Chess for HP48GX
Samuel Herzog
Does anyone know if a chess for the HP48GX exists ?
Tony (Makoto) Kobayashi
A few weeks ago, I posted a chess board program that lets two people play
chess on an HP48. I know that there is a single player chess program but it's
rather slow and the author said that the opponent is not very good. If you're
interested in my chess board program, it's at:
ftp://ftp.cis.com/pub/hp48g/uploads/
There are two files:
cmChess.dir -- the Critter Mayhem Chess directory
cmChess.inf -- the information file
There are also two files for a Go board program.
cmGo.dir -- the Critter Mayhem Go directory
cmGo.inf -- the information file
If you have any comments or suggestions, I would be happy to hear them.
Mach
D Kirkland
: http://www.abc.se/~m9521/hpfiler/hpches08.zip
: this one is very well done!
At over 46K? No way!
It is very slow and plays very bad.
Surely this could be done with less than 46K!
dan
Franck ZIBI
I am the author of HpChess, and I am making a new version.
If you have some suggestions, comments ( apart from the size (46ko) or
the strenght,(yes, it is a weak opponent)):
don't hesitate !
================
In the next release :
using SHUTDN
HpChess gives up in a lost position
En passant move
using clock
7 levels of playing (including blitz)
+ any other suggestions ! :-)
Franck
Jorge M M Cavalheiro
: : >Does anyone know if a chess for the HP48GX exists ?
: : http://www.abc.se/~m9521/hpfiler/hpches08.zip
: : this one is very well done!
: At over 46K? No way!
: It is very slow and plays very bad.
The problem with chess is that the heuristics are quite complex and the algorithm (minimax-alfa-beta) is recursive. So to get a 'decent' chess game, you need to go at least to ply 6! Hpchess also has a opening library that takes a lot of space too. Finally the speed of the HP48 is simply not enough...
: Surely this could be done with less than 46K!
: dan
--
--
Jorge Cavalheiro
l36...@alfa.ist.utl.pt
-= LEIC =-
Instituto Superior Tecnico, Lisbon, Portugal.
D Kirkland
: Hello !
: Franck
Mine WILL be better!
Just something to think about...
dan
D Kirkland
: : : http://www.abc.se/~m9521/hpfiler/hpches08.zip
: : : this one is very well done!
: : At over 46K? No way!
: : It is very slow and plays very bad.
: : Surely this could be done with less than 46K!
: The problem with chess is that the heuristics are quite
: complex and the algorithm (minimax-alfa-beta) is recursive.
: So to get a 'decent' chess game, you need to go at least to
: ply 6! Hpchess also has a opening library that takes a lot
: of space too.
:
:Finally the speed of the HP48 is simply not enough...
I said this because I *have* written a chess program that is
*much* faster and plays better. In it's current unfinished
form, it does this at under 5K! I hope the keep the program
under 10K and I am sure it will stay under 15K (the first
version anyway).
A 3 or 4 ply search will be able the beat most HP48 users.
Yes, hpchess uses an opening library. While it may be large
in HP48 terms, it is very small as for opening libraries.
One can make a good move and still take hpchess out of it's
opening library.
My program does not have an opening library, yet it still
makes pretty good moves. It only takes a couple moves to
take hpchess out of it's opening library, then it kills poor
hpchess.
The speed of the HP48 not enough? That depends. I think my
game, when I get it finished, will beat most HP48 users and
will not be *too* slow doing it.
Anyway, hpchess will have to get a *lot* faster and a bit
better to ever compete against my program.
Mine WILL be better!
dan
Makoto Miyamoto
>Does anyone know if a chess for the HP48GX exists ?
Try HPCHES08.LIB or HPCHESS.GX on
HPCVBBS.EXTERNAL.HP.COM/dist/hp48g/games/board
They are written in ML/Sys-RPL. HPCHES08.LIB has two levels and is
rated by its programmer to have an ELO of aprx. 1200, if I understand
the french DOC right (my last french lesson, is about 3 years back).
It has two limitations:
You can't undergraduade a pawn, and you can't capture en passant. But
thats OK I think. Another limitation: its about 45 kB!!! But its worth
the download.
Hope that helped,
Best wishes and
Tschuess from Germany.
>---------------------------------------------------------<
/ "One Ring to rule them all, One Ring to find them, \
\ One Ring to bring them all and in the darkness bind them" /
>---------------------------------------------------------<
/ Makoto Miyamoto \
\ EMail:Makoto....@rz.ruhr-uni-bochum.de /
>---------------------------------------------------------<
Olivier Mueller
On Thu, 2 Nov 1995, Makoto Miyamoto wrote:
> eche...@ztl.ch (Samuel Herzog) wrote:
>
>
> >Does anyone know if a chess for the HP48GX exists ?
>
> Try HPCHES08.LIB or HPCHESS.GX on
> HPCVBBS.EXTERNAL.HP.COM/dist/hp48g/games/board
> They are written in ML/Sys-RPL. HPCHES08.LIB has two levels and [...]
And for SX, is there also a good chess program (ML/SysRPL) on the market ?
Thanks for any answer !
Olivier
--
---------------------------------------------------------------------
Olivier MULLER, Student in Electrotechnology, | Engelsteinstr 14
Swiss Federal Institute of Technology, Zurich | CH-8344 BAERETSWIL
<omue...@stud.ee.ethz.ch> | Switzerland
---------------------------------------------------------------------
D Kirkland
: In article <474etm$s...@news.cc.utah.edu>, dbk...@u.cc.utah.edu (D
: Kirkland) wrote:
: > Mine WILL be better!
: With respect, Kirkland, whoopy-ding. Get a product out, *then* brag about
: it. Franck already has a chess program out, working, that in spite of its
: limitations still beats me about 30% of the time.
Oh great! Just what I always had in mind! Write a chess program
that, at it's top level, beats some wimpy chess player about 30% of
the time!
Sorry, but if this had been my goal I would have had the program
out 6 months ago. And it would have been MUCH smaller than 46K!
(Less than 10K.)
[With respect? Strange start considering the rest of what you said.]
: So I hope yours *will* be better. But any joker with an internet
: connection can brag on USENET. Franck actually *did* something. Hope you
: do too.
Heh! No, you better just stick with HPChess. When my program
finally does come out (6 months, a year, two years, ten years,
whatever...), please do NOT bother trying it. ANYBODY that is
happy with a chess program only able to beat them 30% of the time
doesn't need a REAL chess program!
Besides, if I were to actually release the chess program, then
I would no longer be able to say:
Mine WILL be better! :)
dan
L W Steventon
: In article <474etm$s...@news.cc.utah.edu>, dbk...@u.cc.utah.edu (D
: Kirkland) wrote:
: > Mine WILL be better!
: With respect, Kirkland, whoopy-ding. Get a product out, *then* brag about
: it. Franck already has a chess program out, working, that in spite of its
: limitations still beats me about 30% of the time.
: So I hope yours *will* be better. But any joker with an internet
: connection can brag on USENET. Franck actually *did* something. Hope you
: do too.
Yes.... and any id... err, intelectual can flame him for doing so. I have
known Dan for a while and if he says that his beta verison is currently 5K
and beating HPCHESS, then it is so.
Would you release ANY code to world that was not completely bug free to the
best of your knowledge? I think not! Dan has been working on his code for
quite a while now, and I'll bet that the wait will be worth it.
-Bye
-Richard
turnerjh
: Heh! No, you better just stick with HPChess. When my program
: finally does come out (6 months, a year, two years, ten years,
: whatever...), please do NOT bother trying it. ANYBODY that is
: happy with a chess program only able to beat them 30% of the time
: doesn't need a REAL chess program!
Well, with the HP's limited memory, I assume you're not hashing. And
with the HP's speed, I'd be suprised if you got 3 plies in a reasonable
amount of time. Just how are you going about making it better than the
existing one? Have you ever written a chess program before for any other
platform?
Enquiring minds...
--
Chip Turner -- turn...@ctrvax.vanderbilt.edu
"Chess is a beautiful mistress to whom we keep coming back, no matter
how many times she rejects us." -- Bent Larsen
Steven J. Edwards
>In fact, bring on HP48 chess programs from other programmers!
>I will be very happy when my program trounces them all! :)
HP48 programmers interested in chess may also test the strength of
chessplaying programs by using some of the various test suites that
contain positions from actual master play.
For chess items, use anonymous ftp at chess.onenet.net:
pub/chess: general chess directory
pub/chess/PGN: Portable Game Notation directory
pub/chess/PGN/Standard: ASCII version of the PGN Standard
pub/chess/PGN/Standard.TOC: Table of Contents for above
pub/chess/PGN/Events: directory of directories of events by year
pub/chess/PGN/Players: directory with many PGN games by player
--> pub/chess/Tests: directory with many chess program test positions
--> pub/chess/Tests/Manifest: description of EPD test files
pub/chess/TB: endgame tablebases
pub/chess/TB/README-TB: tablebase decyphering documentation
pub/chess/TB/tbt.c: ANSI C tablebase test harness
pub/chess/PGN/Tools: PGN tools and utilities directory
pub/chess/Unix/SAN.tar.gz: Standard Algebraic Notation source kit
The test suites are all in EPD format which is explained in section
16.2 of the PGN Standard. I recommend starting with the BK.epd file
(24 problems) and then the WAC.epd test (300 problems). Most
chessplaying programs for PC class hardware can easily solve 80% of
WAC in under three minutes per move.
-- Steven (s...@mv.mv.com)