[LWN Logo]
[Timeline]
From: announce-admin@opennms.org
To: announce@www.opennms.org
Date: Wed, 27 Sep 2000 00:10:48 -0500 (CDT)
Subject: [opennms-announce] OpenNMS Update v1.27

================
 OpenNMS Update
================
Vol 1., Issue 27
================

   In this week's installment...
     * Project Status
          + An Old-Fashioned Cyber-Revival
          + Planning Session: Transactions
          + Coding Projects Underway
     * Still Pending Announcements...
     * OpenNMS: A Clearinghouse for Apache Project Technologies
     * The Wish List
       
==============
Project Status
==============

An Old-Fashioned Cyber-Revival

   It was a good ol' Sunday mornin' tent meeting, well, except for the
   fact that it was on a Tuesday afternoon, in a Hampton Inn, and the
   religion we were gettin' was all Java and XML related. Close enough.
   In reality, as a part of a planning session we had scheduled, we were
   lucky enough to be joined by Doug Tidwell, IBM's "Cyber-evangelist"
   and author of O'Reilly's upcoming book on XSLT. (Yes,
   "Cyber-evangelist" is his real title) Doug dropped by to give us a
   hand on some of the finer points of technologies of some of the tools
   from IBM's Alphaworks site, the Apache Project, and some of the stuff
   he's been working on as well.

   As far as specifics that came out of the meeting, let's just say that
   there were a lot of technologies discussed. Some of the details or
   specific problems that Doug was able to help out with included:

     * Doug helped us to bridge the gap on some of our XML - > FO - > PDF
       conversions. With this, jreporter should see some life very soon.
     * We got an overview of some of the things coming out of the
       upcoming Cocoon2, including some on-the-fly conversion tools that
       might eliminate jreporter altogether.
     * Some real world performance data on DOM vs. SAX, and now
       motivated, we'll be looking a little harder at SAX than we have to
       date.
     * A discussion on namespaces, SAX2, Xalan/XSLT, and amazingly
       enough, an ASN.1-to-XML converter. (Woo-hoo!)
       
   Doug's one of those guys that not only knows his stuff, but can talk
   about it at every level. I saw his presentation on Cocoon at the
   O'Reilly Open Source Conference this year (downloadable from
   ftp://ftp.oreilly.com/pub/conference/oscon2000/699/Tidwell.zip) and
   worth a read if you are interested in Cocoon at all.

   A somewhat more appropriate read for some of the network management
   non-coders amongst us, take a look at Doug's tutorial on XML
   Programming in Java at IBM's DeveloperWorks site. You have to
   register, but it's free, well-written, and a great way to come up to
   speed on some of this stuff fast. You can find it at

   http://www2.software.ibm.com/developer/education.nsf/dw/xml-onlinecour
   se-bytitle?OpenDocument

   Big thanks to Doug Tidwell and to IBM for heartily supporting
   Cyber-evangelism! Hey Doug, will you sign our copies of the XSLT book?
   
Planning Session: Transactions

   During the non-Tidwellian part of our meeting, we were focused on
   defining some of the transactions that would need to occur betwe en
   the distributed poller and the master station to support the
   distributed architecture. We came out of the meeting with a finite
   list of transactions that we'll be supporting (via a combination of
   servlets, the service control manager, and events).

   The good news that came out of this session is that we were able to
   pare the list down to just a handful of required transactions, and by
   re-using the basic transactions, we should be able to get all the
   necessary data back and forth between the DPs and the MS. Some of the
   key things we needed to focus on were informing the Master Station of
   new nodes/capabilities discovered, event forwarding, and configuration
   distribution.

   We've also kicked off a "proof of concept" test on development using
   some of the SOAP tools from the Apache project. If everything goes
   well there, we'll likely be propagating SOAP throughout the
   architecture. If not (or worse, it brings significant development
   overhead), we'll likely just pass our own XML streams back and forth.
   Care to weigh in on the SOAP vs. non-SOAP debate? Drop us a line on
   the [discuss] mailing list .
   
Coding Projects Underway

     * create.sql -- Transaction discussions have prompted a need for a
       few more fields in the database, including better identification
       of events, and some critical timers for the distributed poller
       table.
     * Events -- Jacinta is our "proof of concept" SOAP guinea pig. She's
       our SOAP woman, so in turn has been dubbed "The Cleaning Lady". I
       figure that one of these days she's going to clean my clock.
     * SCM -- Integration work continues. We are starting up distributed
       processes, but need to improve the implementation before we are
       gold.
     * icmpd/Perl -- New version of Net::RawIP out. This should help in
       the development task a little bit. Freakin' memory leaks.
     * jReporter -- Development underway. Sowmya's going to be working on
       some of Doug's suggestions, and ignoring some of the well-intended
       but confusing documentation accompanying Xalan.
     * Filters -- Moved into testing. Jason is working with Ben to verify
       that we get what we think we should be getting.
     * Java MIB Compiler -- Lee has stepped up to the plate on this one,
       but per the info we got today on a possible ASN.1-to-XML
       converter, we might be changing pitchers before he gets his cuts
       in. I'll keep you posted.
       

==============================
Still Pending Announcements...
==============================

   Last week I mentioned an upcoming announcement and was vague about the
   time it would be released.

   Allow me to take this time to be a little less vague.

   Thank you.

   Look for that announcement on Thursday morning (knock wood). The
   announcement itself will be posted on the web site, but I'll send a
   note to the [announce] list to make sure you know when to look.

   Jeez. Do I take care of you or what?
   

========================================================
OpenNMS: A Clearinghouse for Apache Project Technologies
========================================================

   We've made several references to various technologies, each with a
   seemingly more obscure name than the last, and most of these originate
   from the Apache Project, and more specifically, from the Apache XML
   Project at http://xml.apache.org.

   We're using a lot of these tools for a couple very good reasons.
   First, these are the de facto st andards for Java/XML in the open
   source world, and second, there are good integration points between
   most of the tools, so if you use one, it's easier to use others.
   So I just wanted to take a moment to point out some of the components
   we are using from the Apache Project, where we are using them, and
   provide a public sign of our gratitude for their contributions to the
   community.

   Straight outta Compton, here's what we are using and where:

     * Xerces, our XML parser, and we use it throughout the Bluebird
       software.
     * Xalan, an XSLT stylesheet processor, used in the jReporter.
     * Cocoon, for XML-based web publishing, and we're kicking it around
       as the next-generation jReporter.
     * FOP, XSL formatting objects, used in the jReporter.
     * SOAP, Simple Object Access Protocol, and we're using IBM's
       reference implementation which was donated to the Apache XML
       project (and has since been enhanced and extended).
     * Some non-Apache tools we're using include Cup, a parser generator
       and JLex, a lexical analyzer generator.
       
   So why bother mentioning all these? Because we've got a whole lot
   going on, and we're able to do some pretty impressive things pretty
   quickly, but only because tools like these already exist. And man, are
   we glad they do.
   

=============
The Wish List
=============

   OK. So we've stolen a lot of code. That just keeps us on track with
   one of the three virtues of good programmers: laziness. But then
   again, if they wrote it, why should we have to? Dontcha just love the
   GPL.

     * Add SNMPv3 functionality to jSNMP
     * Implement the event correlation spec.
     * Review the database design
     * Consider the integration points for CIM/WBEM support
       
   My previously documented problems with IBM's JDK 1.3 for Linux were
   incorrectly attributed as user error last week. It turns out, there
   were some changes in the behavior of a JTable method from 1.2.2 to 1.3
   that toasted one of our configuration panels. A few slight changes and
   this one is addressed. No prob, Bob.

   So keep this in mind if you are testing on Linux: With 1.2.2, you'll
   likely have drag/drop problems. With 1.3, on some of our panels, you
   won't get far enough to HAVE drag/drop problems.

   Remember, forewarned is forearmed. And after all, where would we be
   without forearms?

   This issue is brought to you by the OpenNMS Olympians--Proud to be
   testing positive for as long as we can remember.

Later,

Shane O.
========
Shane O'Donnell
shaneo@opennms.org

_______________________________________________
announce mailing list (announce@opennms.org)
To subscribe, unsubscribe, or change your list options, go to:
http://www.opennms.org/mailman/listinfo/announce