"Mario." <[EMAIL PROTECTED]> said: > On Sun, Oct 26, 2008 at 11:40 PM, Chris Morley > <[EMAIL PROTECTED]> wrote: > > > > I have been reading a little about RTAI and Xenomai. > > I wonder Why EMC choose RTAI, Other then at the time Xenomai may not have been available (around 2005?). > > Xenomai is only, very, very recent 'invention'
I am sure that others can give you a much better historical review of EMC, but here is some personal remembrances to put things in perspective. I first remember seeing EMC circa. 1998 which was developed by the National Institute of Standards and Technology (NIST). I do not remember what OS it used at that time, but it may have been the first version of RT-Linux. Poking around a bit I found that Victor Yodaiken published some of his first work on RT-Linux in the February 1997 issue of Linux Journal titled "Introducing Real-Time Linux". RT-Linux has, what some would consider, a controversial history. It was basically a patch to the linux kernel. Soon after Yodaiken got it all working he applied for, and was granted, a patent for the hard real-time mechanism he added to the kernel (see http://yro.slashdot.org/article.pl?sid=00/02/10/1048214&mode=thread for details). That had a good bit of the linux community up in arms... Since there there has been release both a freely distributed version and a commercial professional version of their RT-Linux operating system. So, the RTAI group which came up with what they hoped was a new mechanism which solved the RT issues without violating the patent. I have not kept up with the arguments on both sides, but as of 2004, when I last did any serious work on porting EMC (versions 1 and 2) to the latest and greatest RT-Linux Professional and Free version. At that time Yodaiken claimed that RTAI violated his patent, and the RTAI folks claimed they did not. At the time, adoption of either RT-Linux or RTAI for hard real-time systems was hampered by fear of exorbitant royalties on the on hand, and loosing development investment if RTAI lost a legal decision on patent violation. I do not know if the RTAI/RT-Linux patent question was ever resolved or if there is still all the finger wagging going on. I am sure that this post is going to raise a couple of people blood preasure a point or two and might even start a little mud-slinging (which is not my intent at making this post). Actually, it was that king of nasty infighting that completely turned me off to public involvement in EMC/HAL developers, and I have remained a lurker ever since. Now Monta Vista came onto the scene with their preemptive linux kernel. Part of that work was done with the consultation of patent lawyers which made sure that the approach was done in a clean-room style, and that it would not violate the patent. The problem with the preemptive kernel (or at least last I checked) is that it is soft-realtime and cannot guarantee hard determinism. In general it does a GREAT job, but there is a question if GREAT is good enough. Actually, with the resources Monta Vista, et al., put into the preemptive kernel I am personally hoping that they get hard-realtime worked out. Along those lines, Xenomai's FAQ implies that PREEMPTIVE_RT support is one of it's goals. Actually, I have remained interested in Xenomai's progress -- in particular to their efforts to port and maintain RT capabilities on embedded platforms. I am sure I have forgotten things, and have a detail or two wrong, but I hope this helps put things into a little better perspective. Best regards, EBo -- ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
