This week Wind River (Nasdaq WIND) of Alameda CA announced that it had acquired patented technology from FSMLabs to endow embedded Linux with real-time responsiveness normally associated with an RTOS (real-time operating system).    What most industry watchers don't realize is that this move is just the latest in a very long history of cherry picking Free and Open Source Software (FLOSS) by the embedded market leader.

Starting about three years ago (v. ZDNet: Wind River, Red Hat team on embedded Linux), Wind River began a step-wise embrace of Linux and related open source software.  Up until that time, Wind River's public bluster attacking the open source OS belied the company's actual long-term involvement with Free and Open Source Software (FLOSS).  Starting in the early 1990s, Wind River began to deploy components from that other FLOSS OS, Berkeley UNIX (in particular TCP/IP networking components and tools) as well as pieces of GNU project code, in particular the GNU C compiler (gcc) and various utilities (binutils).

In 2001, Wind River acquired BSDi, whose assets included the BSD/OS version of Berkeley UNIX and also significant responsibility for maintenance and distribution of FreeBSD.   That acquisition did not bode well for Wind nor for BSD/OS, with Wind River discontinuing BSD/OS and depriving FreeBSD of even the minimal resources needed to maintain it.

Since its more recent "return" to Open Source, Wind River has made substantial investments in key FLOSS projects, most notably in Eclipse , patches to the Linux kernel, and in its own Open Source version of TIPC .

By acquiring the real-time Linux (RTLinux) assets from FSMLabs, Wind River resolved a key gap in its embedded and real-time product lines: previously, Wind's real-time offering (VxWorks) was not Open Source and its Linux products had no options for real-time responsiveness. Earlier, the company was loathe to cross that gap for fear of cannibalization of VxWorks sales.  The acquisition of RTLinux is in effect an admission of the fact that customers were preselecting for Linux (in spite of its performance limitations) and were not amenable to a proprietary/closed real-time alternative once they had opted for a FLOSS platform.

RTLinux technology in theory applies to a range of application areas - telecommunications infrastructure, networking, control, instrumentation, multimedia, etc. but it appears that Wind River is most eager to see it applied to the nascent position of Linux as a mobile phone OS.  This is a speculative path for Wind, who has had little or no visibility in this dynamic space, but could give the Alameda-based supplier a leg up by enabling design wins in so-called "single core" designs.

Today, Linux (and WindowsMobile and SymbianOS) run on general purpose application processors (APs) while time-critical baseband and CODEC operations are handled by dedicated DSPs and/or encapsulated wireless (GPRS or CDMA) modems, termed baseband processors (BPs).  By deploying Linux on the AP and letting RTLinux-hosted baseband code run on that same single CPU (without need for a separate BP), Wind River could extend the reach of the Open Source OS (and of its offerings) from the relatively rarefied domain of smart phones down to higher volume, lower-cost single-core feature phones and even low-end devices that together account for over 90% of the global handset market.