The father of Microsoft's Singularity OS moves on

James Larus, a 16-year Microsoft veteran and the man who started the Singularity operating system project at Microsoft, has left the company.

Larus will become the new Dean at the Ecole Polytechnique Federale De Lausanne (EPFL) School of Computer and Communications Science (IC) this October.

Larus was a Principal Researcher who contributed to the programming languages, compiler and computer architecture fields inside and outside the company. He joined Microsoft Research in 1998 to start and lead the Software Productivity Tools (SPT) group, according to his bio. He subsequently become a Research Area Manager for programming languages and tools, where he started Singularity, a Microsoft Research operating system project.

Singularity is a microkernel operating system and set of related tools and libraries that is developed completely in managed code. Singularity was not based on Windows; it was written from scratch as a proof-of-concept. Microsoft made the Singularity source code available for download under a non-commercial, academic license back in 2008.

Microsoft describes Singularity as a research project "which demonstrated that modern programming languages and software engineering techniques could fundamentally improve software architectures." Singularity ended up spawning and/or influencing a number of other operating system research projects at Microsoft, including Barrelfish, Helios, Midori and Drawbridge.

Later in his career at Microsoft, Larus helped start the eXtreme Computing Group (XCG), which is a group in Microsoft Research developing hardware and software supporting cloud computing. In XCG, Larus led groups developing the Orleans framework for cloud programming and various computer hardware projects.

“Jim Larus has taken a new position at EPFL, a top academic institution in Switzerland with which we have had many collaborations over the years," said a Microsoft Research spokesperson via an e-mailed statement, when I asked about Larus. "As Microsoft Research’s Faculty Summit (last) week demonstrates, our connections with academics around the world are very important to us. We are proud that such a well-respected institution would look to one of our own for such an important role. We wish Jim all the best in his new position, and we hope to continue working with him and his colleagues long into the future.”

Microsoft officials often publicly downplayed (at least publicly) the significance of Singularity -- not wanting company watchers to consider a research project a threat to Windows, one of Microsoft's biggest commercial cash cows. But at least one member of the Singularity family, Midori, may still end up a commercial project at some point.

The Midori effort is still forging ahead. Midori is currently considered a "technical strategy incubation" project.

An aside: There's a sense of frustration in a recent blog post by Midori team member Joe Duffy about the long and winding road from research to product.  From that post:

"A lot of research organizations value code and building real things, but still keep the group separate from the engineering groups. The building real things part is a step in the right direction, however the tragedy is that most of the time such research ends with a 'prototype'; at best, some number of months (or years!) later, the product team will have had a chance to incorporate those results....

And, man, how painful is it to realize that you could have delivered real customer value and became a true technological trendsetter, but instead sat on the ideas, in the worst case never delivering the idea beyond a paper, and in the best case delaying the delivery and thus giving your competitors an easy headstart and blueprint for cloning the idea."

Hmm. Has Midori -- word of which first leaked back in 2008 -- been hamstrung because of this dichotomy? Just wondering aloud....

Microsoft officials continue to decline to talk about Midori. But as a recent job posting for the team noted, work on the "C#-like language for writing asynchronous and parallel operating system components," now has a name: M#. There's already a language known as M# on the market; the Midori M# seems to be something separate and different.

Galen Hunt, the Principal Researcher in Microsoft Research's operating system group and another key figure in the creation and nurturing of Singularity, is still at Microsoft Research. (Hunt's bio describes him as the person who led the Singularity project; Larus's bio describes him as the one who started the Singularity project.) Hunt is instrumental in the Drawbridge library OS project at Microsoft. Hunt also led the "Experiment 19" project at Microsoft Research, which played a big role in replacing the Windows CE kernel with the Windows NT one (and thus influenced Windows Phone 8).

(Thanks to the many Twitter folk I follow who provided bits and pieces that helped with this post, including @headinthebox, @h0x0d, @JoseFajardo and @MaConfig.)