GNU Chess for Pilot?
Yeechang Lee
Over in comp.sys.newton.programmer a while back someone mentioned
plans for porting GNU Chess to the Newton. Since there is no decent
chess program for the Newton (yeah, I can't believe it either), they
certainly need it. Now that 3MB Pilots are a reality, I wonder
whether gnuchess could be ported to the Pilot too.
Memory is certainly a concern; the gnuchess FAQ
<URL:http://www.research.digital.com/SRC/personal/Tim_Mann/gnuchess/FAQ.html>
implies 8MB is the minimum platform for the program, but if the other
fellow felt it was feasible for the Newton, a device with 5MB total
and a heap architecture that complicates using all of the memory at
once, could it be possible for us too? Or would the 16MHz processor
ultimately prove to be the real bottleneck?
None of this, of course, is to impugn anything against Scott Ludwig's
<URL:http://www.eskimo.com/~scottlu/> PocketChess; it remains arguably
the stellar programming achievement for the Pilot almost a year after
its first release, and a shining example to all Pilot programmers of
the virtues of making a program available for free. Certainly,
though, to have gnuchess in pocket form would be a true coup.
--
<URL:http://www.columbia.edu/~ylee/>
Erich Oetting
In article <slrn5s2i0...@watsun.cc.columbia.edu>, yl...@columbia.edu
(Yeechang Lee) wrote:
> Over in comp.sys.newton.programmer a while back someone mentioned
> plans for porting GNU Chess to the Newton. Since there is no decent
> chess program for the Newton (yeah, I can't believe it either), they
> certainly need it. Now that 3MB Pilots are a reality, I wonder
> whether gnuchess could be ported to the Pilot too.
>
> Memory is certainly a concern; the gnuchess FAQ
> <URL:http://www.research.digital.com/SRC/personal/Tim_Mann/gnuchess/FAQ.html>
> implies 8MB is the minimum platform for the program, but if the other
> fellow felt it was feasible for the Newton, a device with 5MB total
> and a heap architecture that complicates using all of the memory at
> once, could it be possible for us too? Or would the 16MHz processor
> ultimately prove to be the real bottleneck?
The Mac version of GNU Chess runs fine on a 1MB Mac SE. Given this,
a Pilot or Newton port should be feasable.
... Erich
Valavan Manohararajah
Yeechang Lee wrote:
> Over in comp.sys.newton.programmer a while back someone mentioned
> plans for porting GNU Chess to the Newton. Since there is no decent
> chess program for the Newton (yeah, I can't believe it either), they
> certainly need it. Now that 3MB Pilots are a reality, I wonder
> whether gnuchess could be ported to the Pilot too.
>
> Memory is certainly a concern; the gnuchess FAQ
>
> URL:http://www.research.digital.com/SRC/personal/Tim_Mann/gnuchess/FAQ.html>
>
> implies 8MB is the minimum platform for the program, but if the other
> fellow felt it was feasible for the Newton, a device with 5MB total
> and a heap architecture that complicates using all of the memory at
> once, could it be possible for us too? Or would the 16MHz processor
> ultimately prove to be the real bottleneck?
>
> None of this, of course, is to impugn anything against Scott Ludwig's
> <URL:http://www.eskimo.com/~scottlu/> PocketChess; it remains arguably
>
> the stellar programming achievement for the Pilot almost a year after
> its first release, and a shining example to all Pilot programmers of
> the virtues of making a program available for free. Certainly,
> though, to have gnuchess in pocket form would be a true coup.
> --
> <URL:http://www.columbia.edu/~ylee/>
If there is really a great deal of interest in chess program for the
PILOT then maybe it is time to do a port of Rajah to the Pilot. I have
already written a couple of games using the GCC Pilot tools.... actually
one of them even uses alpha-beta (I wrote a Othello/Riversi variant).
Rajah's code is fairly compact (~ 32K) if you take out certain endgame
database tables, and hash tables.... only problem is that it uses
32 bit move structs all over the place and the Pilot has a 16 bit
memory bus... I will have to come up with a more compact move type.
Another problem is that certain extensions are tied to the move
structure format :(
Steve Dicks
In article <slrn5s2i0...@watsun.cc.columbia.edu>,
yl...@columbia.edu (Yeechang Lee) wrote:
> Over in comp.sys.newton.programmer a while back someone mentioned
> plans for porting GNU Chess to the Newton.
>
> Memory is certainly a concern; the gnuchess FAQ
> <URL:http://www.research.digital.com/SRC/personal/Tim_Mann/gnuchess/FAQ.html>
> implies 8MB is the minimum platform for the program, but if the other
> fellow felt it was feasible for the Newton, a device with 5MB total
> and a heap architecture that complicates using all of the memory at
> once, could it be possible for us too? Or would the 16MHz processor
> ultimately prove to be the real bottleneck?
Having ported GNUChess to Acorns RiscOS (ARM chip as in the Newton) I can
assure you that 8Mb is vast overkill - possibly to develop / compile it, but
my Arc version runs quite happily in 720Kb, and at an acceptable speed on
a 25Mhz ARM3 (A5000) where it was first developed - I'm a bit surprised by
your processor speed of 16Mhz (is this an ARM2?) but I doubt that it would
be the bottleneck. I would offer to help if I had access to such a device
but I don't, and wouldn't know where to get a 'C' compiler either.
--
Steve Dicks - Starswan Systems Ltd
Hertfordshire's premier one-man Systems Consultancy
... GUT : God's Undiscovered Thoughts
Stuart Cracraft
Perhaps a better way to go is the SCP (Stanback's stripped down
chess program) for the U.S. Robotics. Very slick. Programmed
by someone in the research group at Microsoft. They added
a front-end GUI to the SCP for the Palm Pilot. A perfect playing
partner. Strength probably 1400-1700.
--Stuart
ems
Is this available for the pilot now? The only chess program I
have ever seen or heard about is PocketChess by Scott Ludwig.
Eric
Christian Moskal
Stuart, is this available from MS's web site? I have never seen
it anywhere else. Or is is in fact Scott's app you were referring
to? Speaking of board games, anyone up to porting one of the
better Go engines to the Pilot? How's that for a challenge?
Best, Chris.
--
On 20 Jul 1997 05:16:53 GMT, "Stuart Cracraft"