Re: [Computer-go] gtp question for topological go program

2015-06-08 Thread Xavier Combelle
It's not clear why you would want that your engine have control on
communication. Can you explain your use case ?

2015-06-08 7:28 GMT+02:00 Ray Tayek rta...@ca.rr.com:

 hi, still implementing gtp for my topological go program. it sorta works
 with gogui.

 looking at the gnugo implementation, it seems all passive in that it
 simply reacts to commands.

 so it looks like one side (or some other controller) has to be the passive
 side and react to the commands (like gnugo), and one side need to be active
 to do the setup and issue genmove commands (like gogui).

 currently my program waits for the player to move before sending it's
 repsonse (the move) back. looks like gogui uses something called:
 runLengthyCommand() for this. and some complicated code.

 is there any documentation on how to be the active side of gtp?

 is it possible for both sides to issue commands?

 any pointers will be appreciated.

 thanks

 --
 Honesty is a very expensive gift. So, don't expect it from cheap people -
 Warren Buffett
 http://tayek.com/

 ___
 Computer-go mailing list
 Computer-go@computer-go.org
 http://computer-go.org/mailman/listinfo/computer-go
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

[Computer-go] Number of Go positions computed modulo 2^64; source code available

2015-06-08 Thread John Tromp
See my updated webpage at

http://tromp.github.io/go/legal.html

regards,
-John
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] gtp question for topological go program

2015-06-08 Thread Ray Tayek

On 6/8/2015 12:58 PM, Xavier Combelle wrote:
It's not clear why you would want that your engine have control on 
communication. Can you explain your use case ?


our main use case is: two humans at different locations with my program 
want to play each other over a socket. seems like one has to be active.


the passive client has a thread that reads a socket and executes 
commands. if this program were to send a command through the socket, 
then the other side's response to the command would be read as a command.


thanks

2015-06-08 7:28 GMT+02:00 Ray Tayek rta...@ca.rr.com 
mailto:rta...@ca.rr.com:


hi, still implementing gtp for my topological go program. it sorta
works with gogui.

looking at the gnugo implementation, it seems all passive in that
it simply reacts to commands.

so it looks like one side (or some other controller) has to be the
passive side and react to the commands (like gnugo), and one side
need to be active to do the setup and issue genmove commands (like
gogui).

currently my program waits for the player to move before sending
it's repsonse (the move) back. looks like gogui uses something
called: runLengthyCommand() for this. and some complicated code.

is there any documentation on how to be the active side of gtp?

is it possible for both sides to issue commands?
...



___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go



--
Honesty is a very expensive gift. So, don't expect it from cheap people - 
Warren Buffett
http://tayek.com/

___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Number of Go positions computed modulo 2^64; source code available

2015-06-08 Thread Robert Jasiek
How much computation time do you expect to reveil the complete exact 
19x19 number? Or is more research necessary before I may ask this?


--
robert jasiek
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Number of Go positions computed modulo 2^64; source code available

2015-06-08 Thread John Tromp
dear Robert,

 How much computation time do you expect to reveil the complete exact 19x19
 number? Or is more research necessary before I may ask this?

this computation of the 64 least significant bits was 1/9 of the total
effort needed. each such job contributes 64 bits to the answer.

regards,
-John
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] gtp question for topological go program

2015-06-08 Thread Ray Tayek

On 6/8/2015 8:52 PM, Xavier Combelle wrote:
I think for this purpose stick with gtp protocol would be an handicap 
more than a bless
it looks that way, but i want to have the program work against other 
engines when it is playing normal go.


I think a simple protocol like this (loosely adapted from gtp would 
fit) ...  # Rules negotiation ...
i could do this (i think). but i would need 2 sockets since on is 
blocked on reading gtp commands.
Actually the gtp protocol could be used with one player as the master 
and the other as an engine but it would let you less flexibility and


would not illustrate well that both player are not in a strictly slave 
master relation but more a pair to pair relation


During the negotiations phase, both player could simultaneously send 
the same command in this case both would agree and it's not a problem

i may try something like this later.

thanks


2015-06-09 4:25 GMT+02:00 Ray Tayek rta...@ca.rr.com 
mailto:rta...@ca.rr.com:


On 6/8/2015 12:58 PM, Xavier Combelle wrote:

It's not clear why you would want that your engine have control
on communication. Can you explain your use case ?


our main use case is: two humans at different locations with my
program want to play each other over a socket. seems like one has
to be active.

the passive client has a thread that reads a socket and executes
commands. if this program were to send a command through the
socket, then the other side's response to the command would be
read as a command.

thanks


2015-06-08 7:28 GMT+02:00 Ray Tayek rta...@ca.rr.com
mailto:rta...@ca.rr.com:

hi, still implementing gtp for my topological go program. it
sorta works with gogui.

looking at the gnugo implementation, it seems all passive in
that it simply reacts to commands.

so it looks like one side (or some other controller) has to
be the passive side and react to the commands (like gnugo),
and one side need to be active to do the setup and issue
genmove commands (like gogui).

currently my program waits for the player to move before
sending it's repsonse (the move) back. looks like gogui uses
something called: runLengthyCommand() for this. and some
complicated code.

is there any documentation on how to be the active side of gtp?

is it possible for both sides to issue commands?
  ...



___
Computer-go mailing list
Computer-go@computer-go.org  mailto:Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go



-- 
Honesty is a very expensive gift. So, don't expect it from cheap people - Warren Buffett

http://tayek.com/


___
Computer-go mailing list
Computer-go@computer-go.org mailto:Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go




___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go



--
Honesty is a very expensive gift. So, don't expect it from cheap people - 
Warren Buffett
http://tayek.com/

___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go