View Single Post
  #12   Spotlight this post!  
Unread 29-04-2008, 10:48
dcbrown dcbrown is offline
Registered User
AKA: Bud
no team
Team Role: Mentor
 
Join Date: Jan 2005
Rookie Year: 2005
Location: Hollis,NH
Posts: 236
dcbrown has much to be proud ofdcbrown has much to be proud ofdcbrown has much to be proud ofdcbrown has much to be proud ofdcbrown has much to be proud ofdcbrown has much to be proud ofdcbrown has much to be proud ofdcbrown has much to be proud ofdcbrown has much to be proud ofdcbrown has much to be proud of
Re: NEW 2009 Control System Released

Comparision of ports of VxWorks on PPC and RTLinux has shown key elements in terms of real time response favors VxWorks by 30-50%. RTLinux isn't a slouch, but does fall short in real time response in comparison.

This article comes to mind, but I remember reading a couple others:
Performance Analysis of VxWorks and RTLinux. In general, even if you are not writing a bunch of ISRs for devices, the nature of data flow programming depends upon the deterministic response of handling hard clock interrupts in order to time slice all your tasks.

Another issue is one of support. RTLinux would end up needing a grass roots support effort on this platform along with a port/creation of its own "WPILIB" equivalent. But the real heartache would come with the FPGA and having to integrate that into RTLinux.

Just my opinion - it would be an interesting academic exercise, but you'd be fighting a huge uphill battle to get it adopted by more than a handful of teams because it doesn't provide significant differentiators in terms of functionality while at the same time having potential detrimental performance issues by comparison.
Reply With Quote