From: ja...@cco.caltech.edu (John Lindal)
Subject: Programming the new RCX (mindstorm)
Date: 1998/06/12
Message-ID: <6lsdkp$861@gap.cco.caltech.edu>#1/1
X-Deja-AN: 362170726
Organization: California Institute of Technology, Pasadena
Newsgroups: rec.toys.lego


Have any of the hardware gurus thought about programming LEGO's new
RCX directly?  If one could figure out how one talks to the computer side
of the IR link, one could write a C library and bypass LEGO's programming
language...

John Lindal

From: Russell Nelson <nel...@crynwr.com>
Subject: Re: Programming the new RCX (mindstorm)
Date: 1998/06/13
Message-ID: <m2lnr1v38d.fsf@desk.crynwr.com>#1/1
X-Deja-AN: 362299978
References: <6lsdkp$861@gap.cco.caltech.edu>
Organization: Crynwr Software
Newsgroups: rec.toys.lego


ja...@cco.caltech.edu (John Lindal) writes:

> Have any of the hardware gurus thought about programming LEGO's new
> RCX directly?  If one could figure out how one talks to the computer side
> of the IR link, one could write a C library and bypass LEGO's programming
> language...

Rumor has it that they're going to tell us how to do that.  Hehe.
Just *imagine* the things we'll be able to do if we don't have to
waste time reverse-engineering it.

-- 
-russ <nel...@crynwr.com>  http://web.crynwr.com/~nelson
Crynwr supports Open Source(tm) Software| PGPok |   Freedom is the primary
521 Pleasant Valley Rd. | +1 315 268 1925 voice |   cause of Peace, Love,
Potsdam, NY 13676-3213  | +1 315 268 9201 FAX   |   Truth and Justice.

From: lan...@iei.net
Subject: Re: Programming the new RCX (mindstorm)
Date: 1998/06/13
Message-ID: <6lv3he$la0$1@nnrp1.dejanews.com>#1/1
X-Deja-AN: 362403459
References: <6lsdkp$861@gap.cco.caltech.edu>
X-Http-User-Agent: Mozilla/4.0 (compatible; MSIE 4.01; Windows NT)
Organization: Deja News - The Leader in Internet Discussion
X-Article-Creation-Date: Sat Jun 13 23:55:26 1998 GMT
Newsgroups: rec.toys.lego


As soon as I get one, I'm going to examine the serial interface.  If it's not
too difficult, I'll cook something up.  I know there was some other kind of
lego device, a dacta thing I think, that had to be sent a secret "sentence"
before it could be programmed.  If TLG is smart, they'll be nice about people
reverse engineering their stuff.  That have to know it's going to happen.
Maybe they'll publish the interface!

brian


In article <6lsdkp$8...@gap.cco.caltech.edu>,
  ja...@cco.caltech.edu (John Lindal) wrote:
>
> Have any of the hardware gurus thought about programming LEGO's new
> RCX directly?  If one could figure out how one talks to the computer side
> of the IR link, one could write a C library and bypass LEGO's programming
> language...
>
> John Lindal
>
>


-----== Posted via Deja News, The Leader in Internet Discussion ==-----
http://www.dejanews.com/   Now offering spam-free web-based newsreading

From: Jeff Elliott <jef...@telepres.com>
Subject: Re: Programming the new RCX (mindstorm)
Date: 1998/06/18
Message-ID: <35889F7C.40A9@telepres.com>#1/1
X-Deja-AN: 364402102
Content-Transfer-Encoding: 7bit
References: <6lsdkp$861@gap.cco.caltech.edu> <6lv3he$la0$1@nnrp1.dejanews.com>
NNTP-Posting-Time: 20 Jun 1998 04:38:20 GMT
Mime-Version: 1.0
Reply-To: jef...@telepres.com
To: lan...@iei.net
Content-Type: text/plain; charset=us-ascii
Organization: Interlog Internet Services
Newsgroups: rec.toys.lego


lan...@iei.net wrote:
> 
> As soon as I get one, I'm going to examine the serial interface.  If it's not
> too difficult, I'll cook something up.  I know there was some other kind of
> lego device, a dacta thing I think, that had to be sent a secret "sentence"
> before it could be programmed.  If TLG is smart, they'll be nice about people
> reverse engineering their stuff.  That have to know it's going to happen.
> Maybe they'll publish the interface!
> 
> brian
> 

I've heard that the sentence  for the old Dacta controller was something
like

"Do you byte when I knock?" (From computer)
"Just a bit off the block." (From controller)

And according to what I've read, Lego puts pretty dire threats about
reverse
engineering into their instructions, etc.  Don't know if they'd bother
going
after Lego users, but they'd definitely like to prevent their
competitors
from doing it.

Jeff

From: Niels Kistrup <NielsK...@compuserve.com>
Subject: Re: Programming the new RCX (mindstorm)
Date: 1998/06/14
Message-ID: <3583EA3D.4B39A9D4@compuserve.com>#1/1
X-Deja-AN: 362551618
Content-Transfer-Encoding: 7bit
References: <6lsdkp$861@gap.cco.caltech.edu>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Newsgroups: rec.toys.lego


John,

Why stop at 'C'; go for a real language like C++ or Delphi.  OOL seem to be a
natural for robotics.  Elements such as sensors and motors easily map to
objects.
If I knew the format for the resulting object code, I'd being working on
writing a very simple C++ compiler (no multiple inheritance, register
variables, strings, namespaces, templates, etc., etc.) this weekend.

Niels

John Lindal wrote:

> Have any of the hardware gurus thought about programming LEGO's new
> RCX directly?  If one could figure out how one talks to the computer side
> of the IR link, one could write a C library and bypass LEGO's programming
> language...
>
> John Lindal

From: ja...@cco.caltech.edu (John Lindal)
Subject: Re: Programming the new RCX (mindstorm)
Date: 1998/06/17
Message-ID: <6m71vo$sco@gap.cco.caltech.edu>#1/1
X-Deja-AN: 363359396
References: <6lsdkp$861@gap.cco.caltech.edu> <3583EA3D.4B39A9D4@compuserve.com>
Organization: California Institute of Technology, Pasadena
Newsgroups: rec.toys.lego


Niels Kistrup <NielsK...@compuserve.com> writes:

>Why stop at 'C'; go for a real language like C++ or Delphi.  OOL seem to be a
>natural for robotics.  Elements such as sensors and motors easily map to
>objects.

Personally, I use C++, but I said C because I figured there was a bigger
following for that.  Once the specs are published, I would expect C, C++,
BASIC, etc, etc libraries.

John

From: lan...@iei.net
Subject: Re: Programming the new RCX (mindstorm)
Date: 1998/06/19
Message-ID: <6mcj9p$flr$1@nnrp1.dejanews.com>#1/1
X-Deja-AN: 364073697
References: <6lsdkp$861@gap.cco.caltech.edu> 
<3583EA3D.4B39A9D4@compuserve.com> <6m71vo$sco@gap.cco.caltech.edu> 
<35871C66.BB4665C8@compuserve.com>
X-Http-User-Agent: Mozilla/4.0 (compatible; MSIE 4.01; Windows NT)
Organization: Deja News - The Leader in Internet Discussion
X-Article-Creation-Date: Fri Jun 19 02:44:09 1998 GMT
Newsgroups: rec.toys.lego


You could make the argument that my toilet is object oriented.	Hey, it has a
seat position property and a flush method.  I'll let the object oriented
biased posts go as long as you promise not to tell us how woderful Java is.

brian

P.S.  after 10 years of procedural programming, Microsoft Foundation Classes
is the first and only practial application of C++ and object oriented
technology I've seen.  If the only programming interface to the rcx is a c++
object, I'll have to vomit.

In article <35871C66...@compuserve.com>,
  Niels Kistrup <NielsK...@compuserve.com> wrote:
>
> John,
>
> Unfortunately I would have to agree that 'C' is a lower common denominator.
> Still, for sensors and motors, an OO approach is natural.
>
> Niels
>
> John Lindal wrote:
>
> > Niels Kistrup <NielsK...@compuserve.com> writes:
> >
> > >Why stop at 'C'; go for a real language like C++ or Delphi.  OOL seem to
be a
> > >natural for robotics.  Elements such as sensors and motors easily map to
> > >objects.
> >
> > Personally, I use C++, but I said C because I figured there was a bigger
> > following for that.  Once the specs are published, I would expect C, C++,
> > BASIC, etc, etc libraries.
> >
> > John
>
>


-----== Posted via Deja News, The Leader in Internet Discussion ==-----
http://www.dejanews.com/   Now offering spam-free web-based newsreading

From: sillywiz@excession.demon.co._delete_this_.uk (Keith M. Lucas)
Subject: Re: Programming the new RCX (mindstorm)
Date: 1998/06/19
Message-ID: <EutCBK.MML@excession.demon.co._delete_this_.uk>#1/1
X-Deja-AN: 364651465
References: <6lsdkp$861@gap.cco.caltech.edu> 
<6m71vo$sco@gap.cco.caltech.edu> <35871C66.BB4665C8@compuserve.com> 
<6mcj9p$flr$1@nnrp1.dejanews.com>
X-Complaints-To: ab...@demon.net
X-Mail2News-Path: news.demon.net!post-12.mail.demon.net!
post.mail.demon.net![194.222.65.210]!excession.demon.co.uk
X-Trace: mail2news.demon.co.uk 898410658 mail2news:
8173 mail2news mail2news.demon.co.uk
Organization: Excession
Newsgroups: rec.toys.lego


In article <6mcj9p$flr$1...@nnrp1.dejanews.com>,  <lan...@iei.net> wrote:

>P.S.  after 10 years of procedural programming, Microsoft Foundation Classes
>is the first and only practial application of C++ and object oriented
>technology I've seen.  If the only programming interface to the rcx is a c++
>object, I'll have to vomit.

Oh dear. After several years of OO programming and more than a decade
in various procedural languages, MFC is not the first and only
practical application of C++ I've come across, but it's the first
that's made me wake up screaming at night. (Most of the others have
just depressed me.) Unfortunately, it's lugging around a lot of crap.

Firstly, the first versions of the compiler didn't support run-time
type info as native and didn't do exception handling properly. Then
there are the message maps, when any conscious C++ programmer would
use virtual functions, only VC++ 1.0 wasn't fast enough, and now
that's locked in as well. Add to that that it's tied to an absolutely
not at all OO back end, and all that luvverly Hungarian (which closely
resembles vomiting on a keyboard) and it's just pants.

Most C++ frameworks are in this state, bending over backwards for
compatibility with things and as a result don't have the unbearable
lightness of being proper C++.

I use Think C Library at work (we're locked into it by historical
decision. Seeing a trend here ?) and the whole time you're jumping
through hoops, eg: change a window name, remember to ask the menu
system to change what it thinks the window, simply because the whole
point of OO has been totally, totally missed.

-------------------------------------------------------------------------------
sillywizATexcessionDOTdemonDOTcoDOTuk"It's not a personality..it's a bulldozer"
--------------------- The Official Irony Generator for the 1998 World Cup -----

From: mat...@disco.tln.net (Matthew Miller)
Subject: Re: Programming the new RCX (mindstorm)
Date: 1998/06/20
Message-ID: <slrn6ommjf.e9k.mattdm@jadzia.bu.edu>#1/1
X-Deja-AN: 364422745
References: <6lsdkp$861@gap.cco.caltech.edu> 
<6lv3he$la0$1@nnrp1.dejanews.com> <35889F7C.40A9@telepres.com>
Reply-To: mat...@disco.tln.net
Organization: Boston University
Newsgroups: rec.toys.lego


Jeff Elliott <jef...@telepres.com> wrote:
>And according to what I've read, Lego puts pretty dire threats about
>reverse engineering into their instructions, etc.  Don't know if they'd
>bother

Man. Guess I won't be reading the instructions. :)

-- 
Matthew Miller                      --->                  mat...@mattdm.org
Quotes 'R' Us                       --->             http://quotes-r-us.org/

From: Dave Baum <dbaum@no.spam>
Subject: Re: Programming the new RCX (mindstorm)
Date: 1998/06/20
Message-ID: <358BFED1.FF6F5B35@no.spam>#1/1
X-Deja-AN: 364760184
Content-Transfer-Encoding: 7bit
References: <6lsdkp$861@gap.cco.caltech.edu> 
<6m71vo$sco@gap.cco.caltech.edu> <35871C66.BB4665C8@compuserve.com> 
<6mcj9p$flr$1@nnrp1.dejanews.com> 
<EutCBK.MML@excession.demon.co._delete_this_.uk>
Content-Type: text/plain; charset=us-ascii
Organization: EnterAct L.L.C. Turbo-Elite News Server
Mime-Version: 1.0
Newsgroups: rec.toys.lego


Keith M. Lucas wrote:
> 
> <snip> 
> 
> Most C++ frameworks are in this state, bending over backwards for
> compatibility with things and as a result don't have the unbearable
> lightness of being proper C++.
> 
> I use Think C Library at work (we're locked into it by historical
> decision. Seeing a trend here ?) and the whole time you're jumping
> through hoops, eg: change a window name, remember to ask the menu
> system to change what it thinks the window, simply because the whole
> point of OO has been totally, totally missed.
> 

I have to agree completely on the fact that many GUI frameworks are hindered
by being built on top of GUI primitives that were not object oriented.

The "Next" was a rather interesting computer - it was truly object oriented
from the ground up.  Objective C was the implementation language of choice,
and the entire GUI (NextStep) was built very cleanly around this.  Programming
under NextStep was IMHO a far better OO experience than any of today's C/C++
frameworks (MFC, OWL, Think C Library, Power Plant, etc).  Too bad Next didn't win.

Dave

-- 
-----------------------------------
Reply to dbaum at enteract dot com
-----------------------------------