[7254] in Kerberos
Re: OpenVision donates Kerberos admin system to MIT
daemon@ATHENA.MIT.EDU (Jonathan Kamens)
Sat May 11 15:03:11 1996
To: kerberos@MIT.EDU
Date: 11 May 1996 18:50:49 GMT
From: jik@annex-1-slip-jik.cam.ov.com (Jonathan Kamens)
In article <199605102306.TAA07486@hedgefund.com>, dhip@hedgefund.com (Dhip Sotb) writes:
|> I applaud OpenVision for donating such a significant
|> piece of technology to the public domain. However, I am
|> confused as to their motivation. Perhaps the stock owner
|> can explain it to me.
(Note: I speak for myself, not for OV. But I do work for OV, and I
also own stock in the company. :-)
I think that there are several reasons why we are doing this, including:
* One of the surest ways to further the acceptance of new standards is
to make free, useable code available which implements them. There is
free code available that implements Kerberos V5 (i.e., the MIT
release), but the Admin system included with that code is really
sub-standard, and the fact of the matter is that a Kerberos V5
implementation without a good Admin system is *far* less likely to be
used in production environments than one with a good Admin system.
Therefore, by donating a good Admin system to MIT, OV is helping
Kerberos V5 to gain more widespread acceptance. Which is good for the
Kerberos community, good for computer security in general, and good for
OV because it means we'll be able to sell more of our Kerberos product.
* Many people use the MIT Kerberos product. By making our code
available, including its source code, we greatly enlarge our user base,
as well as putting our code into the hands of many people who are
willing to track down and fix bugs (not that there are any bugs in the
code, of course :-) and even implement and donate back to MIT
improvements to our code. Those fixes and improvements can then be
incorporated back into our commercial product.
So, by donating the code, we're acquiring a bigger user base, test
base, and developer base for our product.
* Having our Admin system be the one that is included as the "official"
Admin system in the MIT distribution gives us a leg up on our
competition. It turns our system into a de facto standard, and if it
ends up being used by enough sites, our competition will be forced to
support it. But since we wrote it and have the most familiarity with
it, we're more prepared to support it than our competitors are.
This particular reason is pretty much totally for our own benefit :-).
But I'd be lying if I didn't list it, because it is certainly a real
consideration.
As one of the engineers who was involved from day one with the
development of OV's Admin system, I can say with confidence that we
always hoped to donate it back to MIT, regardless of whether it would
give us an advantage against our competitors. After all, pretty much
everyone who originally worked on the system was either a student of
MIT or an employee of Project Athena at one point or another :-).
Nevertheless, I recognize the "business reality," which is that if
there wasn't a good business case for us to donate the code, we
wouldn't have been able to do it.
Incidentally, I should mention that pretty much all of the fixes we've
made to MIT code have been donated back to MIT. Also, the GSS-API
implementation in the MIT distribution (at least the first one; I think
that others may have been donated since then) wsa donated by OV. Alas,
we haven't had the manpower to make sure that some of our recent fixes
have found their way back to MIT, but I hope to make sure that happens
when I'm back in the States and working for OV full-time again
(I'm living in Israel until July, and I'm only working for OV part-time
right now).
Again, I want to stress that I can't speak for OV. This is my
perception of how this donation came to happen, but I haven't been
involved in any of the nitty-gritty decision-making about it, both
because I'm just a lowly grunt and because it's hard to be involved in
political dealings of this sort over E-mail from a different country.
