[Mulgara-dev] Subversion

Tue Jun 6 16:58:03 UTC 2006

Hey everyone

Saw the emails going around last week and wanted to introduce myself to
the list.  My name is Ben, I've been developing on top of tks since one
of the earlier versions...as I write this I am leafing through the
"Tucana Intelligent Connections Developer Training" book...David Wood
actually came to our company and did a three day training course on tks
and we have been using tks ever since.  

Although we are using tks2.2 right now on our production system we have
downloaded, compiled, and tested kowari with semi-success, (I sent out a
few emails a few weeks ago with restore issues we are having and not
being able to shut the system down and bring it back up).  Luckily we
are not using any of the extra features of tks itself, so we are going
to be able to transition to mulgara fairly easily. 

I think we are doing some interesting things with tks and are continuing
to push forward with developing and using the system on a daily basis.
I'm not sure what constitutes a large database (our backups are around
80 megs), but I think we have a very nice test system were we can push
new builds through their paces and hopefully we can do some testing.

Looking forward to working on mulgara...  Thanks

-ben    

________________________________

From: mulgara-dev-bounces at mulgara.org
[mailto:mulgara-dev-bounces at mulgara.org] On Behalf Of Paul Gearon
Sent: Friday, June 02, 2006 2:23 PM
To: mulgara-dev at mulgara.org
Subject: Re: [Mulgara-dev] Subversion

Hi Brian,

If you check it in before I do, then I have no problems with this.  :-)

My concerns with Eclipse and Subversion were because I think it might do
a refactor as an "svn remove" and an "svn add" instead of an "svn mv".
I wanted the "mv" command so we wouldn't lose the history behind any
files.

However, in tagging kowari-1.2 it looks like we don't have a history for
any of the files in the latest tag anyway.  (Not sure if this was
intentional or what, but that's how it is).  So at this point it doesn't
actually matter if there's a remove/add or a mv.

Also, we'll want to do a general clean up of "kowari" strings in the
code as well.  This is so we can catch the internal URIs which are used
for special operations.  We also need to pick the the "tucana" URIs as
well.  We were supposed to convert the tucana URIs for Kowari as well,
but that doesn't matter now.  This will include test code, data files,
etc.

Once we're done, the only reference I want to see to "Kowari" is in any
legal statements we make on copyright (since we have to attribute
copyright of the original source to the new owner).

As for introductions: of course I can see the list of people who've
subscribed to the list, but it would be rude for me to "out" them.  I
recognize a number of the names though.

For anyone who doesn't know me, I was one of the first developers on
TKS.  Of course, the core of TKS was eventually open-sourced and called
Kowari.  I worked at Plugged In (which became Tucana) from early 2000,
until the company was dissolved at the end of 2004.

During my time at Tucana I co-wrote the storage layer for TKS with David
Makepeace.  I also worked on the query resolution with Simon Raboczi.  I
then spent about 8 months "fixing bugs" and adding features at every
layer of the code.  This was frustrating work at the time (since I
couldn't concentrate on anything for more than a week or two), but it
gave me a good overview of the whole system.

During 2004 I started working on RDFS and OWL for Kowari, and this
continued into 2005.  I wrote the rules engine called Krule, which can
currently process inferences for RDFS.  Krule is also capable of
OWL-Lite, though the required rule set is not yet published.  I can (and
have) done OWL processing with Kowari, but it's mostly been in external
applications.  Hopefully I can have this integrated and committed in the
near future.

My current job description includes the word "semantics", but I've still
yet to work out what that means.  However, it means Herzum Software lets
me use Kowari/Mulgara, and has agreed to let me spend some of my working
time on the project.  Unfortunately, that time has recently been sucked
up into learning Mailman/Postfix/Tomcat/Jira/Confluence, but I'm hoping
that I'm past most of that now and can get back to development.  :-)

Regards,

Paul Gearon

On Jun 2, 2006, at 12:47 PM, Brian Sletten wrote:

In my experience, Eclipse has excellent support for this kind of
migration with Subversion.

If you want some help with this, let me know. If you want me to do it,
let me know.

I'll take a look at the Jira site to see what is there and start to get
going. Should we do a round of introductions? I don't know everyone who
is here.

I need to send you another htpasswd entry since I drew a blank trying to
remember it the other night. :)

--

Brian J. Sletten (brian at bosatsu.net)

GPG: http://www.bosatsu.net/gpg/brian.asc

FOAF:http://www.bosatsu.net/foaf/brian.rdf

_______________________________________________

Mulgara-dev mailing list

Mulgara-dev at mulgara.org

http://mulgara.org/mailman/listinfo/mulgara-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mulgara.org/pipermail/mulgara-dev/attachments/20060606/9df172d7/attachment.htm>