The first technical talk of the 10th Plone conference is by Jan Jongboom, who works for Cloud9.
Writing synchronous code is easy. Asynchronous code however… A simple thread on a Linux system uses 2Mb. And every time someone does a request, a thread is started. This can become a problem when there are requests that take a long time. And if e.g. Apache is limited to a certain amount of memory, you are limiting the number of possible requests. With LibUV this is managed so you don’t need the 2Mb per thread. So if you have a web application that requires a lot of concurrent connections, you are no longer limited by the memory usage of the threads.
With Cloud9 you get a free Linux OpenShift VM. Including a real terminal. But the coolest feature is collaboration. Someone else can come into your IDE and see the same things you see. He/she can see your cursor and see you type. (Also: If you close a file, the files is also closed in the IDE of the other person.) So you can easily debug and work together on something. You also don’t even have to leave your browser to deploy your code.
Plone doesn’t run on the OpenShift VM yet because of a LibXML problem. But you can also bring your own server. So Jan used an Amazon EC2 instance for his VM. The Plone Unified installer ran fine there. He can just open and edit the files (e.g. the buildout configuration files) from his browser.
Cloud9 is very proud of their code completion. It should become available for Python as well in the next few months.
Since Cloud9 in open source, you can fork the GitHub repository and contribute if you want to. Or join Cloud9 since they are hiring.