What is desktop integration? What does it mean? Is there a right way to do it? And a wrong way? Who are the stakeholders? And do the stakeholders communicate and if so, how?

In GNU/Linux, we probably have the best CPU scheduler around, the most awesome virtual memory subsystem, a kick-ass IPC system but at the end of the day (whether it's fair or not) many many people dismiss an OS unless things Just Work™. Yet, other people dismiss the OS if they can't configure exactly the way they want. And that's why we have many different desktops and OS distributions. Sometimes each with their own verions of the same bugs.

Desktop integration can be characterized as defining, providing and maintaing interfaces to make a wide range of experiences easy to implement and maintain. As the desktop is the top-level user interface for a wide range of users, desktop integration is thus concerned with ensuring that proper interfaces are available. One important part that we all should be using the same code (thus benefitting from each others bug fixes) or at the very least the same interfaces. Desktop integration can be thought of as a way to avoid huge incompatible silos.

In this MC I'd like to attempt to answer some of these questions, find some common ground and come up with recommendations on how to improve the eco system around the free desktop. I'd like the MC to be very specific and focus on real world problems.

The MC will start with presentation (I'm planning to give it) that gives an detailed overview of desktop GNU/Linux anno 2008 with respect to the current state of the art of desktop integration and what's planned. This will include a brief account of all the shared code and specifications used by the two major desktops (GNOME and KDE).

For the second presentation I think I've convinced William Jon McCann to give a presentation about the session management, login management and authentication problems he's working on. This includes making things like fast-user-switching, multi-seat, terminal services and modern multi-factor authentication (this includes an awesome rant about PAM) Just Work™.

After that I'd like to have a round-table discussion where we identify the most urgent (or concrete) topics to focus on and we'll form 2-3 breakout sessions to work on these. Then we'll conclude with with short presentations from the breakout sessions.

Total time: 2 hr 30 min

Kernel hackers, core userland software maintainers, desktop applications developers, with a keen interest in improving the desktop experience on GNU/Linux is encouraged to show up.