Tuesday, February 21, 2006

midterm plan

------------------------A to Z mid-term proposal-------------------------------------

A text game: I am over-my-head with this, but I think I could pull out a fe w elements to start. First off, I think I should begin with text file (one player) –solitaire.

Even with one text file, because I am talking about storing text files in different structures for different types of analysis, I think I would do best to do each part in a separate code file and then work to join them, piece-by-piece. Along the way, I expect I will figure out more concise rules or play –or ditch the idea all-together.

I don’t know yet what I should attack 1st.

I like working the idea of this –but I dread even beginning the programming, because I suck at all of this.

(Can you hear my tears?)

-------------------------------------------------------------------------------------------------------------------------------------------------------------------------

******************************************************************************
Context War
*** A Text Game for Two *******************************************************************************
Because we are wedded to our own cognitive interpretation, language is conflict or communion. We flexibly mount context upon what we hear based on what we know: what we think we fear, what we think we love, and the rest we ignore.
Could “a game” inspire enough investment to create context for the users?
Relationship to context is not empirical like numbers. –-More like a debate. -------How is debate won or lost?
*******Constructionist Theory*****Start with something you care about, you something you have read and related to. Put that against (or put that toward) another’s choice
The Object of the game is "a conversation" even if an argument: It is meant to be an experiment to see if players will attach a new context to half-random expression within a random competition. If it really worked--points should not matter at all.
I’m concerned that a point-based structure to determine Win/Lose is not best way to go. Can a game be “finished” without empirical determination?
I know for me, it’s a downfall of Scrabble: when points are most important to a player, technique
beats resourcefulness: axe beats aviary when aligned on triple word square. Only using all 7 tiles in a single play has a real feeling of victory --if you can do that—it doesn’t matter where on the board it falls! There luck --your tiles--available connecting space, but its cleverness over-all.
If one player could produce context that the other could understand...
I think through actually working the programming, I would be able to resolve a more ephemeral en to the game and still make clear victory for one player.

Ultimately, if I were to complete this project, I would like to negotiate the XPort to a design a more satisfying interface for the players. And maybe Vicki could read it aloud.
---------------------------------------------------------------------------------------------------------------------- ********************************************************************************

Rules for Play (and in-part, pseudo code)
Begin Game and Set-up
Each player chooses an author and an online text file from that author.
If there is only one player against the program (me) the second text is from Herman Melville.
Players choose author (types into program)
Connect to Guggenheim Library (Spider other online text sources)
Search online text library
Search for author
Locate available text
Display list of available text files
Player must choose from text available
Import the text file
Java IO
Make new String
Store as content T1

Repeat for player 2 (or random Melville)
Make new String
Store as content T2

There are now two distinct text files:
T1 and T2
Keep them separate.
****************************************************************************************************
T1 {split into words (and indexOf)
Insert into binary tree (possible to keep indexOf and embed this number in other structure?
Later a duplicate text file will be split through regex to manage complete sentences)
}
Hold T1tree in storage
Repeat for T2
*****************************************************************************************************
Duplicate Text1 {regex for first capital letter following “.”, “!”, “?”
and all words up to next “.”, “!”, “?”
Make new Strings for complete sentences
Can these full sentences be stored in a HashMap?
Repeat for Text2
*********************************************************************************************
Turns | Points | Death | Bonuses

Begin Game:
Players take turns choosing words they believe will be found in their own text.
Game begins blank.
The First Player to chose a word that results in a display sentence receives x points.
Player1 types a word into the program
Search Tree for the word
Does the word appear in own text? True/False
If True;
Search for concordance in Land-mine String
Is the word a “Land-mine”* in that Player1’s text?
If True {
Player1 Blows-up}
Else (the word is in that Player1’s text){
The sentence that the word appears And Player1 receives x points. Turn over
(Player2’s turn)
}
If False (word in not found in text) {
Check to see if the word appears in Player2’s text.
If True {
If the word is a Land-mine {
Blow-up Player2
Else {
Player2 receives x/2 points
If False {
Do nothing And turn over (next player’s turn)
} ****************************************************************************************************
*Land-mines are randomly generated words drawn from each user’s text. If one of these words is chosen directly, the player fro ms oe whoes text it is in “blows-up” (I don’t know yet if that will terminate game). If a word is a Land-mine not directly chosen but within a sentence displayed for a word chosen, that is a good thing for that player (don’t know yet how good).
******************************************************************************************************
**Land-mine: See “Random Number Generator” below
**Make Land-mines:
Random Number Generator:
Concordance
Bayesian Analysis
Compare T1 and T2
Find common words
If a word appears in both T1 and T2
Keeping text file separate, eliminate common words.
For T1 and T2,
Analyze last IndexOf #
Take some % of the number
Locate the corresponding words attached to those numbers,
Make new String called TLm1 and TLm2
****************************************************************************************************
Next Play
àPlayer2
If Player1’s word choice did not result in a sentence written to screen,
Player2 thinks of a word and types it into the program.
Play is same as first play.
*****************************************************************************************************
If previous player’s turn did result in a sentence to screen, Player2 can choose from any
of the words displayed on screen to receive double points.

System goes though same procedure.

Points are stored and hopefully context is imagined.
*************************************************************************************************