We made a massive Sam’s run today. Several cases of soda, a couple cases of Red Bull, coffee, and both salty and sweet snacks. We based a lot of our purchasing decisions on last years conference, so we hope we bought enough this time.

Today both jeffg and ranger showed up. joed’s plane flight got cancelled so he doesn’t make it here until tomorrow, and Alejandro’s flight was also delayed, so instead of getting in around midnight he’ll be here around noon on Sunday. Pretty much everyone else shows up tomorrow as well.

We are hoping to tag 1.3.6 this weekend. All of the C-based JNI code has been excised from the main OpenNMS code, so now jrrd (to support RRDtool files) is optional, iplike (to allow for powerful IP address based matches when querying the database) is optional but highly recommended, and jicmp (which allows for Java to perform pings) is required but provided as a separate package.

What does this mean for OpenNMS? A number of things. First of all the opennms, opennms-webapp and opennms-docs packages are now “noarch”. This will allow us to build and release in a fraction of the time it used to take, and we will be able to provide nightly snapshots. Second, OpenNMS should build on Windows. Now, I didn’t say it would run on Windows, but it should build – which is the first step (we still need to work on any hardcoded file separators).

Since I was responsible for building on the umpteen platforms we support, I at least am ecstatic. We are also setting up some yum repos in addition to the debian/ubuntu repos, so installation should be a breeze.

In other news, OpenNMS has been appearing in a couple of blogs. One of our clients in Europe who we’d hoped could make it to Dev-Jam talks about his experiences with the app, and Coté over at Redmonk links to a post I made in the past about the glory days of OpenView.

He writes “Developers seem to have a diminished role in IT management.” This is very true in the case of OpenNMS. Unlike some other open source management projects, all of us came from a background in enterprise management, and only a few of us were “classically” trained developers. This has meant that getting to a critical mass of developers has taken us a long time, and it has made our annual Dev-Jam conference essential for the success of the project (plus it is a whole lot of fun). Most other open source projects are written by developers for developers.

On the upside, by bringing together network management professionals and not strictly developers, we’ve managed to create a product focused on their needs.

Anyway, tomorrow is the big reunion day. It’s nice being back at UMN. We went out to Big 10 for lunch and one of our favorite waitresses from last year, Meegen, was there. She’s a blast. Having an unusual name myself, we swapped stories. She won. She said that often when she tells someone her name, they don’t believe her (it’s pronounced “Me” vs. “May”), so they often ask her to spell it, as if she spelled it “Megan” they could go “aha! you’re saying it wrong”.

Well, it was funny in the bar.