Archive for February, 2007

JRobin Alignment Fix In

Wednesday, February 28th, 2007

We’ve gone ahead in the 1.3 branch and made JRobin the default storage model (versus RRDTool, but RRDTool is still available). Since the snmp-graph.properties file is in an RRDTool format, we have to translate all of the commands into objects that JRobin can understand.

When Sasha re-wrote the way graphs were produced in the 1.5 branch of JRobin, this changed the newline (\\n) functionality so that JRobin-based reports didn’t line up nicely.

I tried to fix it a couple of months ago and failed.

Craig Miskell (OGP), however, figured out that “\\\\l” would perform the same function, so I committed a small patch to trunk today that should make the JRobin reports look a lot nicer. No changes to the graph data are necessary.

Will Fraley has suggested some more changes to JRobin, but since that was to the JRobin library itself we’ll have to adopt them a little bit later.

Note that the 1.2 branch uses JRobin 1.4 and is not affected.

Also note, if you want a workaround for 1.3.2 simply replace “\\n” with “\\\\l” (four slashes and an “ell”) in your snmp-graph.properties file.

The Curse of the Upgrade

Monday, February 26th, 2007

We travel a lot, and our airline of choice is American. They have their own terminal at Raleigh/Durham (the nearest airport) and they tend to be competitive with our second airline choice, Southwest.

The benefit that pushes the needle in favor of American is their BusinessExtrAA program, which give points to both travelers as well as businesses. So we earn points good for travel and the poor soul who has to fly a lot gets ‘em too.

Now a little while ago I noticed that I had these things called “upgrade points”. I found out that I could get upgraded from coach to first class using them.

But I didn’t know about the curse.



The class of airfare when you get upgraded is “X”. For me that means “Cancelled”.

(more…)

Amen, Nat – Preach it, Chris

Saturday, February 24th, 2007

Today has been an interesting day for me. For quite awhile now I’ve been crusading against using the term “open source” as a marketing strategy. The phrase has an exact meaning, and just making the source available does not make it open. Historically there have always been ways to get access to commercial source code with the right team of lawyers and the right contracts (i.e. money), but that is a far cry from being able to change it and distribute the derivative work – two pillars of the open source philosophy.

Well today a site called Datamation published “Ten Leading Open Source Innovators” which included a large number of companies that are neither open source nor innovators.

The one that caught my eye was Zenoss. As was reported at Infoworld Zenoss is based on Nagios. I have a hard time understanding how building a product on a five year old application achieves the definition of “innovative”. Nagios was innovative, we have yet to see if Zenoss can be.

But may day was made when I saw that this article made the usually peaceful Chris Dibona irate. Chris, as the open source guru at Google, has the second best job on the planet, and he has probably forgotten more about open source than I’ll ever know. “My outage is quite present” had me shouting alleluias in the hotel lobby.

“Damn kids. Get off my damn lawn.”

Heh. Priceless.

But then when Nat Torkington over at O’Reilly jumped into the fray, I started to get my hopes up that this might be the start of a trend of really questioning those companies who claim to be open source but aren’t.

Now listen – I don’t believe that software has to be open source to be good. I doubt that Blizzard would get any benefit from making World of Warcraft open source. I own several Macs, so I pay a premium price on hardware and often pay for both closed and open software. All I care about is using the best tool for the job. My goal with OpenNMS is to create the de facto platform for both open and closed management software, and the best way to do that is by nurturing the community that builds it.

So don’t think I’m trying to disrespect SugarCRM. We use the free version, and if it works for you, great. But let’s not call it open source.

Oh, I’m in California for the second week in a row, but will be heading home tomorrow. I promise to blog next week on more OpenNMS-centric things.

“OpenNMS is a great monitoring system that continues to improve at a rapid pace!”

Sunday, February 18th, 2007

I think I might have been a little harsh in some of my comments lately, but it does get frustrating sometimes and at the end of the week I tend to be a little more brittle than normal.

However, it was nice to see a real positive article on us from SCaLE. I’m still waiting for the presentations to go on-line before posting my final thoughts.

Open Source Is Not A Marketing Term – Part Deux

Thursday, February 15th, 2007

Slashdot has an interesting article Has Open Source Lost Its Halo?”. It seems to mirror a lot of my ideas in the first Open Source Is Not A Marketing Term post.

In a post by Gordon Haff he brings up the term “predatory open source” to describe the actions by some companies to open source non-strategic products in order to undermine commercial competition.

I’m not sure how I feel about that. I have been saying for a long time that software is going to take two paths: it will either become a commodity or open.

Commodity software is not necessarily bad. The success of such things as World of Warcraft goes to show that there is a huge market for inexpensive, high-value software that “just works”.

But enterprise “anything” is hard – from network management, to CRM and ERP systems, to content management systems and databases. In these service intensive fields, solutions are best achieved by combining skilled experts with the proper software tools. I posit that it is impossible to come up with a “point and click” solution to enterprise problems.

As the goal of OpenNMS is nothing short of becoming the de facto network management platform of choice, I guess our project could be seen as being predatory to IBM’s Tivoli/Netcool, CA’s Unicenter, HP’s OpenView and BMC’s PATROL and other products, but I see it as just the natural evolution of the market. I doubt very seriously that we show up on the radars of those companies, but I bet we will soon.

The reason is that the fact that something is open source doesn’t make it successful, useful or cost saving. First and foremost it has to be better than other options. Eclipse isn’t the default Java IDE because it’s open source, it’s the default Java IDE because it’s extremely useful.

Which gets me harping on the community again. At SCaLE (presentations coming soon I’ve been promised) I had a slide paraphrasing Bill Clinton with “It’s the community, stupid”. If IBM open sourced Lotus 1-2-3, I doubt Steve Ballmer would be calling up Bill Gates screaming that Microsoft Excel was doomed.

The article mentions something else:

Rosenberg is more disturbed by the bandwagon jumpers: the companies, mostly startups, belatedly going open-source in order to ride a trend, while paying only lip service to the community and its values.

While jumping on the open source marketing bandwagon my produce a flurry of press and attention, in the long term it is doomed to fail. You can’t own a community and you can’t force a community to do something it doesn’t want to do, and in that fact is an inherent freedom. Piss off the community and they can just up and fork to another project. Try to co-opt a community and they’ll ignore you. But working with a community can change an entire industry.

I only have two rules with respect to OpenNMS:

  1. OpenNMS will never suck.
  2. OpenNMS will always be 100% free.

Luckily the community agrees with me.