[Mulgara-dev] The Date to Fork Kowari

[David Wood]

Hi all,

This message proposes a different fork date for Mulgara.  It is  
lengthy, but I hope that all Mulgara users, contributors and  
supporters will read it carefully.

Mulgara is a fork of Kowari [1].  A fork is necessary for the simple  
reason that Northrop Grumman Corporation (NGC, [2]) purchased the  
"Initial Contributor" rights to Kowari from Tucana Technologies, Inc  
in 2005 and has seen fit to threaten me [3] (formerly a Kowari admin)  
and fail to support Andrae Muys (the current Kowari admin).   
Additionally, NGC has responded very slowly and incompletely to the  
Software Freedom Law Center [4] when requested.  They seem intent  
upon slowing the Kowari project to the point of stagnation.  I *do  
not know* why, but presume it is to buy time for their Tucana product  
to enter the market.

NGC contracted for significant bug fixes and features to be added to  
Kowari between August and December 2005.  Unfortunately, they have  
informed a lawyer with the Software Freedom Law Center that they  
object to the inclusion of some (unspecified) portions of that code.   
A legal review informed Kowari admins and ex-admins that (a) Kowari's  
MPL is on firm legal ground and (b) NGC does not appear to have  
grounds for their objection.  Nevertheless, NGC's slowness at  
defining the code which they object to and their failure to produce  
evidence of their intent to withhold permission to release that code  
to an Open Source cvs repository has stopped Kowari development.  I  
fear that Mulgara may be impacted similarly if we proceed with a  
Kowari fork which predates August 2005.

** TO AVOID ANY POSSIBLE LEGAL ENTANGLEMENT, I suggest forking Kowari  
as of 01 August 2005 and performing clean-room implementations of all  
necessary bug fixes and critical features from that point. **  Such a  
step will ensure that Mulgara can move forward cleanly without fear  
of legal action by NGC.

The most critical feature to re-add is, IMHO, the permanent model  
names code added by Brian Sletten.  Brian has graciously agreed to  
perform a clean-room reimplementation of that feature during this  
summer.  Andrae Muys and Paul Gearon, the Mulgara Project founders,  
have agreed to this approach in recent private conversations.  I now  
ask for their public support and the support of those on this list.

There are some very, very cool features upcoming in Mulgara.  Andrae  
plans to add his RDB Resolver.  Brian and I have started working  
toward NetKernel integration.  We have been working to define the  
best way to provide SPARQL support.  I would rather see us do these  
cool things than wait for NGC's lawyers and/or SFLC to give us a  
green light.  Nor do I want to see these things added to a  
potentially tainted code base!  We need to build things which change  
the world, not argue with those who don't want us to.  Forking  
earlier allows us to proceed without any question of legitimacy.

It is hoped that this step will reassure Mulgara's potential users  
that Mulgara is the semantic store to use.  Few will migrate to  
Mulgara if any question of legality, however bogus, remains.

For those of you who don't know me, I was a founder at Tucana  
Technologies and Plugged In Software.  I am now entrepreneur-in- 
residence at the University of Maryland, College Park and consult  
part time.  The original idea for a specialized metadata database  
with scalability as a feature was mine, in early 2000 (or was in  
1999?).  The idea came from a drastic simplification of the horribly  
bloated metadata architecture of Invisible Worlds, a company founded  
by IETF pioneer Marshall Rose.  Over time, in conjunction with Simon  
Raboczi, Tate Jones and others, Kowari's early predecessor was born.   
David Makepeace, Paul Gearon and many others (including relative  
newcomer but amazing coder Andrae Muys) made it cooler than I  
imagined.  Andrew Newman gets the credit for not letting the Open  
Sourcing die.  Now, I would like involve all of you and see Mulgara  
thrive.

Thanks for listening.

Regards,
Dave

[1] http://kowari.org/
[2] http://www.es.northropgrumman.com/
[3] http://prototypo.blogspot.com/2006/01/northrop-grummans-position- 
on-kowari.html
[4] http://www.softwarefreedom.org/