Re: [Amforth] rshift not zero padded

[tur b. <mai...@gm...>]

If you are trying to make a led version of a cylons eyes a delay shift or
using pwm will always result in a led coming on when not wanted. To
overcome this you need to use Mirror Imaged Bit Angle Modulation to get the
fading and reverse direction to run smoothly

On Fri, 18 Sep 2020 19:24 Mark Roth, <cab...@gm...> wrote:

> Is there a reason that when using rshift or 2/ the most significant bit
> isn't zero padded? lshift and 2* do this correctly. I was trying to
> implement a simple back and forth led on a port by starting with a value of
> 1 in the the port. Moving from right to left in the port works perfectly
> but when moving from left to right using either rshift or 2/ every other
> bit is set high so after a couple of back and forths the port is filled
> with 1s.
>
> I can see that the lshift says it is zero padded but the rshift doesn't say
> that it is. Is there a reason for this? The 2012 specs seem to say that
> rshift is zero padded as well.
> http://lars.nocrew.org/forth2012/core/RSHIFT.html
>
> Perhaps it's just my noob code that is the problem.
>
> : cylon
> >     $ff ddra ! 1 porta !
> >     begin 40 ms
> >         8 0 do porta @ 2* porta ! 40 ms loop
> >         8 0 do porta @ 2/ porta ! 40 ms loop
> >     key? until ;
>
>
> It appears to work correctly in one direction but not the other.
>
> Thanks,
> Mark
>
> _______________________________________________
> Amforth-devel mailing list for http://amforth.sf.net/
> Amf...@li...
> https://lists.sourceforge.net/lists/listinfo/amforth-devel
>