BOB@earth.arc.nasa.gov (Bob Day) (11/29/88)
I've just started writing an application in which I need crypt (the password encryption function), but much to my surprise I've discovered that it's not there (Xenix 2.2.3). It is in the lint library (/usr/lib/llibc), but it's not in any of the object libraries (/lib/[LMS]libc.a). Does anyone know why SCO left this one out?
vipoon@kepler1.UUCP (Victor I Poon) (11/29/88)
In article <1238@amelia.nas.nasa.gov> BOB@earth.arc.nasa.gov.arpa (Bob Day) writes: >I've just started writing an application in which I need crypt (the password >encryption function), but much to my surprise I've discovered that it's not >there (Xenix 2.2.3). It is in the lint library (/usr/lib/llibc), but it's >not in any of the object libraries (/lib/[LMS]libc.a). Does anyone know why >SCO left this one out? Since the DES encrytion algorithm is used by so many computers, it is illegal to ship the routines out of the country, (not that this will REALLY stop someone if they want it). You must call up SCO and they will send you the disks with the crypt libraries on them (as long as your address is in the US). You must then compile with a -lcrypt flag. Hope this helps. -- Victor I. Poon {acsm, sbcs, polyof}!kepler1!vipoon Kepler Financial Management, Ltd. You make more of me, I'll make more of you - Red Jack
jbayer@ispi.UUCP (Jonathan Bayer) (11/30/88)
In article <1238@amelia.nas.nasa.gov>, BOB@earth.arc.nasa.gov (Bob Day) writes: > I've just started writing an application in which I need crypt (the password > encryption function), but much to my surprise I've discovered that it's not > there (Xenix 2.2.3). It is in the lint library (/usr/lib/llibc), but it's > not in any of the object libraries (/lib/[LMS]libc.a). Does anyone know why > SCO left this one out? Crypt is available by calling SCO. My understanding is that crypt cannot be shipped overseas (no, it doesn't make sense). Any domestic installation can obtain it by calling sco and asking for it. Jonathan Bayer Intelligent Software Products, Inc.
rac@jc3b21.UUCP (Roger A. Cornelius) (11/30/88)
From article <1238@amelia.nas.nasa.gov>, by BOB@earth.arc.nasa.gov (Bob Day): < I've just started writing an application in which I need crypt (the password < encryption function), but much to my surprise I've discovered that it's not < there (Xenix 2.2.3). It is in the lint library (/usr/lib/llibc), but it's < not in any of the object libraries (/lib/[LMS]libc.a). Does anyone know why < SCO left this one out? Check the release notes. You can get it by calling SCO. Roger Cornelius uunet!pdn!jc3b21!sherpa!rac
mike@mks.UUCP (Michael Brookbank) (12/01/88)
> >I've just started writing an application in which I need crypt (the password > >encryption function), but much to my surprise I've discovered that it's not > >there (Xenix 2.2.3). It is in the lint library (/usr/lib/llibc), but it's > >not in any of the object libraries (/lib/[LMS]libc.a). Does anyone know why > >SCO left this one out? > I think SCO doesn't ship CRYPT automatically because it is illegal for an American company to ship CRYPT out of the US for security reasons. Rather than have separate inventory they deliver only on request. If it is any easier my firm, MKS Inc., has a product called MKS Trilogy which includes CRYPT, AWK and the Korn Shell designed to run on SCO XENIX 286 & 386, Microport, and 386ix. -- Mike Brookbank Phone: (519)884-2251 Mortice Kern Systems Inc. UUCP: uunet!watmath!mks!mike 35 King St. North BIX: join mks Waterloo, Ontario N2J 2W9 CompuServe: 73260,1043
BOB@earth.arc.nasa.gov (Bob Day) (12/01/88)
Thanks to everyone who replied about crypt. We called SCO tuesday morning and crypt arrived in the mail the next morning. Pretty good response time.
domo@riddle.UUCP (Dominic Dunlop) (12/01/88)
In article <1238@amelia.nas.nasa.gov> BOB@earth.arc.nasa.gov.arpa (Bob Day) writes: >I've just started writing an application in which I need crypt (the password >encryption function), but much to my surprise I've discovered that it's not >there (Xenix 2.2.3). It is in the lint library (/usr/lib/llibc), but it's >not in any of the object libraries (/lib/[LMS]libc.a). Does anyone know why >SCO left this one out? Hmmm. Could be that you've got the export version of Xenix. We aliens definitely don't get crypt(3X). Instead we have to look in _UNIX System Security_ by Kochan and Wood (Hayden, 1985) for a listing of a C language implementation of the Data Encryption Standard, or in Scientific American, November and December 1988, for a reasonably exhaustive discussion. Unfortunately, neither of these sources detail how crypt(3X) peturbs the algorithm in one of 4,096 different ways according to the ``salt'', so we would still have difficulty if we wanted to reimplement. Sigh. -- Dominic Dunlop domo@sphinx.co.uk domo@riddle.uucp
george@genat.UUCP (George Gorsline) (12/02/88)
Why no crypt? It is available for US customers only as a special item.
Omitting crypt is required by US export regulations to keep un-friendly
countries from gaining access to this valuable technology. The manual
has a reference to this effect.
Whether Canada is an un-friendly country or the question of
extra-territoritality is not at issue. There are PD crypts available;
whether they meet the US requirements is not known.
--
George Gorsline VE3FIU / K8HI
One of the VE3YDX gang... Y DX? Because it's there!(Y not?)
__... ...__ . ... _.. _.._
Genamation, Inc., 351 Steelcase Rd. W, Markham Ontario L3R 3W1
uunet{!mnetor,!utzoo}!genat!george (416) 475-9434jfh@rpp386.Dallas.TX.US (The Beach Bum) (12/03/88)
In article <83@genat.UUCP> george@genat.UUCP (George Gorsline) writes: >Whether Canada is an un-friendly country or the question of >extra-territoritality is not at issue. There are PD crypts available; >whether they meet the US requirements is not known. The crypt program in the archives is a DES-based three rotor German enigma. It is exportable as the code was, as I recall, written by a Frenchman. It is not compatible with the crypt from System V - if you convert the code you will have to have all of your users re-crypt their files. -- John F. Haugh II +-Cat of the Week:--------------_ /|- VoiceNet: (214) 250-3311 Data: -6272 |Aren't you absolutely sick and \'o.O' InterNet: jfh@rpp386.Dallas.TX.US |tired of looking at these damn =(___)= UucpNet : <backbone>!killer!rpp386!jfh +things in everybody's .sig?-------U---
rudolf@acadch.UUCP (Rudolf Kuenzli) (12/09/88)
What I never will understand is the fact that DES has been described officially
and everybody in this world may obtain a copy of it. I think the organization
is named ASA(?). So why bother about programs written using DES?
Rudolf Kuenzli {sun|uunet}!acad!acadch!rudolfchip@vector.UUCP (Chip Rosenthal) (12/15/88)
In article <509@acadch.UUCP> rudolf@curano.UUCP (Rudolf Kuenzli) writes: >What I never will understand is the fact that DES has been described officially >and everybody in this world may obtain a copy of it. The issue here isn't reason or logic, it is government bureaucracy. DES is a controlled item for export. Therefore your choice is either fight the bureaucracy or don't. Unbundling crypt allows SCO to take the second route. -- Chip Rosenthal chip@vector.UUCP | Choke me in the shallow water Dallas Semiconductor 214-450-5337 | before I get too deep.
cjh@hpausla.HP.COM (Clifford Heath) (12/30/88)
> The crypt program in the archives is a DES-based three rotor German > enigma. It is exportable as the code was, as I recall, written by a > Frenchman. > John F. Haugh II Sorry, that's not true. No implementation of any bidirectional encryption program may be exported or re-exported from the US. That's why anyone can get the DES papers, the algorithms, but not necessarily the programs. Almost everyone in the world who wants programs has them anyway, of course, but don't tell the NSA :-). We wrote one in less than a week direct from papers on the subject. In any case, DES was promoted as a standard only after it was reduced to 56 bits from the original research which used 128 bits, since it is usually possible to brute-force crack it at 56 bits. It's unlikely that this ploy fooled the Ruskies, though. > The crypt program in the archives is a DES-based three rotor German The rotor has nothing whatever to do with DES. The crypt program is a simple triple rotor, no DES anywhere. My opinions here have nothing whatever to do with my employer. Clifford Heath. cjh%hpausla@hplabs.HP.COM
donegan@stanton.UUCP (Steve Donegan) (06/23/89)
Does anyone have a replacement for the libcrypt.a that is in the public domain for XENIX 386? Sources preferred. Please email any responses. Thank you.
edhew@egvideo.UUCP (Ed Hew) (06/28/89)
In article <870@stanton.UUCP> donegan@stanton.UUCP (Steve Donegan) writes: >Does anyone have a replacement for the libcrypt.a that is in the public >domain for XENIX 386? Sources preferred. Please email any responses. >Thank you. It's not public domain, but I understand that Mortice-Kern Systems in Waterloo, Ontario, Canada have a commercially offered crypt replacement. You'd have to ask them about specific library routines. Try <uid>@mks (I don't know specific loginid's there). If in doubt you could always 'postmaster' or even 'root'. Hope this helps. --ed {edhew@egvideo.uucp}