Systems management, what will happen when the VCs want their money ?
Tarus is happy not to have VC's on his back. He doesn't want to be responsible for turning a 15 Milj investment into a 150 Milj cashout. Others chose to go that way.
Back when he wrote the article the chances were small that he already knew that Qlusters was going to be shut down with still sooo much money in the bank, but the VC's wanted it back.
So how do open source companies plan on making those tenfold roi reality.
Apart from selling out to a bigger company I think thats a very difficult task.
Especially when you keep in mind how to manage both the Open Source community and your customers. The figures he mentions that VC's require surely start pushing vendors into violating Fabrizio Capobianco rules.
Now the story changes when indeed you cn go to a model where you are selling a large scalable service to your customers, even with microsized payments it becomes a possibility, but that's a totally different business model from what e.g. the Open source systems management shops are doing .
So will the Zenoss, Hyperics or the Groundworks of this world survive the demands of their VC
Luckily these projects are Open Source, so when the company dissapears, the project can continue, and grow even better. Like openQRM did
Comments
#1 Tarus Balog : But who gets the icing and who gets the crumbs?
The problem with Mark's example is that the Zenoss model is not open source. The code is controlled by a commercial entity, and it is that entity that decides what is and what is not added to the "open" part of the project, not the community.
If the community wants or needs a feature that is only available in the proprietary version of the product, their choices are limited. They could write it themselves, or they could fork it.
Unfortunately, neither of those options are usually viable. Supposed the reason most people paid for the closed source product was Feature X, and some enterprising person decided to write that feature. Do you think it would be accepted and committed to the "open" part of the code. I doubt it. If the powers that be thought that feature should be free, they would have opened sourced it themselves (thus saving everyone the time to rewrite it). The reason they don't is that it drives software revenue, and at its heart Zenoss.com is a commercial software company. Even if people inside the company wanted to accept the code, I doubt the investors would allow it, especially if it negatively impacted the bottom line.
The other option, forking, is rarely done (Joomla v. Mambo is the only example I can think of). You've got to have a serious commitment to do this and these proprietary companies know this. Plus these proprietary software companies have groups of lawyers who will be on top of it, thus discouraging the casual person from taking this route.
Of course they could also buy it, but the code they purchase is not free and they are not granted the freedoms that come with true FOSS.
My point is that I get tired of the "ignore the man behind the curtain" tact that "commercial open source" companies take. They want all the advantages of a vibrant community without giving up anything. Not that there is anything basically wrong with this: HP did this for years with OpenView and the OpenView Forum. But I dislike their use of the label "open source" when it has little to do with the true spirit of free and open software.
I was delighted to see on Larry Augustin's blog that Europeans as a group understand the difference. I think this is one reason OpenNMS is doing so well overseas (at the last training session only two out of seven people were from the US). We will continue with our determination that OpenNMS will a) never suck and b) always be free, and I'm betting the farm that this strategy will work in the long run.
#2 Mark Hinkle : Not Again
I could debate this until the end of time with the OpenNMS group but I would simply say this. Last time I checked the OSI defines open source licenses not Tarus Balog and it's a software development methodology not a model. I can only quote The Bard, when I say, "The lady doth protest to much, methinks."
Zenoss Core is supported by an open transparent community and our software is licensed under the GPLv2 and OSI-approved license. It's good software. However, I am just choosing to quote many of our non-paying users/members. We do accept and post many pieces of code and post them on our website for anyone to see even code that compete with our commercial offerings.
Our commercial offerings are all centered around services and technical support. There are also certified monitors and other extensions called ZenPacks that we develop and sell as part of those packages. Recently we ran a contest were we challenged our community to develop and share their ZenPacks. Many of those ZenPacks are competitive with our commercial monitors and we got no benefit other than positive energy in our community.
We also announced that to all our community members (including those that pay money for our proprietary extensions) in our monthly newsletter seems pretty transparent to me. I just am not smart enough to have some devious ulterior motive.
Really there's little difference here. Doesn't the privately-held OpenNMS group "control" the code of their project through their commits? Despite that I would take Tarus' word that he will always try to do the right thing by his community of users. I have no reason to believe that he wouldn't. I am not sure what is the source of his venom and why anyone would imply we would act differently.
Arguing with someone who's software does similar things is just a waste of time. I will just agree to disagree. So here's the advice I give anyone who reads this tiresome and reoccurring thread. Go to Sourceforge download the software of your choice and if you need help with it I am sure there's a community and/or a company that can help you out we are all easy to find.
#3 Benjamin Reed : Actually...
Not to belabor the rest of the argument, I'd like to clarify: the OpenNMS Group is out-voted by the other OGP members. If they disagree with the direction The OpenNMS Group wants to take OpenNMS in, they have the right and the ability to override us, as well as take over the reins of the project if something were to go wrong and TOG disappeared. We "run" the project because they trust us, and much like the Trolltech's "Qt goes BSD if we do something bad" license, our community has an exit strategy. OpenNMS can be picked up by someone else and continued as if nothing happened.
If Zenoss, Inc. disappears, the community will spend a long time picking up the pieces and reimplementing things that have gone "missing." That's the real issue, in my mind.
#4 Tarus Balog : The Cake is a Lie
I noticed you said "Zenoss Core" and not "Zenoss" is licensed under an OSI approved license. My comments were not directed at that bit of your software that is open source, but at the motivations behind a software company with US$14 million in investment and a business model built on selling software under non-OSI approved licenses.
Both Microsoft and Google release a lot of code under OSI approved licenses (more than Zenoss and OpenNMS combined I would bet), but neither of those companies would claim to be "open source" companies like Zenoss does. Just because you say it over and over again does not make it true.
If you will publicly state that Zenoss, Inc. does not make money selling commercial software licenses, and that all of your revenue is from services and support, I'll gladly shut up. Heck, if you say that less than 10% of your revenue comes from software licenses I'll shut up. But I'm betting you can't say that, because it's not true.
If you believe so much in open source, why not jump in 100%? Stop using it as a marketing term and commit. Stop making your community duplicate your work. Then you'll see some serious positive energy, even from me.
#5 jeffg : Elbows off the table, you're eating that cake in a glass house!
Mark Hinkle wrote:
Also OpenQRM was released under a permissive though not really open source license (the OpenQRM license was the MPL with an attribution clause so while it was a good step not adherent to the open source definition).
What's Zenoss licensed under, then? You did not say "Zenoss Core", so I take your comments to pertain broadly to both the GPL-licensed Core and the commercially-licensed, more-fully-featured Enterprise version. Under that nomenclature, Zenoss turns out to be quite a bit further from adherence to the OSD than is MPL-with-attribution.
#6 Mark Hinkle : Sometimes You Can Have Your Cake and Eat it Too
Qlusters is an example of where a venture-backed company came first and then the open source software was released. Also OpenQRM was released under a permissive though not really open source license (the OpenQRM license was the MPL with an attribution clause so while it was a good step not adherent to the open source definition). Luckily guys like you and Matt are continuing work on OpenQRM and I believe will it will be a successful project for a long time to come without the VC backing. Though as I analyze Qlusters I don't think it's a good comparison to Zenoss.
At Zenoss we have and continue to build an open source project and a community that is symbiotic with a company. We got the OSS project rolling before we took investment from VCs. Also, I would argue that being a good open source software developer and being a good investment for VCs is NOT mutually exclusive. If you do one right it's usually good for the other and vice-versa.
At Zenoss we took VC and we do have obligations to that end - but everyone has obligations. As an open source project leader you have obligations to your community to try to develop good software and include them. As a business we have obligations to our employees and our shareholders. Luckily, our investors realize the reason JBoss, Red Hat, MySQL, and others were so successful was because they had a good product developed in cooperation with a community.
Responsibly growing and positively interacting with our users is to open source success. And that's the mandate we have from investors -- which has translated to the financial success they expect. They haven't pressured us to do anything unnatural to recoup their investment. I believe our investors know that the biggest asset is the community and that doing anything to alienate them puts their investment at risk.
Bottom line -- to me at least -- is that if the open source project is good, the community around the project is healthy and happy, and the business associated with it is thriving then whether they have venture-backing or not is a moot point.