[6249] in Kerberos
Kerberos on Linux (was KERBEROS?)
daemon@ATHENA.MIT.EDU (Jonathan Kamens)
Fri Nov 17 05:04:26 1995
To: kerberos@MIT.EDU
Date: 17 Nov 1995 08:53:08 GMT
From: jik@jik.datasrv.co.il (Jonathan Kamens)
In article <48dn1f$di@news.magi.com>, dmtech@magi.com (Andrew Mileski) writes:
|> Andy Smith (smitha@worf.netins.net) wrote:
|> > Has anyone ever gotten kerberos to compile on linux?
|>
|> Kerberos is, sadly, junk. It is not worth the effort.
There are many people, happily using Kerberos on Linux and many other
operating systems, who do not share your opinion.
|> I've yet to hear of a success story on Linux.
Well, then, you haven't been paying very much attention.
First of all, I believe that the Cygnus Kerberos IV distribution compiles and
works with very little (if any) pain on Linux. The exception to this *might*
be the Berkeley r-commands, e.g., rlogin and rsh, since there's so much
OS-specific hair in them that's hard to port.
Second, the MIT Kerberos IV release also works with little trouble on Linux if
you know what you're doing (i.e., any problems you might encounter while
compiling it are obvious and easy to fix). Again, the r-commands may be a
problem.
Third, I believe that the current MIT Kerberos V release, which includes
Kerberos IV backward-compatibility, compiles and works out-of-the-box on
Linux. And in that case, you can avoid any problems with the r-commands by
using the Kerberized telnet included with the release.
Actually, I'll amend that last point by stating that there are some known bugs
in the current Kerberos V release, but they're not (for the most part)
Linux-specific, and in any case, people have fixed them and made patches
available (when MIT will get around to putting out a patch release of their
own is anybody's guess -- several people have asked in
comp.protocols.kerberos, but MIT hasn't answered yet). If you're interested,
check old postings in comp.protocols.kerberos to find out where the patches
are available, or ask there if there aren't any postings at your site that
mention the patches.
Have you ever worked with the MIT or Cygnus Kerberos code base? I have,
extensively, including porting it to Linux. If all you're basing your
comments on is anecdotal evidence, perhaps you should refrain from commenting.
