Loadstar 213/213.d81/t.gin

Text File | 2002-01-01 | 8.7 KB | 308 lines

u
            G I N   1 2 8
         by K & R Slaminko

   This program is for the C128 in 80
column mode. That's not the way I
intended to write it, but there it is.
See my rant session at the bottom for
more info.

   Gin is a simple, old, and very
popular card game played with a
standard deck. You have 10 cards in
your hand, and if you can arrange
those cards into two sets of 3 and one
set of four, you have Gin and win the
hand. Your opponent is, of course,
trying to do the same thing.

   A 3 card 'set' consists of:
    - 3 cards of the same rank or
    - 3 consecutive cards of the same
      suit (called a Run).

   The Ace may be considered high or
low in a run, but wraparound is not
permitted. Q-K-A and A-2-3 are legit
but K-A-2 is not.

   A 4 card set is the same with an
extra card: 4 of a kind or 4
consecutive cards of like suit.

   You must achieve a hand with two
sets of 3 and one set of 4. Other
combinations (e.g. two runs of 5)
don't count. A card may only be
counted as part of one set, so if you
hold 3-4-5 clubs and the other 3
fives, you can consider this as either
a set of 4 (the four 5's) or as two
sets of 3 (the run and the other 3
5's).

   A dealer is randomly chosen, and
alternates every hand. The dealer's
opponent is dealt an extra, 11th card
and begins play by discarding one
card. The other player then has the
choice of either picking up that card
or taking one from the deck. Then he
discards and the first player is faced
with the same choice. Play alternates
this way until someone gets Gin, or
all cards in the deck are used (a
tie). Scoring is simply a matter of
tracking how many hands each player
wins.

   In Gin 128, your hand will be
displayed face up at the bottom of
the screen. Your opponent (the evil
Knees) will play from the top. The
deck and discard piles are shown in
the middle. The computer checks for
Gin after each discard.

   You can choose to play with a 1351
mouse in port 1, a joystick in port 2,
or the keyboard. Once you make this
choice, you're stuck using that
interface forever (or at least until
you restart the program). The mouse
and joystick interface involve
pointing and clicking, the keyboard
controls use different keys to
indicate your choices.

   Knees' cards may be displayed face
down unless he has Gin, face down
until one of you gets Gin, or face up.
This last option was used for
debugging the program but I decided to
leave it in as a sort of learning
tool. Of course it's cheating to peek
at your opponent's cards this way, but
who'll know?

   Your hand may be displayed sorted
by suit, by rank with Aces low, or by
rank with Aces high. Any time you're
being prompted to pick up or discard,
you can change the sort order by
pressing S (keyboard) or clicking on
the word 'Re-sort' (mouse/joystick).

   To pick up a card with the
keyboard, press 0 for the top card in
the discard pile or 1 for a card from
the deck. Mouse/joystick users can
click on the appropriate pile or,
alternatively, on the words 'discard'
or 'deck'. The card you pick up will
be inserted into your hand according
to your sort preference, and will be
highlighted by a flashing asterisk.

   Discarding is accomplished by
pressing a letter key (a-k) or by
clicking on the unwanted card.

   At the conclusion of each hand, you
may choose to play again by pressing Y
or N or by clicking on 'Y' or 'N' in
the prompt.

   As a 128 program, Gin 128 does not
return to Loadstar.

   -------------------------------

              RON'S RANT

   This was an exceptionally
frustrating project. I initially
thought Gin would be a very simple
program to write and had, in fact,
proposed it as a group project to try
to get more people interested in
programming for Loadstar. It's just
as well that I got very little
response to my call, as it turned out
[way] more difficult than anticipated.

   I'd written several strategy card
games (Hearts, Pinochle, etc) for the
defunct LS Quarterly for the 128, and
since the strategy for Gin was much
simpler, I figured Gin would be a
snap. [Wrong-O!] But since I didn't
realize that yet, I started writing
it in 128 mode since most of the
gruntwork (card design, display,
loading of the custom font, mouse
interface, etc) would be identical to
my earlier work so I could just edit
one of those old programs. The
intent was to eventually port it over
to 64 mode once the logic was
finished. Incidently, if you should
be brave enough to look at the source
code, you'll see that I stuck to
BASIC 2.0 commands rather than 7.0 so
the translation would be easier.

   As expected, programming the
strategy was pretty easy: discard the
card least like the rest of the hand,
pick up the discard if it improved
the hand otherwise pick from the
deck. No big deal. The problem was
getting the computer to reliably
recognize the sets in a hand. Oh
sure, 3 eights, 3 threes, and 4 tens
-- these are are pretty easy to spot.
But what if you have overlapping sets?
Consider this hand (currently holding
an 11th card):

Clubs:    QKA234
Diamonds: 4K
Spades:   4K
Hearts:   K

OK, you've got a run of 4 (A234
clubs), a run of 3 (QKA of clubs -
oops that uses the Ace again), 4
kings, and 3 fours. Every card is
part of a legitimate set, but 3 of
them (the K,A, and 4 of clubs) are
each part of 2 sets. This hand can
be arranged to form gin as:

    Set of 4 kings
    Run A23 of clubs
    Set of 3 fours
    Discard the Queen.

Trying to get the computer to
recognize that possible arrangement
and ignore all others was a real pain.
Yes, Knees, it really is OK to toss
that Queen even though it's part of a
set.

Turns out there's a whole bunch of
similar hands that may or may not be
Gin, depending on how the cards are
grouped. In some, every card is part
of a group and there's even a set of
four, but no matter what card you
throw away, the rest won't add up to
Gin. I couldn't have Knees just sit
there paralyzed with indecision.
Optimum grouping of the cards is one
of those tasks that the human brain
handles fairly easily (at least once
you get used to it),  but it just
doesn't translate well into computer
code.

   So I wrote some code, played a few
hands (with Knees' cards face up) and
watched the mistakes pile up. Fiddled
with the code so those mistakes
wouldn't recur and played some more.
Knees made more stupid plays. Fiddled
some more. Corrected those misplays
but some of the earlier ones came
back. Grrrr. Went on this way for a
while until everything seemed right.
Figured if I could play 20 straight
hands without an error I'd be good to
go. First 19 were fine, then on hand
#20 Knees claimed Gin when he didn't
really have it. Sigh. Fixed that and
started over. This time on hand #20 I
had Gin but the stupid program
wouldn't recognize it. Aaaarrgggghhh!

At this point I shoved the whole thing
aside for a few weeks so I could calm
down.  Finally came back to it, found
the problem, and have now played 100+
consecutive games without Knees making
a gross error. Does this mean I've
covered every eventuality? Probably
not but I've had enough of this
frustration.

I quit! I'm giving up on my intentions
of converting this program to C64 zip
BASIC. All the various patches to
handle special situations have made
the code too large & too complex for
that to be a simple matter.  Besides,
I'm tired of messing with it.  I
program for the Commodore because it's
fun -  much like solving logic
problems or doing other puzzles.  This
program is no longer fun and I want to
move on to something else.

Having said all that, and even though
I'm fairly confident that it's working
properly, I'd like to invite you all
to let me know of any problems with
Gin 128. If the computer credits you
or Knees with a Gin you don't have, or
conversely refuses to recognize a
legal Gin, make a note of the cards in
question and email them to me at

         [e-mail]

and maybe I can figure out a fix.

             End of Rant.

 RS & KS


 [Dave's Confession]: Yes, it is true.
Knees Calhoun has found the Holly
Tower! Being geographically
challenged, his only clue was an old
packet of Holly Sugar he found under
an old table leg. The packet said that
the sugar was manufactured from sugar
beets.

 So Knees put his nose in the air, and
followed the sweet smell of the annual
sugar beet migration. Unfortunately,
Kit Carson had already driven the
sugar beet herds to near extinction
providing sugar packets for Harvey
Houses along the Santa Fe Railroad.

 But by the time Knees discovered this
ecologically pathetic fact, he was
already in Garden City, KS. At the
truck stop, he asked if anyone knew
where Holly was, and some thoughtless
person said, "Sure -- just 80 miles
west on US 50."

 I discovered him moving his stuff
into the Penthouse offices on the
213th floor. A ruckus broke out, and
by the time we caught it and put it
back in its cage, we had missed the
deadline. Now you know the real reason
this issue is late!

 DMM