[LWN Logo]
[Timeline]
From: "Shane O'Donnell" <shane@www.opennms.org>
To: "General@Opennms. Org" <general@opennms.org>
Subject: OpenNMS: OpenNMS Update v1.20
Date: Tue, 8 Aug 2000 18:01:22 -0400

================
 OpenNMS Update
================
Vol 1., Issue 20
================

In this week's installment...

 - Project Status
   - Correlation Engine Spec
   - JSDT Update
   - Documentation Updated
   - Coding Projects Underway
 - New Technologies Under Review
 - The Wish List

==============
Project Status
==============

Correlation Engine Spec:

It is with extreme excitement that I announce that we have received the
initial draft of a PHENOMENAL spec for a correlation engine which will, at
some point in the future, bolt on to the Bluebird event subsystem.

The spec is currently in review and is being proofread and technologically
adapted to eliminate some of the overlap between the spec and the existing
event subsystem spec.

Countless thanks go out to Doug Stevenson for creating the hefty tome (21
pages and counting) which quite obviously draws on Doug's years of
experience with other, lesser correlation products. Once built, this engine
engine will be a reference for commercial products, and if instrumented
correctly, could replace many of them. I cannot sing the praises of this
document highly enough (without an operation, anyway.)

But I'll let you judge for yourself. The spec will be available later this
week at http://www.opennms.org/files/code_specs/maji

It's reading documents like this that make me remember why I got into this
business in the first place. It's great working with people you can learn
from.


JSDT Update:

I wish the news was better, but we'll take what we can get...

Latest word from Danese Cooper (Manager of Licensing at Sun) is that there
is a license in the works for us, but nobody seems to know the status. She's
checking on this and we should have word this week.


Documentation Updated:

Effective earlier this week, Steve was able to get some of the more recent
architectural changes to Bluebird reflected in the on-line documentation.

Specifically, the Master Station Design Config and Control has been revved
from 1.0p to 1.0q, as well as converted (courtesy M$ Word) to html format.
Also, the Real-time Console Design Control and Config doc has gone from 1.0g
to 1.0k (don't ask).

To access these documents, go to the web site, sign in, select "Developers
Portal" and then click on "Developer Documentation" on the left.

Give these docs the once-over. Good stuff, Maynard.


Coding Projects Underway:

Eventd & Event Correlation Spec -- One rev of eventd is done. Next step is
to read, parse, and format per eventd.conf. Event Correlation spec is in
review, per the blurb above.

discovery/capsd -- Internal testing is complete. We are now waiting on Weave
who, following a weekend of who-knows-what, went belly-up with a stomach
virus. Informed sources tell us that we won't see him until late this week,
if at all. Hey sicko! You're cramping my productivity over here!

AWT <-> SWING Conversion -- Still in preliminary phases. Waiting on an
update from the community.

icmpd/Perl -- The porting team spent last weekend at DefCon and as of this
writing, were _still_ not sober enough to respond.

Service Control Manager -- Task assigned. Development will soon be underway.

Bug fixes for jReporter -- Task assigned. Development begins tomorrow.

Fred Reimer dropped us a note last week that a bug he had uncovered with
Sun's Java2 JRE for Linux has been closed. This was bug # 4185644 (for those
of you keeping score) and was related to a drag/drop problem in the GUI. The
fix is not yet available on all platforms yet, so if you've stumbled across
this, you may want to keep your eyes and ears open for the next Java2 drop
from Sun (or just switch to IBM's JDK for Linux) ;-)

Props to Fred for keepin' an eye on the Man!


=============================
New Technologies Under Review
=============================

It's crunch time for some of the final architectural decisions and we wanted
you to be aware of what we are currently thinking...

For Java-XML middleware, we're looking long and hard at SOAP. It's still
only in reference implementation, but it's been adopted by the Apache
project and has industry support from IBM, Micro$oft, and Sun (kinda). Looks
safe and does what we need it to... You can check it out further at
http://xml.apache.org/soap/

Out of curiosity, we've been watching OpenJMS (and not just because of the
name similarity). They are building an open source implementation of Sun's
Java Message Service API 1.0.2. You can check them out at
http://www.openjms.org

This is old news, but the bbparser (Bluebird Parser), written by Sowmya,
extends XML4J (from IBM) which eventually became Xerces (part of the Apache
project). You can find more info on this at http://xml.apache.org/xerces-j/

And last, but certainly not least, we've been reviewing CIM for potential
inclusion in an upcoming release. We agree that CIM is slowly becoming a
more important management technology, but we're strapped for resources,
especially given our current scope for release by the end of the year. It
will take significant community involvement to get this built soon. I've
solicited Andrea Westerinen, co-author of THE book on CIM (which I now own
and have since read and partially understood) to provide some input on how
and where CIM can be implemented in Bluebird for maximum impact and minimum
delay. Our inclusion of CIM support will be largely directed by her input.
You can "read all about it" at http://www.dmtf.org/educ/


The Wish List

  "I wish I were an Oscar Meyer weiner,
   for that is what I truly'd like to be.
   'Cuz if I were an Oscar Meyer weiner,
   everyone could send their bytes to me."
           -- Popular Jingle, adapted by me.

OK. Get ready. We're overhauling. Seeing what works and what doesn't.
Separating the wheat from the chaff, the runts from the litter, and the men
from the sheep. Here we go!

 - Java coders need to grab the event correlation spec and bring it to life.

 - Hopefully, Mr. Stomach Virus is working on the Poller Spec this week.
Watch this space for further updates and abuse.

 - CIM wonks to help us figure out how this fits. This means architecturally
as well as code (and database work). If you've seen the CIM spec and our
database, you won't keep beating me up about this not being a database
issue. It is.


In closing, let's just say that I think Lieberman is an OK guy and all, but
if Gore was looking to spice up the ticket, he would've been better off
picking his nose. At least that would have garnered some ongoing press
coverage.

Wake me when the election's over,

Shane O.
========
Shane O'Donnell
OpenNMS
shane@opennms.org
===================
It's Revolution, not Evolution
at http://www.opennms.org !!!!

= = = = = = = OpenNMS.org List Message = = = = = = = = =
To unsubscribe from this list, send a message with a body of
"unsubscribe general <email>" to majordomo@opennms.org.