|
[DPRG] Processing unit for mobile robot
Subject: [DPRG] Processing unit for mobile robot
From: ed at okerson.com
ed at okerson.com
Date: Thu Apr 24 08:58:55 CDT 2008
>> http://rt.wiki.kernel.org/index.php/Main_Page
>
> Thanks for the link, I did some reading/scanning. It is
> difficult to get far thought. I don't know if you can appreciate
> it or not, but to an outsider, you can't read a single page
> without running into a dozen acronyms or terse references that
> can't be understood.
Did you start with the FAQ as I recommended? I don't see anything there
that is "terse" or difficult to understand, unless you just don't want to
understand.
>> 10-30us scheduler and interrupt latencies;
>
> So my interpretation is that a PC is just now approaching
> latencies of a 2MHz 68HC11, which had a longest latency of a
> divide instruction of 41 cycles or 21uS plus a call time of
> something like... Was it 11 cycles?
>
> Or am I not understanding what this means on a PC? Now
> admittedly, the HC11 couldn't do as much once it got to a
> routine. 1000x Clearly. But I'm thinking this is a measure of
> how long a response to a real time request runs?
>
>> After all, humans are inherently non-realtime, but we do
>> much more than just move around.
>
> I don't know if I agree. Humans have latencies in sub-second
> range, and we largely consider that normal response times. We
> consider quadraplegics, unable to feed themselves or move more
> than their eyes as severly disabled. While everyone is amazed at
> Stephen Hawkings mind, his situation is not to be envied. So I
> think what we do is still motion related.
Yes, but how many times have you seen someone you have not seen in years,
and it takes you two days to remember their name? Or someone tells a joke
and everyone else laughs, but you don't "get it" for a few hours, then you
burst out laughing when your brain finally parses it?
Given a choice, would you rather be in Stephen Hawkings condition, or that
of a lobotomized person? I never said we didn't need motion, only that
there is more to it than just motion.
>> What I see as so desireable is to not have to create
>> everything from scratch myself. By using an Operating
>> System, many people can work to the same API and build parts
>> of the system. But by using an open source operating system,
>> you still get all the source code so you can tweak it if you
>> so desire.
>
> Okay, I hear that you want to make use of others code. I know
> there are some vision programs. I know you can make use of
> network codes. Believe me I'm not trying to be facetious, but,
> does useful code for robots exist? Is there OpenPID? OpenNav
> OpenSteering?
GPS handler
http://gpsd.berlios.de/
Steering/path planning software:
http://opensteer.sourceforge.net/
Robot Navigation
http://clunixchit.blogspot.com/2006/04/carmen-carnegie-mellon-robot.html
Realtime and Kinematics:
http://www.orocos.org/
Just to name a few. Need more? Google is your friend.
> Maybe there are. I tried Open GPS and got several hits. Not so
> sure the others worked. .PID and .NAV are apparently file names,
> and there are lots of articles on how to "open" those files.
A pid is a Process ID in most operating systems, so searching for that
will probably not turn up much usefull info without more qualifying terms.
> Maybe I should look at LEAF some more. I have largely rejected
> the Open software movement for reasons I won't dive into right
> here.
>
>> I choose Linux because of consistency. I can use the same
>> tools regardless of what platform I am developing for, and my
>> code compiles and runs on several platforms with consistent
>> interfaces and API's. And this is all possible without
>> spending large sums on development tools.
>
> Okay.
>
> The same is true for me and Forth, except I can work on even
> smaller platforms. So I can see that side of the argument.
Since the predominant language of Linux is C, then a lot of my code will
run on things much smaller as well. But there is a lot more open source
stuff available in C than there is in Forth. Granted that all open source
software is not perfect, but neither is most of the stuff you pay for.
There have been many occasions that I have evaluated several open source
packages for a particular use, and ended up writing my own from scratch
anyway, but that is the exception not the rule. Most of the time I find
open source apps/libraries that do exactly what I need, or come very
close.
Ed
More information about the DPRG mailing list
|
