Return Styles: Pseud0ch, Terminal, Valhalla, NES, Geocities, Blue Moon.

Pages: 1-4041-

DataBase-chess

Name: n3n7i 2011-08-27 22:56

...A database almost wholly built of Keys.. =)

[tblPiece]    [tblGroup]    [tblHalf]    [tblFull]
*PID        *GID        *HID        *FID
Px        ^PID0        ^GID0        ^HID0
Py        ^PID1        ^GID1        ^HID1
PVal        ^PID2        ^GID2
        ^PID3        ^GID3

*X = Primary Key X
^Y = Look-up Key Y

Px / Py / PVal << Currently the only Solidly-Represented-Data...?

Name: n3n7i 2011-08-27 23:27

Ugh... try again..
[tblPiece]    [tblGroup]    [tblHalf]    [tblFull]
*PID          *GID          *HID         *FID
Px            ^PID0         ^GID0        ^HID0
Py            ^PID1         ^GID1        ^HID1
PVal          ^PID2         ^GID2   
              ^PID3         ^GID3


It takes 16 Pieces / 4 Groups / 1 Half >> To produce the first Fullmap (Halfmap doubles up // both sides being identical to start with /+/ Inverted black 'Py' Position [per black halves])

-Compression?-
Each of the Next 20 Fullmaps (the first 20 moves) Require only one extra Piece entry [15 / 16 Pieces don't move!] + 1 extra Group entry, 1 Halfmap and the new FullMap entry...

-More?-
From there, the next possible 400 maps [20 Black first moves per each of 20 of whites] require NO extra Half-maps [All have already been created by whites identical moves...], thereby no extra Groups or Piece entries either, and only the new Fullmap entries are added....

...Not bad eh? =)

Name: n3n7i 2011-08-28 3:14

Now....
FullMap -> FullMap [Initial Synaptics?]
-*Should Allow Reading x Plies Deep / 4 moves wide Fairly Quickly?- //Read straight from file?!

[tblAlpha]
*AID [or ^FID] (AID <---> FID) One to One relation to Fullmap Entries
^FID0
^FID1  }- Next Four -Best?- maps
^FID2    // Fixed numbers, like 4 are a bit of a must (atm.?)..
^FID3

...Not so much 'Learning' as just 'Remembering'...? Hopefully quite efficiently too...Recursive Assessments [over deeper plies] can only Improve in accuracy > ??

Just needs a 'smart' // Data-driven?? Assessment Function...? That would be the key? =)

Name: Anonymous 2011-08-28 3:51

crank

Name: Anonymous 2011-08-28 4:14

Kill yourself faggot.

Name: Anonymous 2011-08-28 4:26

Looks like the OP just re-invented the Trie which is already used in databases and for character classification in many of Unicode text processing libraries.

http://en.wikipedia.org/wiki/Trie

What's up with the these types? It's like they drop out of first year CompSci due to mental-health problems or autism, but they don't give it up as a personal hobby and go on to come up completely retarded reimplementations of existing algorithms with their own nomenclature and try to pass it off as something new.

Name: Anonymous 2011-08-28 5:05

fuck i love autism

Name: Anonymous 2011-08-28 5:42

When civilization fails, >>6 will be struggling to program without his books while >>1 will be writing shit all day.

Name: ( ≖‿≖) 2011-08-28 6:32

>>8
derp
( ≖‿≖)

Name: Anonymous 2011-08-28 6:36

>>6
I reinvented RC4 without knowing it, in the days before I had Internet access.

"Great minds think alike."

Name: Anonymous 2011-08-28 6:37

WAABT!

Name: Anonymous 2011-08-28 6:38

>>10
I invented binary trees, B-trees and hash tables in the days before I had Internet access. Then I found out I wasn't really all that clever :/

Name: Anonymous 2011-08-28 6:42

>>10
No. Great minds think differently of the same problems. That's what make them great.

Name: Anonymous 2011-08-28 6:55

I invented message-passing object oriented programming before I've seen Win32 API. I've been destroying my mind with drugs and alcohol ever since, a gesture that would remain unappreciated by the world.

Name: !L33tUKZj5I 2011-08-28 9:26

I reinvented CSS using shell scripting to generate the HTML before I knew what CSS was. This was ridiculous.

Name: Anonymous 2011-08-28 9:31

>>15
Your greatest achievement. And you're still ridiculous.

Name: !L33tUKZj5I 2011-08-28 9:37

>>16
My greatest achievement is annoying you so much that you have to skip from thread to thread, bleeding from the ass to insult me whenever you see me.

Name: Anonymous 2011-08-28 12:08

>>17
No, that's me, actually.

Name: !L33tUKZj5I 2011-08-28 12:26

In that case, I have achieved twice as much.

Name: 1000 posts 2011-08-28 12:45

1000 posts

Name: Not >>16 or >>17 2011-08-28 12:56

>>19
I remember putting a tripfag down in a big thread, but my memory is fuzzy now. Was that you? Actually don't bother.

Name: VIPPER 2011-08-28 13:02

Im taking a JEWS in this thread.

J
 E
  W
   S

Name: !L33tUKZj5I 2011-08-28 13:10

>>21
Don't think so.

Name: n3n7i 2011-08-30 5:47

*Drags this thread back up

Trie's sound quite interesting... Did you read about Fredkin? =) What a champ =D

He was mucking around with Chess and AI too

I take it you can use these types of structures for a few kinds of data... Natural Language aka 'talking robots' maybe?
...You might need a few more than one trie, i think

...Someone make a Universal Data storage / retrieval trie ?

Is there such a thing as an information-base // knowledge-base?? >In electronic terms? ...I don't want data, I want knowledge =)

Name: Anonymous 2011-08-30 6:10

>>24
Look at the Unicode Consortium paper on Folded Tries, which nests them like in the OP post to achieve optimal compression and performance.

http://unicode.org/iuc/iuc21/a347.html

Name: n3n7i 2011-08-30 9:43

^^

Swap chess pieces for pixels

Train on faces / objects??

Name: Anonymous 2011-08-30 11:21

>>26
Penis in your mouth. That's what you mean.

Name: n3n7i 2011-08-30 20:22

>>26
If i was talking about you?

Why troll /prog ?   //  Why /prog troll(s) ?
...Average anon doe'nt/can't code?? =( poor you... go google some tutorial's =)

>>4,6,7,8... I'm Quite impressed by this thread, it's sort-of bordering on being constructive =) And i've learnt about Trie's =)

Name: n3n7i 2011-08-30 21:02

So, Folded == Nested... you would count just [tblPieces] & [tblGroup] as the first(?) // base trie ?

[tblGroup] & [tblHalf] as the second // first nest of the base?

What about Branched & Nested ??
Eg. Alpha(Computer's Best) And Beta(Player's Best)??

Name: Anonymous 2011-08-30 21:13

sdlfjsad spoiler

Name: n3n7i 2011-08-30 21:25

A database is just two (or more) tables in a one to many relationship ? (Not neccessarily? but commonly...)

A trie is that too, but also has a secondary ratio (named '?') which is greater than one to one (1:2 or 2:1??)

So the relation becomes (two?) to many?? or one/one : many/many?

...Terrible explanation i know

Name: Anonymous 2011-08-30 21:31

databaseeeee adfjaskjdfhaks

onetomanymanytoonejasdfkjahsgkjsfg

trinary trinary trinary trinary trinangry trinagrardfjkaks

u must vizit zorkleeeeeeeeeeeeeetttttttttttttt

FISTING

Name: Anonymous 2011-08-30 21:33

DUBBBBZ

Name: n3n7i 2011-08-30 23:42

Name: Anonymous 2011-08-30 23:50

>>34
Why don't you go back, you and your stupid emoticons whence you came from?

Name: n3n7i 2011-08-31 0:01

...Lower level branch would make a better example maybe

Eg. Have another table connected to the [tblHalf] table (tblHalf is where half-boards are kept// eg all of whites pieces or all of blacks...)

Our new table can be called [tblNextHalf] and may store the next n Related Halves... (a -compact?- direct Half-Half table [Movestore ?]) // similar to Alpha/Beta :: FullMap (direct Full-Full table)

Name: 2011-08-31 0:06

Name: n3n7i 2011-08-31 0:40

...Some data?
[tblPiece]
/* PID is a primary Key // Must be unique for each entry */

PID: 0,    Px: 0,    Py: 0,    PVal: 2  (<< One table entry//
on file would just be 0,0,0,2 )

//^Piece id// x-coord// y-coord// Pieceval (2 = Castle)
//0,0 is Top-left corner

PID: 1,    Px: 0,    Py: 1,    PVal: 1
//positive Y coord is moving forward // (Pieceval 1 = Pawn)

PID: 2,    Px: 1,    Py: 0,    PVal: 3 (Horse)
PID: 3,    Px: 1,    Py: 1,    PVal: 1 (Pawn in front of the horse)
[/tblPiece]    /--------------

And the group table
[tblGroup]
GID: 0,    ^PID0: 0,   ^PID1: 1,   ^PID2: 2,    ^PID3: 3
[/tblGroup]

// ^Primary Key // Piece ID lookup 0 (Val 0 >> returns [0],0,0,2)// Lookup 1 (V1 >> [1],0,1,1) // (Val 2 >> [2],1,0,3) // v3...

Piece
0,0,0,2
1,0,1,1
2,1,0,3
3,1,1,1

Group
0,0,1,2,3

...etc referencing Group id [0...n]

Name: Anonymous 2011-08-31 3:09

Poor fuck, bumping his own thread in the hope of someone being suddenly interested.

Name: Anonymous 2011-08-31 3:55

>>39
No, he does it to troll you, and with moderate success to boot. Stop responding.

Name: Anonymous 2011-08-31 6:45

>>40
But you just responded, generating more interest in the thread. It's like he's converted you to troll. Or maybe like he fucked you bareback and dropped his load in your ass and he was TROLL+ and now you have TROLL. (this disease is incurable btw)
Oops forgot to sage!!

Name: ENTERPRISE MEME FAN 2011-08-31 7:23

>40
Is retarded.
>41
Is a troll

ENTERPRISE POST IDENTIFIERS

Name: n3n7i 2011-08-31 7:40

...Wonder how useful trie's would be for natural language processing..?

Can a computer understand words?
>> is that a Circular logic problem?? Eg. How do you define the meaning of a word, without using more words..?

Name: Anonymous 2011-08-31 9:53

How do you hax an anus without first haxing an anus?

Name: Anonymous 2011-08-31 12:33

>>43
>How do you define the meaning of a word, without using more words..?
That called non-verbal communication. Its actually how children learn the language;
Equate some object/action/event with mental picture described by a word uttered by person who already has has the mental picture.

Name: 2011-08-31 13:01

Name: 2011-08-31 13:01

Name: 2011-08-31 13:02

Name: 2011-08-31 13:02

Name: 2011-08-31 13:03

Name: 2011-08-31 13:20

Name: Anonymous 2013-06-19 1:27

good thread wasshoi

Don't change these.
Name: Email:
Entire Thread Thread List