I got a call from my friend Scott this weekend. He is one of my friends that only calls when he needs computer help. As a school teacher in California, he had been given a new tablet PC, and he had lost the ability to connect to the network.

I couldn’t help him. I’ve been using Macs and UNIX almost exclusively for over four years and my Windows-fu is weak. Of course I made the usual suggestions such as reinstall the drivers, but it didn’t help.

But the fact remains that Windows is still the dominant desktop platform out there and a lot of people feel more comfortable with it than with Linux or OSX. We often get asked the question: when will OpenNMS run on Windows.

The answer is now.

A couple of months ago I got an e-mail from Bindu Dandapani at HCL in India announcing that they had managed to get OpenNMS 1.2 running on Windows. Since this was before we split out the C based libraries from the main OpenNMS Java code, that was quite a feat. They accomplished this by using some of the new Java classes in 1.5 to replace the C code we use for ICMP, so I asked them to look into 1.3 and to let us know if they could use their changes against our development branch.

Ever since we broke out the C code in 1.3.6 I’ve been wondering just how hard it would be to run on Windows. At a minimum we’d need to be able to compile jicmp, but that should be possible with mingw, and since the rest of the code is Java we should be able to “just work”. So earlier in the week I asked Ben Reed to take some time to see just how hard it would be and how long it would take.

Apparently, slightly less than a week.

As Ben’s post mentions, there are a few small things to work out, but it is definitely a salute to Java that a quarter of a million lines of code built on Unix can be ported to another O/S in such a short time.

In a strange coincidence, I got an e-mail from Bindu today that they, too, had gotten OpenNMS 1.3 to run on Windows. I’m hoping that they can help out with the last few remaining issues and perhaps help us work on an installer so that we can get the Windows port ready for prime time before the 1.3.8 release.

For many of you, this really won’t matter, since you use Linux/Solaris/OSX/etc. but the hope is that this will expose OpenNMS to a whole new group of people who didn’t want to try it before, or who work in Windows-only environments and are eager to have better tools.

I’ve decided to run January’s OpenNMS Training on Windows. Space is limited, so be sure to sign up early. Also, there is still time left to sign up for November’s OpenNMS Training on Linux. It will be the last Linux training for awhile (until probably March of next year).