On 7/21/07, . <_(a)whats-your.name> wrote: Some email software does not put who posted every time, relax, it really isn't that big of a deal. realtime isnt a module, its a big kernel patch. my experience (not on ia32) I have been compiling kernels since the 2.4 series myself. All of the 2.6series, as the rt patch became availiable I compiled with it, and have had extremely good results. The only times any of the above happened was because of my own stupidity, I have never had it not compile. Any time it didn't boot was because of my own fault. Never had it freeze the boot sequence, and any freezes I experienced afterwards(extremely few) I have other suspects well in line before the patch. Seablade

On Sat, 2007-07-21 at 22:53 -0400, . wrote: it's word wrap doing the most damage there. no. "realtime" is a condition. RT patched refers to ingo molnars realtime preempt patches. ( http://people.redhat.com/mingo/realtime-preempt/ ) that people refer to all kinds of things as "realtime" just makes it harder to figure out what they are referring to without some specific context. confusingly, the realtime lsm *module* is basically a tool for assigning the correct permissions to your apps. (lsm stands for linux security module) it's also redundant in the face of rlimits which is another method of doing the same thing but one that is included in the mainstream kernel. more confusingly still, rlimits can be configured by different methods too. via a supporting PAM (/etc/security/limits.conf) or by set_rtlimits which is a command line tool with it's own configuration file. even more confusingly, you may not need to patch your kernel at all for several reasons. 1. over time elements of the rt patches have been merged into the mainline kernel. 2. some distros feature prebuilt low latency kernels 3. some distros are dedicated to audio and are built from the ground up with respect to low latency issues. i'd advise those not familiar with the details to avoid building a kernel unless they've tried what's available in their own distro first, and found it lacking. anyway check out these two links for some background: "Realtime Preemption" http://tapas.affenbande.org/wordpress/?page_id=6 "Linux Audio/MIDI Realtime System Setup" http://tapas.affenbande.org/wordpress/?page_id=73 i'm not sure if these pages are kept rigidly up to date but they are great for getting a handle on the issues. hats off to tapas (Florian Schmidt) i think they manage quite well over at 64studio. http://64studio.com/faq_developer indeed. well, perhaps you too should stick to blogs where overheated, unproductive and self indulgent diatribes are the norm. here, you risk being the exception. pete.

On Sun, 2007-07-22 at 19:53 +0200, Marc-Olivier Barre wrote: There is an increasingly interresting wiki being build over here: http://rt.wiki.kernel.org/index.php/Main_Page .. with all kinds of CONFIG information (including a pointer to https://wiki.ubuntu.com/RealTime ... :-) as well as RT programming examples. Also various RT-news, like this Java related one from IBM: WebSphere Real Time [...] * Response time measured in milliseconds * Unique Real Time Garbage Collection technology: Avoids unpredictable pauses to Java applications for garbage collection * Ahead-of-Time Compilation: Pre-compile code to achieve better performance than interpreted compilation and more predictable response times than Just-in-Time Compilation http://www.ibm.com/software/webservers/realtime/ So - as far as I can see - the Java language is now aboard as a first class RT citizen. Who would have thunk? -- mvh // Jens M andreasen

On 7/25/07, Lee Revell <rlrevell(a)joe-job.com> wrote: Yeah, and before building a realtime-safe GC, they'd better try build *at least* a safe GC... __________________ Marc-Olivier Barre.