Thanks - apart from reminding of my days of typing in ZX81 programs
from magazine listings and then trying to save space (defining
variables from text values, I seem to recall was one method) you
have inspired me to fire up my copy of QL2K and try your listing
out. I had to use an editor to strip out the CR characters as you
suggest and then QLAYT.EXE to link the file into my win1_ directory.
It mostly works until density% becomes 5 at which point it fails at
line 310 with "bad parameter", presumably because it is trying to do
a RND(1 to 0).
Any suggestions?
Well spotted!
It actually runs OK on QPC (more brownie points to Marcel). RND(1 TO
0) in SBASIC always returns 1, while even better RND(1 TO -1) can
produce negative numbers, presumably it takes the signed second
parameter as an unsigned 16 or 32 bit integer, produces a random
number in this range, then signs it all again to return it!
Should work with changing that line to 310 IF RND(1 TO 6-density%)=1
THEN
which should make sure it gives a snow density of 1 to 5, whichis what
I'd intended.
Varying that line would probably just increase the snow density even
further (and make it slower) but changing it so that it returns a
negative number would probably try to generate negative snow (the
wrong kind of snow???) and that would cause all sorts of chaos :-)
It's one of those programs you can tinker with forever, subject to the
whims of the creaded Sinclair random number generator.
When I figure out how to copy it from QPC to QemuLator (floppy disk?)
or QL2K I'll test it in QDOS. Just goes to show the dangers of
assuming that because it ran OK on one computer doesn't mean it'll
always run OK on another! I didn't know until you pointed this out
that the RND function works in this way in SBASIC.
Dilwyn Jones
_______________________________________________
QL-Users Mailing List
http://www.q-v-d.demon.co.uk/smsqe.htm
