The Road to Closed Source Software, Eucalyptus

« previous entry | next entry »
Jun. 23rd, 2010 | 09:44 am

It was funny to wake up to this article on "
Marten Mickos says open source doesn't have to be fully open
".

I can remember the morning of the first keynotes for the MySQL Conference after Sun had acquired MySQL. You have Jonathan Swartz and Rich Green delivering keynotes where the underlying message was "we continue to allow MySQL to run its own business".

Why was this?

Because Marten was going to announce the close sourcing of part of the MySQL Server. For years there were conversations around "if we did XYZ could we take out a critical...". These conversations were always met with a dead silence. The codebase was neither modular, nor did any of the developers resonate with the message. The backup code had never been designed to be a standalone component so the entire message of "we are close sourcing it" was a delusion. We had no ability to do it.

The reaction from the audience at the conference?

One big giant boo'ing, and later in the day a talk was given where the audience chanted "we will not ship crippleware".

Marten mentions Apache as being a failure at monetization, this patently false. I know several of the Apache authors well, and they are doing just fine. They made plenty of money, and the Apache project has brought wealth to many individuals. Sure, not a concentrated wealth to an individual company, but when I look around to the Yahoo's, Google's, and other companies that made use of open source software, I see plenty of wealth that was created. MySQL fell into a lucky path where it got sold for a billion dollars, but the real wealth in MySQL was in all of the companies that made money from using it (check out the article by Stephen O'Grady on the The Economics of Open Source: Why the Billion Dollar Barrier is Irrelevant).

I am curious to see what is going to happen with Eucalyptus, if anything. It has been no secret in the open source world that the Canonical developers have had no love for the processes and policies around Eucalyptus, and Mark has said he would like to see more alternatives. I have been told that there isn't really all that much to Eucalyptus, so for Canonical to decide to just fork it wouldn't be difficult.

I cannot see the Ubuntu world just blindly accepting crippleware into their stack, that seems like it would be against the basic tenants of the community.

A final thought on the "open core allows the creation of more open source code". I have yet to ever see this actually happen. More often some piece of code makes a very small pot of money, not often paying for the development or hassle of its maintenance, and then there is an internal rush inside of the company to monetize every nugget of functionality in the software.

Redhat? The company is open source, not open core. The MySQL database? It was open source, not open core.

There has been no successful launch of an open core company that has reached any significant size, especially of the size that Marten hints at in the article.

Open Core is a dead on arrival business model. We should realize this and just move on.

Link | Leave a comment | Share

Comments {1}

It's funny

from: patrick-crews [launchpad.net]
date: Jun. 23rd, 2010 11:45 pm (UTC)
Link

When he mentions Apache that he talks about how there are no VCs or CEOs as a lack of success.

An ungenerous reading makes the term 'carpetbagger' spring to mind.

Part of what makes Drizzle and Rackspace so appealing is that success depends on the ability to deliver excellent software that people want to use - providing service rather than charging people just to get in the door. Expertise over legalese! Of course, these are just my $.02

Reply | Thread