Oracle is planning to deactivate Kenai for community developers. Yes, it is sad but let’s move on and spread our code before it is destroyed. It is also time to make public some ideas I’ve been sharing with Sun’s folks last year. I have criticized Oracle for some decisions they have made, that’s true, but I don’t want to criticize and shut up. I have to propose something better and realistic to show that they are lazy thinkers, growing by buying other people’s innovation, not actually doing their own.
I want to discuss 2 ideas. Each one solves a particular need of entrepreneur developers.
I need integrated services to manage my software project and also to deploy it: It would be a mix of SourceForge with Amazon EC2, or Assembla with Google App Engine. Once my code is committed to the server and an integration test is performed, then my application would be automatically deployed in a server. Today, we need at least two different service providers to make it happen and it means more bureaucracy and waste of time. If Oracle or RedHat adopts something like this, the payment for such services would financially support their open-source projects, such as OpenSolaris, Glassfish, JBoss and MySQL.
I’m an expert in my own open-source product, but I need help with its infrastructure: I have a complex and large JEE5 application and my expertise allows me to provide good support in terms of application features and bug fixing. As a matter of fact, my application is using Glassfish as application server and MySQL as database. However, all I know about these technologies are enough to develop and deploy the application. If a critical error occurs, avoiding the application to run because of an external and unexpected problem, probably my own knowledge is not enough. In order to solve that, Oracle or RedHat would make a partnership with developers, providing services that complements developer’s services. Considering RedHat, the customer would pay directly to them for the services and RedHat would transfer part of the payment to the developer, who provided part of the service too. With this deal, every developer becomes a potential retailer of RedHat’s services. That’s one more solution to increase investments on open-source projects.
Both ideas above were explained to Sun engineers some months ago (At the time I was thinking that IBM would buy Sun :D). To sell open source products, companies should innovate, however Sun was good when innovating their products, but not their selling and services processes. These parts of their business were always traditional. RedHat, on the other hand, has far better selling and services processes for their open source products than Sun because they are agile, have simple procedures and minimal distance between the customer and the specialist. That’s why they are one of the only companies to get profit from open-source.
Companies that already offer software project management (track system, version control, wiki, etc.), such as Atlassian, Collabnet, Google or even Oracle with Kenai, have strategic advantage because they already have something done, with large experience on it. I only wish that people out there get this idea and make it happen. This is not what I love to do, so I won’t appropriately exploit it. But I’m confident that these are good ideas and I hope to blog about any future service that provides such features and I’m probably going to be one of the first customers of the pioneer.