| |
Main
Date: 31 Mar 2006 00:44:25
From:
Subject: 6man tablebases available
|
Now everybody get 6man tablebases free.
It works fine, read more at:
http://kd.lab.nig.ac.jp/chess/tablebases-online/
All you need is a reasonable internet connection (down/upload) and some
free hard disk storage. You need only 42 TBs files to cover over 90% of
cases occurring in human games. That is only about 200 Gb.
Regards,
Tykki
|
|
| |
Date: 03 Apr 2006 09:25:30
From: Francis
Subject: Re: 6man tablebases available
|
Which 42 files?
--Fran
<[email protected] > wrote in message
news:[email protected]...
: Now everybody get 6man tablebases free.
:
: It works fine, read more at:
: http://kd.lab.nig.ac.jp/chess/tablebases-online/
:
: All you need is a reasonable internet connection (down/upload) and some
: free hard disk storage. You need only 42 TBs files to cover over 90% of
: cases occurring in human games. That is only about 200 Gb.
:
: Regards,
: Tykki
:
|
| | |
Date: 03 Apr 2006 17:27:25
From: Joost de Heer
Subject: Re: 6man tablebases available
|
Francis wrote:
>>It works fine, read more at:
>>http://kd.lab.nig.ac.jp/chess/tablebases-online/
>>You need only 42 TBs files to cover over 90% of
>>cases occurring in human games. That is only about 200 Gb.
> Which 42 files?
That's described at
http://kd.lab.nig.ac.jp/chess/discussion-board/viewtopic.php?t=74 (which is
linked to in the text on the original page)
Joost
|
| | | |
Date: 03 Apr 2006 10:46:54
From: Francis
Subject: Re: 6man tablebases available
|
"Joost de Heer" <[email protected] > wrote in message
news:[email protected]...
: Francis wrote:
:
: >>It works fine, read more at:
: >>http://kd.lab.nig.ac.jp/chess/tablebases-online/
:
: >>You need only 42 TBs files to cover over 90% of
: >>cases occurring in human games. That is only about 200 Gb.
:
: > Which 42 files?
:
: That's described at
: http://kd.lab.nig.ac.jp/chess/discussion-board/viewtopic.php?t=74 (which
is
: linked to in the text on the original page)
Again, which 42 files?
Post it here, please, for the whole nntp world to read! :)
That http link requires too much reading and is too complicated.
I don't have that much time to waste!?
Thanks.
--Fran
|
| | | | |
Date: 04 Apr 2006 10:14:13
From: David Richerby
Subject: Re: 6man tablebases available
|
Francis <[email protected] > wrote:
> "Joost de Heer" <[email protected]> wrote
>> http://kd.lab.nig.ac.jp/chess/discussion-board/viewtopic.php?t=74
>
> Again, which 42 files?
The first forty-two in the list `Details for human games'.
However, my understanding is that tablebases don't work like that.
As I understand it, for a tablebase to work correctly, you need to
have all the tablebases for positions that could arise from that
material. So, for, say, the KQvKP tablebase, you need KQvK (to cover
cases where the pawn is captured), KQvKQ/KR/KB/KN (to cover cases
where the pawn is promoted) and KvKQ/KR/KB/KN (to cover cases where
the pawn is promoted and the queen captured). Since the forty-two
most common six-man endgames include KPPvKPP (3rd) and KPPPvKP (4th),
you need every KXXvKXX and KXXXvKX tablebase (where X represents any
piece or pawn) if things are going to work properly. This means that,
of the 365 files, the only ones not required are the seventy KXXXXvK
tablebases.
Could anyone who is more of an expert on tablebases confirm this?
> I don't have that much time to waste!?
But you do have enough time to play a game of chess against a computer
until there are only six pieces on the board. And you have time to
order a new 400Gb hard drive, wait for it to be delivered and fit it
to your computer. And you have time to download 200GB of files.
Dave.
--
David Richerby Electronic Fluorescent Toy (TM): it's
www.chiark.greenend.org.uk/~davidr/ like a fun child's toy but it'll hurt
your eyes and it uses electricity!
|
| | | | | |
Date: 05 Apr 2006 13:25:21
From: gimmick
Subject: Re: 6man tablebases available
|
While what you've stated is true: that "you need all derivative
tablebases for a given position"...
It doesn't matter in practice so much as the chess engine that's using
the tablebases will have an overwhelming force advantage. Enough for
it's computer brain to squish the opponent, computer or otherwise.
Also remember the engine will think or fall back on the 5 man bases at
it's own discretion.
G.
David Richerby wrote:
> Francis <[email protected]> wrote:
>> "Joost de Heer" <[email protected]> wrote
>>> http://kd.lab.nig.ac.jp/chess/discussion-board/viewtopic.php?t=74
>> Again, which 42 files?
>
> The first forty-two in the list `Details for human games'.
>
> However, my understanding is that tablebases don't work like that.
> As I understand it, for a tablebase to work correctly, you need to
> have all the tablebases for positions that could arise from that
> material. So, for, say, the KQvKP tablebase, you need KQvK (to cover
> cases where the pawn is captured), KQvKQ/KR/KB/KN (to cover cases
> where the pawn is promoted) and KvKQ/KR/KB/KN (to cover cases where
> the pawn is promoted and the queen captured). Since the forty-two
> most common six-man endgames include KPPvKPP (3rd) and KPPPvKP (4th),
> you need every KXXvKXX and KXXXvKX tablebase (where X represents any
> piece or pawn) if things are going to work properly. This means that,
> of the 365 files, the only ones not required are the seventy KXXXXvK
> tablebases.
>
> Could anyone who is more of an expert on tablebases confirm this?
>
>
>> I don't have that much time to waste!?
>
> But you do have enough time to play a game of chess against a computer
> until there are only six pieces on the board. And you have time to
> order a new 400Gb hard drive, wait for it to be delivered and fit it
> to your computer. And you have time to download 200GB of files.
>
>
> Dave.
>
|
| | | | | |
Date: 05 Apr 2006 01:36:47
From: Francis
Subject: Re: 6man tablebases available
|
Of course, congratulations, you solved the riddle. The original post and
claim is of course ridiculous!
--Fran
|
| | | | |
Date: 03 Apr 2006 18:23:06
From: Joost de Heer
Subject: Re: 6man tablebases available
|
> : That's described at
> : http://kd.lab.nig.ac.jp/chess/discussion-board/viewtopic.php?t=74 (which
> is
> : linked to in the text on the original page)
>
> Again, which 42 files?
Read post 1. There's a whole list. Every line with a TB has a number before
it, starting with 1, the next one one higher than the previous one. Now which
42 tablebases would it be? The bottom?
Joost
--
Take my love, take my land
Take me where I cannot stand [Firefly]
I don't care, I'm still free
You can't take the sky from me
|
| | | | | |
Date: 04 Apr 2006 02:41:36
From: Francis
Subject: Re: 6man tablebases available
|
"Joost de Heer" <[email protected] > wrote in message
news:[email protected]...
: > : That's described at
: > : http://kd.lab.nig.ac.jp/chess/discussion-board/viewtopic.php?t=74
(which
: > is
: > : linked to in the text on the original page)
: >
: > Again, which 42 files?
:
: Read post 1. There's a whole list. Every line with a TB has a number
before
: it, starting with 1, the next one one higher than the previous one. Now
which
: 42 tablebases would it be? The bottom?
How would I know? I'm not an expert on tablebases. So why not just list the
42 here so we all can know what you know?! Then I'll run right out and buy a
big hard drive to download them, saw a 400GB for about $200. Also noticed a
1 terabyte drive for about $800 but that's too much...
--Fran
|
|
