm for their "international version" > of System V which provides encryption only. Does anyone know what > algorithm is used in this new scheme, in comparison to the almost-but-not- > quite DES algorithm used in earlier versions of UNIX? The issue here is > that many of our customer sites have deemed the older DES-based password > encryption "adequate", but are unprepared to accept the new AT&T algorithm > without some background on what it is and how it differs from the old one. > > Is the new algorithm still based on DES? > -- > /Steve Dyer > {decvax,linus,ima,ihnp4}!bbncca!sdyer > sdyer@bbnccv.ARPA My System V manual says: "Crypt implements a one-rotor machine designed along the lines of the German Enigma, but with a 256-element rotor." This says to me that the old Crypt algorithm was not based on DES. -- Mark ihnp4!ihlpm!mcb -- -- Mark
henry@utzoo.UUCP (Henry Spencer) (06/07/85)
> My System V manual says: > "Crypt implements a one-rotor machine designed along the lines of the > German Enigma, but with a 256-element rotor." > > This says to me that the old Crypt algorithm was not based on DES. You are reading crypt(1). He is talking about crypt(3). Different code, different uses, different encryption algorithms. -- Henry Spencer @ U of Toronto Zoology {allegra,ihnp4,linus,decvax}!utzoo!henry
jim@ISM780B.UUCP (06/08/85)
>My System V manual says: >"Crypt implements a one-rotor machine designed along the lines of the >German Enigma, but with a 256-element rotor." > >This says to me that the old Crypt algorithm was not based on DES. You are reading crypt(1); encrypt in in crypt(3). The nature of the difference between the two-way crypt and the one-way crypt is only available to those with domestic source licenses. Those with access to the difference are likely to be amused. -- Jim Balter, INTERACTIVE Systems (ima!jim)
guy@sun.uucp (Guy Harris) (06/08/85)
> > AT&T produced a new encrypt() algorithm for their "international version" > > of System V which provides encryption only. Does anyone know what > > algorithm is used in this new scheme, in comparison to the almost-but-not- > > quite DES algorithm used in earlier versions of UNIX? > > > > Is the new algorithm still based on DES? I believe it's still DES based but has been "crippled" so that the Russkies can't get a hold of DES technology :-(. > My System V manual says: > "Crypt implements a one-rotor machine designed along the lines of the > German Enigma, but with a 256-element rotor." > > This says to me that the old Crypt algorithm was not based on DES. Different other "crypt". There's the "crypt" routine, which is DES-based (at least in pre-stupid-export-control UNIX and in domestic System V) and the "crypt" command which is rotor-based (and has been cracked, but it's *still* not available in international System V). Guy Harris
jas@rtech.UUCP (Jim Shankland) (06/11/85)
> > AT&T produced a new encrypt() algorithm for their "international version" > > of System V which provides encryption only.... Is the new algorithm > > still based on DES? > > -- > > /Steve Dyer > > {decvax,linus,ima,ihnp4}!bbncca!sdyer > > sdyer@bbnccv.ARPA > > My System V manual says: > "Crypt implements a one-rotor machine designed along the lines of the > German Enigma, but with a 256-element rotor." > > This says to me that the old Crypt algorithm was not based on DES. > > -- Mark > ihnp4!ihlpm!mcb Just in case no-one else responds: crypt(1) is the one-rotor Enigma-like thingy. crypt(3), which includes encrypt() and setkey(), and is used for password encryption, IS based on DES. And I, too, would like to know what AT&T uses in its international System V. Anybody know? Jim Shankland ..!ucbvax!mtxinu!rtech!jas ..!ihnp4!pegasus!rtech!jas
lee@eel.UUCP (06/15/85)
Crypt(3) in international SystemV is exactly the same as before except that the decryption capability has been removed. This is easily verified without access to the source by comparing the encryption of login passwords on systems with and without the decryption capability.
ed@mtxinu.UUCP (Ed Gould) (06/18/85)
In article <2284@sun.uucp> guy@sun.uucp (Guy Harris) writes: >> > AT&T produced a new encrypt() algorithm for their "international version" >> > of System V which provides encryption only. Does anyone know what >> > algorithm is used in this new scheme, in comparison to the almost-but-not- >> > quite DES algorithm used in earlier versions of UNIX? >> > >> > Is the new algorithm still based on DES? > >I believe it's still DES based but has been "crippled" so that the Russkies >can't get a hold of DES technology :-(. Close, but not quite. There is an export restriction on programs that can *decrypt* DES, but not on those that can *encrypt*. Removing setkey() is sufficient to meet the export requirements. -- Ed Gould mt Xinu, 2910 Seventh St., Berkeley, CA 94710 USA {ucbvax,decvax}!mtxinu!ed +1 415 644 0146