🚀 go-pugleaf

s2874066@techst02.technion.ac.il (Eyal Peleg) writes:
>  Draw a vertical line that starts from home [0 0] and ends at [0 90],
>  This line should not be a continuous one but should be 4 pixels on and 1 off
>  something like "---- ---- ---- ---- ---- ...."

Boy, when I read this message the first time I thought "that's trivial, what's
the deal?"  But when I actually tried it I learned two things, one was the
fencepost problem about pixels, and the other is that it's even harder in
Berkeley Logo, because it tries to set the aspect ratio according to what the
graphics hardware says is right, and for my CGA laptop that comes out some
non-integer ratio.

(Berkeley Logo tries to arrange things so that FD 100 fits within the
splitscreen graphics area.  For CGA that turns out to mean that one
vertical turtle step is only (about) half a screen pixel.  If I make it
exactly half a screen pixel, I can get the desired dotted line with
PD FD 2 PU FD 3!!  If I use the setting that Berkeley Logo comes up with
on its own, there's no way I can get uniform dotted lines.  You can't win.)

In article <1af08oINN3cc@agate.berkeley.edu> bh@anarres.CS.Berkeley.EDU (Brian Harvey) writes:
>s2874066@techst02.technion.ac.il (Eyal Peleg) writes:
>>  Draw a vertical line that starts from home [0 0] and ends at [0 90],
>>  This line should not be a continuous one but should be 4 pixels on and 1 off
>>  something like "---- ---- ---- ---- ---- ...."
>
>Boy, when I read this message the first time I thought "that's trivial, what's
>the deal?"  But when I actually tried it I learned two things, one was the
>fencepost problem about pixels, and the other is that it's even harder in
>Berkeley Logo, because it tries to set the aspect ratio according to what the
>graphics hardware says is right, and for my CGA laptop that comes out some
>non-integer ratio.
>
>(Berkeley Logo tries to arrange things so that FD 100 fits within the
>splitscreen graphics area.  For CGA that turns out to mean that one
>vertical turtle step is only (about) half a screen pixel.  If I make it
>exactly half a screen pixel, I can get the desired dotted line with
>PD FD 2 PU FD 3!!  If I use the setting that Berkeley Logo comes up with
>on its own, there's no way I can get uniform dotted lines.  You can't win.)

No problem at all with Berkeley Logo on a sparc running X11. I bet my kids
wish they had this machine. I haven't had a chance to run it on my
goofy almost EGA card at home. I wrote the driver for that one, &
I suspect the resolution at 640 x 400 is good enough, but the dithered
colors are another matter. At certain angles they come unglued, giving
underlying undithered colors. I've considered tripling the lines widths to
get rid of incomplete dithering on some line angles, but I've decided
the real solution is to:
1) stay with skinny lines & use clean colors (I've got 8 clean & 8
dithered with that board), the clean ones suffice for most of our uses
2) spring for a real machine with EVGA one of these days (is EVGA at
the top or superseded? has the Christmas price drop come yet? it's been
half year since I last checked, and my sparc is spoiling me for at home work)

CGA is like tiling. The turtle drops bricks on the sidewalk.


Dale Parson, Bell Labs, dale@mhcnet.att.com

 ---------------------------------------------------------------------- 
|    "These words are too solid, they don't move fast enough           |
|     to catch the blur in the brain that flies by, and is gone..."    |
 ----------------------------------------------------------------------
|     Suzanne Vega                                                     |
 ----------------------------------------------------------------------

Ok, Harvey:  That's what I did, and you are absolutely right!  My nomitor
is 13" RGB; resolution 640 horiz pixels by 480 "vertical lines"  (don't
they call 'em pixels both ways?)  Scaling down to the portion of the
screen used by the turtle, that's 495 X 221 pixels - exactly the same as
turtlesteps, after allowing for your "fencepost" effect.

Thanks for your clear explanation of the 3/2 vs 4/1 fd count!

Dave Kressen