jrate-devel Mailing List for jRate

Hi Pablo,

On Sat, Apr 29, 2006 at 01:47:20PM -0400, Pablo Daniel Martinez wrote:
> Any body there ?
> I'm interested in the project, but it seems there's no too much activity.
> Is the project dead ?

No, the project is not dead, though I've had little time to work on it
since last fall.  I've been using jRate quite a bit, but not working
on the internals.

At present, I'm not prepared to release the current CVS as an official
release, though it does contain a number of enhancements since the
last beta.  Check out the ChangeLog and NEWS files in the CVS repo for
details.

Are you interested in using or contributing to this project?

Morgan
-- 
Morgan Deters
md...@mo...

Any body there ?
I'm interested in the project, but it seems there's no too much activity.
Is the project dead ?

Best regards,
                       pablo

On Mon, Apr 03, 2006 at 02:06:50PM +0000, Alex Craddock wrote:
> Hello,
> 
> I was just wondering what OS's people have been able to compile jRate 
> on, as i need to try and get it to work on my laptop and so far the few 
> i have tried (gentoo linux and ubuntu linux) have not been succesful in 
> installing

Hm, if it's building (that is, "make" works), but not installing
("make install"), is it an issue with the demos/ directory?  Some
releases were partly broken (I think that may include the current
stable), and it would sometimes bomb out.

If something in demos/ fails, then you're probably okay; jRate is
installed properly before descending into demos/... it's just a
makefile issue in that directory.

Morgan
-- 
Morgan Deters
md...@mo...

On Apr 3, 2006, at 7:06 AM, Alex Craddock wrote:

> I was just wondering what OS's people have been able to compile  
> jRate on, as i need to try and get it to work on my laptop and so  
> far the few i have tried (gentoo linux and ubuntu linux) have not  
> been succesful in installing

Seems to work fine on Fedora Core, at least.

Trevor

Hello,

I was just wondering what OS's people have been able to compile jRate 
on, as i need to try and get it to work on my laptop and so far the few 
i have tried (gentoo linux and ubuntu linux) have not been succesful in 
installing

Thanks

Alex.

Trevor,

On Sun, Apr 02, 2006 at 12:33:50AM -0800, Trevor Harmon wrote:
> Interesting. How do you define "amazing" with regard to timing  
> characteristics?

I didn't put in numbers because I feared that my memory might fail me
if I tried.  :-)

In particular I recall very tight wake up after nanosleep().  The
spread is perhaps a few tens of usec (even in heavily-loaded
environments).  I think I still have some test kernels around, if I
get a chance this week I'll play with them and post some more
observations.

I should mention that jRate can, I'm sure, get a lot more out of this
kernel than it presently is.  In particular, jRate's periodic thread
release mechanism needs a rewrite to reduce overhead and increase
predictability.  (It will also then be more appropriate for use with
RTAI/fusion.)

> Also, am I correct in my understanding that a vanilla kernel, patched  
> with realtime-preempt, and RTNPTL running on top of it, is the best  
> possible environment for jRate at the moment? (Not counting pricey  
> commercial Linux kernels here)

Probably.  RTAI/fusion might also show good performance now (in
"compatibility" mode), and, like I mentioned, full hard-rt mode
RTAI/fusion support is on the horizon too.

Morgan
-- 
Morgan Deters
md...@mo...

On Mar 31, 2006, at 12:50 PM, Morgan Deters wrote:

> On Wed, Mar 22, 2006 at 01:19:04PM -0800, Trevor Harmon wrote:
>> This is a user question, not a devel question, but since there isn't
>> a user mailing list for jRate, I'm posting to jRate-devel.
>
> There is, actually, though it's not hosted through sourceforge and
> isn't very active.  Subscription info at
> http://jrate.sf.net/MailingLists.php

Oh, I see. I missed it because I was only looking at the Mail section  
of the project page. You might consider moving the other lists to  
SourceForge to avoid confusion. (Or move the devel list to wustl.edu.  
Doesn't matter to me as long as they're all in the same place.)

> Minimally, to support priority inheritance, you need RTNPTL.  As an
> additional benefit, RTNPTL is careful in other places to ensure that
> high-priority threads always run in preference to low-priority ones.
> There are some notes on setting up jRate with RTNPTL in the INSTALL
> file in the distribution.  It requires you to build a patched kernel
> and patched glibc, then point jRate to it when you ./configure.
>
> Also, there are several patches to the Linux kernel to make it
> maximally preemptible and low-latency.  In particular I suggest
> looking at Ingo Molnar's realtime-preempt patch.  I've played with
> this thing in the lab and verified some really amazing timing
> characteristics.  Never played with it on embedded though.
> http://people.redhat.com/~mingo/realtime-preempt/

Interesting. How do you define "amazing" with regard to timing  
characteristics?

Also, am I correct in my understanding that a vanilla kernel, patched  
with realtime-preempt, and RTNPTL running on top of it, is the best  
possible environment for jRate at the moment? (Not counting pricey  
commercial Linux kernels here)

Trevor

On Apr 1, 2006, at 5:07 AM, Luigi Talamona wrote:

> Dear JRate developers,
> my name is Luigi Talamona. I live in Italy.
> My question is: I need to generate digital signals on RS-232 port, to
> drive electronic custom devices (I'll want to project a software
> test-bench for them). Time gap between signals is near 1  
> microsecond or
> less (1 microsecond tipical). Is it possible to obtain these features
> using Jrate on a Linux x86 based host ? If not, do you have any other
> idea about my problem ?

RTSJ is probably overkill for that, and more importantly, I don't  
think it is designed to handle sub-microsecond accuracy.

I think you'd be happier using a microcontroller for your design. The  
MAX232, for example.

Trevor

Dear JRate developers,
my name is Luigi Talamona. I live in Italy. 
My question is: I need to generate digital signals on RS-232 port, to
drive electronic custom devices (I'll want to project a software
test-bench for them). Time gap between signals is near 1 microsecond or
less (1 microsecond tipical). Is it possible to obtain these features
using Jrate on a Linux x86 based host ? If not, do you have any other
idea about my problem ? 

Apologize for my bad english...

Thanks in advance

Luigi T.