Subject:
|
Re: Bug in RCX 2.0 beta (Was: RIS 2.0 software now in beta)
|
Newsgroups:
|
lugnet.robotics.rcx, lugnet.robotics
|
Date:
|
Wed, 21 Mar 2001 14:29:56 GMT
|
Viewed:
|
2043 times
|
| |
|
|
In lugnet.robotics.rcx, Sebastian Wangnick writes:
> Dear fellow Lego enthusiasts,
>
> so far I did not receive any feedback to my problem report, neither from the
> newsgroup, nor from Lego technical support. Could somebody please try to
> confirm that I am not completely stupid in my analysis!
>
> Regards,
> Sebastian
Hi Sebastian
I think you're right in your analysis and TLG may silently have fixed the
problem, because if you look at the Vision Command installation, you'll find
that some of the SDK DLLs are newer in that release. For fun, I copied the
DLLs across to the SDK and the problem seemed to go away. Hope this helps.
-- Michael
> --
> Familie Wangick <familie.wangnick@t-online.de> schrieb in im Newsbeitrag:
> G9opDv.B47@lugnet.com...
> > Dear all,
> >
> > I have managed to pin down the problem a bit further. Please look at the
> > following program:
> >
> > program foo {
> > const c10 = 10
> > main { forever {} }
> > task test {
> > local v2f
> > local v2e
> > local v2d
> > v2f = 1
> > v2e = 2
> > v2d = v2e * c10
> > }
> > }
> >
> > If you download this script to the brick, and the upload it again, you • will
> > receive the following:
> >
> > task,0,2,0,0x27,0x81
>
> task,1,25,0,0x14,0x2f,0x02,0x01,0x00,0x14,0x2e,0x02,0x02,0x00,0x14,0x2f,0x00
> > ,0x2e,0x00,0x54,0x2f,0x02,0x0a,0x00,0x14,0x2d,0x00,0x2f,0x00
> >
> > Task 1 (the task test) translates as following into LASM:
> >
> > task
> > 1
> > 25
> > 0
> > 0x14 setv 0x2f const 0x0001
> > 0x2f
> > 0x02
> > 0x01
> > 0x00
> > 0x14 setv 0x2e const 0x0002
> > 0x2e
> > 0x02
> > 0x02
> > 0x00
> > 0x14 setv 0x2f var 0x002e
> > 0x2f
> > 0x00
> > 0x2e
> > 0x00
> > 0x54 mulv 0x2f const 0x000a
> > 0x2f
> > 0x02
> > 0x0a
> > 0x00
> > 0x14 setv 0x2d var 0x002f
> > 0x2d
> > 0x00
> > 0x2f
> > 0x00
> >
> > As you can see the compiler translates the general multiplication v2d = • v2e
> > * c10 into a setv tmp v2e, mulv tmp c10, setv v2d tmp sequence. When
> > allocating the temporary local variable if fails to respect the • user-defined
> > local variables, though -- it should have used the next free local • variable
> > 0x2c for this operation!
> >
> > Hope this help fixing the problem.
> >
> > Regards,
> > Sebastian
> > --
> > Dipl.-Inform. Sebastian Wangnick
> >
> > Familie Wangick <familie.wangnick@t-online.de> schrieb in im Newsbeitrag:
> > G9oKJq.1y2@lugnet.com...
> > > Dear all,
> > >
> > > I'm experiencing a strange phenomenon in my MindScript program which I
> > > believe to be a bug, where within a task the assignment of a • multiplication
> > > of a local variable (containing value 0) and a constant (set to 10) to a
> > > second local variable would set a third local variable (containing 1) to
> > > zero! If rewritten as a simple assignment and then a multiplying • assignment
> > > the phenomenon disappears.
> > >
> > > Please find attached the script exemplifying the issue. It expects a • light
> > > sensor on 1, a motor turning the light sensor on B and a touch sensor on • 2
> > > being triggered when the motor turned the light sensor to 180° • resp. -180°.
> > >
> > > As a newbie to LEGO robotics (which both my son and me find fantastic) I
> > > don't exclude that I overlooked some finesse of the RCX environment • though
> > > ...
> >
> >
> >
|
|
Message is in Reply To:
8 Messages in This Thread:
  
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
