|
|||||
|
So DevJam officially ended Friday, and now it’s time for LinuxWorld. If you’re here in San Francisco, please, stop by the .org pavilion and say “hi!” We will be showing off OpenNMS 1.3.6 and hanging out with the other über-geeks. 😉 Share on FacebookSo I’ve been having a lot of fun at my first OpenNMS DevJam. We’ve had a lot of fun geeking out and I’ve had a chance to work on some neat bits of code. First of all, we released OpenNMS 1.3.6, which as I mentioned before, removed all of the JNI code from the build, simplifying packaging and distribution immensely. In addition, we’re now working on embedding Jetty so it won’t be necessary to install a separate servlet container like Tomcat to be able to use the web interface. While I was at it, I implemented a pet project, Bonjour service advertising — so your OpenNMS web instance will be easily located from any Bonjour-capable web browser, like Safari or Camino. There’s also a lot of cool work going on with a Google Web Toolkit UI that’s looking very promising. I’ve been very impressed by the facilities provided by the University of Minnesota. The dorm rooms are much nicer than what I had back in the day, and our conference area is great — blazing fast . . . → Read More: DevJam Hacking Share on FacebookAs of yesterday, Brozow and I (and with the help of others) have completely exorcised the C code from the OpenNMS codebase. That is not to say that it isn’t required, we still rely on C code for performing ICMP pings (java has no raw socket support), but at least it is split out into it’s own package, which makes packaging much easier. Now, instead of building umpteen different RPMs for different platforms every time we make an OpenNMS release, we can make a single .noarch RPM for all platforms — only updating the platform-specific code when necessary, which is rarely. That code has barely changed over the years, so chances are, we won’t have to do it again for a while. Thanks to this, it is now easy to create a yum repository with this stuff, so I have gone ahead and done so: It only contains stuff for RHEL4/CentOS4 and RHEL5/CentOS5 right now (i386 and x86_64 versions), but the intention is to expand it to all supported RPM-based platforms. Share on FacebookOpenNMS 1.3.5 is now in Fink 10.4 unstable. It’s mostly a bugfix release, with a few changes to notifications and thresholding. For a general idea of what’s changed, see the New and Noteworthy page at the OpenNMS wiki. For more details, see the full release notes. Share on FacebookIt’s been a crazy month, and it’s gonna keep getting crazier. June 11th I went to WWDC. Then, last weekend, I was in New York for the Origami USA convention. Now I’m headed out to Glasgow to speak at aKademy. Then, in July it’s off to the OpenNMS Dev-Jam, and immediately after that, LinuxWorld Expo, where we’ll be in the .org pavilion. Phew! So anyways, if you’ll be at any of those events (heh), drop me a line! Share on FacebookIf you have not seen the announcement yet, OpenNMS 1.3.3 was released. It includes lots of bugfixes, as well as a start on separating out the native C code so that packaging a pure-java release of OpenNMS will be easier. We hope to have that wrapped up by the 1.3.4 release. I’ve released 1.3.3 into 10.4 Unstable, as well as the new iplike package for PostgreSQL. Unfortunately, I had to remove OpenNMS from 10.3 because the 1.3 series requires Java 1.5. I will probably go back and revert 10.3 to the latest 1.2 version, just so that something is available still. I also spent some time reworking the OpenNMS front page to be more useful for finding information. It was a pretty big mess of random links without much navigation to help people get to the info they need. There’s still more to do, but it’s definitely easier to figure out where to go now. The only problem is, enough people can actually see the link to the demo site that the demo site is hitting tomcat’s maximum connection limits. <grin> Hopefully that will get fixed up shortly, but . . . → Read More: OpenNMS 1.3.3 Released Share on FacebookSo we’re in the process of trying to get OpenNMS 1.3.3 out the door. There are still a few bugs that are considered blockers but for the most part trunk is looking pretty nice. One of the big things that’s getting a lot of attention is packaging. I’ve autoconf-ized iplike and we’re going to do other work to get all of the native code out of the main java build as much as possible. Not all of that can happen in the 1.3.3 timeframe, but at the least we want to chop things up so there’s only a single opennms RPM, and then platform-specific packages for only the native code. That should make updates considerably easier, and open up the door for nightly builds and other spiffy stuff. W00t! Also, a reminder: I’m going to be in San Francisco this week, at WWDC. See you there! Share on FacebookI finished packaging the latest OpenNMS in Fink over the weekend. It’s now split into 2 packages — opennms (the main OpenNMS code), and opennms-pgsqlXX (the PostgreSQL database-specific code). You can now use OpenNMS with any version of PostgreSQL you like, just install the proper pgsql entry (after optionally dumping and migrating your data), start up the proper version of postgresql, and you’re set. Also, I spent some time figuring out how the Debian packages are built, and updated them to 1.2.9 in the OpenNMS APT repository. Share on FacebookSo I started my new job with OpenNMS monday, and after the requisite “help us clean up the support ticket queue” work that any newbie should be forced to go through (grin), I’ve been trying to get the latest OpenNMS release working in Fink. I must admit, despite the fact that 1.3.2 came out in January, I’d been putting off updating it because they it had been converted to maven. Now, maven is awesome. However, it’s not very compatible with the “build offline” way that packaging works in Fink — not to mention a number of other packaging projects. I’ve figured out the magic to make maven treat a downloaded tarball of dependencies as a repository, and have everything building/working now. (Yay) I’ll put it in Fink tomorrow, after some testing, but it’s looking good. Share on FacebookSo that was quick. It turns out I got an offer I couldn’t refuse. In an odd twist of fate, I will be working on the OpenNMS code base, for the third time. (This one’s the charm, right?) I will be doing community-facing stuff, as well as helping out with all the little things anyone has to do at a very small company. I get to do what I love, which is: a little bit of everything. 😉 Thanks to everyone for their support, I guess I’ll see you folks in #opennms on Monday! (OK, technically I’m already there, but man, that’s not as poetic.) Share on Facebook |
|||||
|
Copyright © 2024 Tales of the Raccoon Fink - All Rights Reserved |
|||||
Comments are closed 