Subject:
|
RE: The pain of communicating with the RCX
|
Newsgroups:
|
lugnet.robotics.rcx.pbforth
|
Date:
|
Thu, 9 Mar 2000 18:37:44 GMT
|
Viewed:
|
1570 times
|
| |
|
|
> The power of Forth lies in the ease with which a system can be built and
> and tested both incrementally and interactively. Unfortunately I'm experiencing
> severe difficulties estasblishing a reliable communication channel between my
> PC (Linux) and the RCX (RIS 1.5). I'm hoping you people can help me (and
> others) to annihilate those difficulties, so forth can be used the way it is
> intended to, and not in a compile-with-as and download manner.
Good. I'm glad your on-side so far :-)
> 1) How do you ship forth code as text to the RCX? Downloading an srec file
> is no problem, but when I try to do an ascii download, I always get some
> communication failure somewhere. I tried increasing line and character
> delays up to 200ms and 80ms resp., but although this made the problem less
> frequent it did not cure it. When working interactively comm failures occur
> too, but retyping a single line is not as annoying as trying to reload an
> entire file. Deminishing file size helps a bit, but downloading a single
> tiny file for every word is a pain.
You are right. I have found that getting a reliable connection is very
dependent on ambient light, the voltage of the RCX batteries, and the tower itself.
Try to use the tower in NEAR mode so that it does not swamp the RCX receiver.
I'm trying to get my Tcl GUI back in order so that we can all have a common
base to work with. There are so many comm programs out there that it
is difficult to get them all working.
> 2) How can I upload an augmented forth system from the RCX to my PC? I think
> rx (=Receive X-Modem; from the lrzsz package) gets confused by the echo of
> the IR tower: It reports an `025 received error'. This is a NAK in octal.
> Presumably the very NAK rx has sent to initiate the transfer. I found
> an old program called xcomm, which fares a little better. It downloads the
> entire file, but reports a timeout in the end. When I try
>
> : z S" Hello" ;
> z X-SEND
>
> and start the transfer by typing ^U, the header and string is transferred,
> then a little pause occurs, and the checksum is sent. I don't see any 0s to
> fill the block to 128 bytes. Doing "0 emit" doesn't give me a 0 either. This
> explains the timeout error, but not why a 0 isn't transmitted. Does the IR
> tower eat them?
This I'm not too sure about. I'll have to look at what the source code
does. It is getting more difficult to find an XMODEM program for all
three platforms, which is another reason for getting the Tcl script back up.
> I'm suspecting the IR tower of subversive activities (see how guiltily it
> blinks with its single green eye!), even more so because information gathered
> on the web (in particular Kekoa Proudfoot's page) seems to indicate that it
> was designed to handle only a very restrictive query-answer protocol. This
> would mean every (few) character(s) should be wrapped in an IR-tower packet
> which is subsequently transmitted. This requires of course special software
> on both the PC and the forth side, so the simplicity of using just any
> terminal emulator is gone:(
Yes and no. The first time you hit CR on your terminal, you get garbage
back from the tower. If the comm program is waiting too long before
retrying, it will always get junk back as the first few chars.
> Has anybody more information on this, or maybe even (the beginning of) a
> solution?
Not yet, but so far developing under Windows has not been too bad.....
Let me work on the TclTk script and keep everyone posted here.
Cheers,
Ralph Hempel - P.Eng
--------------------------------------------------------
Check out pbFORTH for LEGO Mindstorms at:
<http://www.hempeldesigngroup.com/lego/pbFORTH>
--------------------------------------------------------
Reply to: rhempel at bmts dot com
--------------------------------------------------------
|
|
Message has 1 Reply:
 | | Re: The pain of communicating with the RCX
|
| (...) Does that count as `tempted by the dark side' or `seen the light'? :) (...) itself. (...) I tried near and far mode, several distances, putting the tower and the RCX under a dark cloth... Everything I could think of except putting 'em where (...) (25 years ago, 9-Mar-00, to lugnet.robotics.rcx.pbforth)
|
Message is in Reply To:
| | The pain of communicating with the RCX
|
| The power of Forth lies in the ease with which a system can be built and and tested both incrementally and interactively. Unfortunately I'm experiencing severe difficulties estasblishing a reliable communication channel between my PC (Linux) and the (...) (25 years ago, 9-Mar-00, to lugnet.robotics.rcx.pbforth)
|
9 Messages in This Thread:
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
