Sunday, March 29, 2009

Post Immersion Update 0001

CINGO! has exited the Pre-Alpha stage, passed Alpha testing, and is now called 'Imp' Beta 0.1 (Each future release will be named after a mythical creature).

Below are some screenshots of the final product for your enjoyment:

The Game Screen (Note that pieces about to be captured are clearly denoted with a red X for the turn duration):

The Change Appearance Screen now has support for the legacy 'Alpha' look from my old version:

New Instructions View (Note the helpful images at the bottom of the screen are much clearer than in the old version):

When I present my project on the 7th of April, I will try to post some images from the event. Today, I used Photoshop Elements to create a 2ft wide iPhone that I then pasted onto two stacked poster boards. I am certain that this will draw enough attention that people will actually come to see what I did; technology-based immersion projects are generally not viewed by many people because the projects have a lack of visual appeal in comparison to those involving pictures of journeys and artistic creations.

Monday, March 23, 2009

Post Immersion Update 0000

Ladies and Gentlemen......

After days of struggling, I have finally added Data Persistence to Cingo. After asking a question on a programming forum called Stack Overflow, I realized that I had to re-engineer my entire application. For the past few days, I spent much of my time recoding Cingo from scratch. Doing so has allowed me to make a much more intuitive interface, to make my code more compact and understandable, and to add data persistence.

I have also begun making a nicer icon-set, have found a way to get rid of the annoying 'Execute Move' button, and have produced a beautiful logo using Blender 3D.

Here is a quick preview of the new main menu:

...and some of the new pieces in the change appearance screen:

Friday, March 13, 2009

Today, I planned to finish data persistence and finalize my application. After many hours of coding and debugging, I was still unable to save and load data effectively. While the data appears fine after loading, my program throws random memory access errors no matter what I do, and locating these errors is practically impossible because they are not reproducible. I also met with Mr.Collias for the last time during this project, but he was unable to help me discover how to fix the errors. He suggested that I should ask questions on programming forums. I will see what I can do on the weekend, but I still have to finish an essay and take the SAT.
My application works very well, it just loses undo functionality when the view is changed and does not save games when terminated.

Quick Update (9:59):
The "Tie" functionality was broken, and I just fixed it.

Thursday, March 12, 2009

iPhone Journal (asc(Q)-63)

I only have one day of my uninterrupted programming experience left. Today, I planned to complete the views of my application and to begin working on saving data. I started out by filling the New Game screen, without which the application would be basically useless. Then, I met several people who fall within the target age group and asked them to test the application. It was very useful to observe their interactions with the menu systems and controls, and I was then able to change the interface to more closely match what people expect. For example, none of my test subjects was able to locate the way to view the instructions without searching for several seconds. Instead of using the arrows on the screen to move on, they clicked on the white section of the screen where the word "Instructions" appears. I thus placed an invisible button on this section and moved all of the arrows away from the toolbar onto the individual instruction pages.
I also caused the change appearance view to function mechanically (it still needs some aesthetic improvements such as a piece and board appearance browser) and started to rearrange my code for ease of saving key variables. In the process, I was able to fix large amounts of memory leaks and managed to discover that the >100 unused variable messages that I had been receiving were due to some variable declarations that were being imported into almost every file in my project. Finally, I began implementing data persistence itself by creating an object meant to house the data for one of the views and to save half of my data to it. The important steps now are to create a new object for storing the rest of my data and to actually save to/load from the hard drive on the iPhone.

(Note the red background squares. These result from a board appearance setting - the only one other than the default so far)

(This message is displayed when New Game is pressed and a game is in progress. Yes, I realized I misspelled "want." I fixed it in the latest Debug Build)

(The switch in this view is for controlling whether the controls are described in the game view once Begin Game is pressed. I noticed that none of my test subjects pressed the help button purposefully. The default setting is ON to make sure people know how to use the controls.)

Quick Update (12:22):

Just wanted to finish this view really quickly. Goodnight!

Wednesday, March 11, 2009

iPhone Journal (asc(R)-65)

Today, I planned to finish the undo functionality, add instructions, and to reduce memory leaks. I was able to accomplish all this and more. I finished the undo button much more rapidly than I had hoped and reduced some of the memory problems at the same time. I then created the instructions view. I soon discovered that the instructions cannot all be displayed on the same page, and so I modified the view controller for the instructions to handle multiple pages. During this process, I discovered a much simpler way to add and remove views than I had been using previously. I also fixed a major error present in the game view in regard to promoting a diamond to a mixed "power" piece.
Finally, I met with Mr. Collias in order to discuss final changes. He taught me methods of assessing the usability of a user interface and then provided several suggestions about my project's interface and about the gameplay element.
I followed almost all of these suggestions with minor adjustments and now have the following images to demonstrate my progress:

(note the red X on the piece surrounded by diamonds and the dot in the center of the X indicating the final destination of the jump.)

Mr. Collias noted that the score display was confusing and that the execute move button was unclear. Also, it was previously difficult to determine what pieces were removed by a CINGO move and what the destination of a given move is. All of these problems have now been addressed. Note that I also added a question mark icon to the game view that, when pressed, discloses information about how to use the controls.
Tomorrow, I will attempt to finish the "Change Appearance" and "New Game..." views and to begin working on saving the state of the application on view switch/exit.

Quick Update (11:32):
Just added a picker with some randomly named components.

iPhone Journal 0x10

Today (3/10/09), I managed to implement 90% of the "Undo" functionality of Cingo and to fix all of yesterday's bugs after hours of troubleshooting. I was, however, unable to work on any of the other views.
The special move for which the game is named took the longest time to add undo functionality to because the data size is variable - it is possible to surround more than one piece in a single slide or jump. I was also able to take the undo button beyond my early expectations such that it may be pressed until the game looks like it did at the very start. The only move that still needs to be able to be undone is upgrading a diamond to a mixed piece. This can be accomplished with a simple BOOL value, and I used a similar method for allowing jumps to be undone, so this should be a relatively painless process.
I still need to work on removing several memory leaks (see the spikes in the image below) that were introduced due to some strange allocation problems. Tomorrow, I will begin by finishing the undo button, fill the instructions view with information, and see if I can figure out how to remove the leaks. The latter is probably not very important at the moment because of the small quantity of leaks, and I may leave this for later.
I will then work on the change appearance view and new game view if time permits. After that, all that will be left is saving on view switch/exit and adding some other ways to win than capturing all opponent pieces. Everything seems to be coming together nicely at the moment.

(Note: the final spike is from killing the process, not from a lack of deallocation on program exit)

I would post more images, but the application has not changed in appearance, just functionality, since yesterday.

Monday, March 9, 2009

iPhone Journal 0xF

(click on images to enlarge)

Today was the start of the final week of my immersion. As promised, I linked all of the views of my application together so that the buttons at the bottom of each view work properly. I then decided to start working on the suggestions Mr. Collias provided me with. First, I edited the mixed piece icon to be more unique, and then I began to implement the Undo button. I created a linked list of object pointers to store each move and rewrote most of the screen touch code to allow each move to be planned out before completion. As visible in the screenshots below, I have added a special square selector icon that indicates the path a piece will follow as it moves. When the switch button is clicked, it now causes an entire move to be completed. With the progress I have made, actually getting the undo button to do something should not be comparatively difficult.
There are also several bugs to work out, including that a memory access error occurs when attempting to re-plan the path that a piece jumps if the jump consists of several steps. Another problem arises when an already visited square is re-visited by a piece in the same turn. I plan to make this impossible in order to make the game more fair and to reduce user confusion resulting from being unable to tell if a square is to be visited twice or not (note the image with 3 white pieces captured and a mixed piece selected). Although I have not attempted this yet, I imagine that jumping over the same piece twice from different angles with a mixed piece is currently allowed, and I need to prevent this.
I am very glad that I have not yet added data persistence because I would have had to rewrite that code had I already done so.
Tomorrow, I intend to finally remove all of the bugs I can find, finish the undo button, and see if I can get started on adding data persistence. With my newfound experience in adding the linked list functionality, the latter should probably wait until I have populated the other views with content and functionality.

Friday, March 6, 2009

iPhone Journal 0xE

Today, I was not able to make as much progress as I had hoped. While I intended to work on data persistence, I spent most of my time experimenting with code in order to change the current view without having to rewrite everything I have already produced. I was eventually able to achieve this with some interesting tricks, but did not have time to actually implement the data displayed on these views, nor did I have time to fully connect everything. By now, however, I know enough to finish this process with relative ease.
I also met with Mr. Collias today, and he provided me with some very useful feedback on usability, features, and overall look. He suggested making more of a distinction between the appearance of the pieces, producing less of a blending between the pieces and the board in terms of color, and adding undo and replay options. Furthermore, he provided a solution for making the "switch turn" button less frustrating/ambiguous and indicated that the square borders are too large.
I now hope to use me newfound knowledge of switching views to finish the interface and to use some of Mr. Collias' suggestions to make the game more playable. I then plan to finally work on data persistence.

(Sorry, no screenshots today; they would just consist of different views with short labels like "Instruction View")

Thursday, March 5, 2009

iPhone Journal 0xD

Once again, I exceeded my expectations. I finished all of the movement and jumping code, allowed for the "CINGO" move, added the ability to promote diamond pieces reaching the other side of the board, and even added a current player indicator and two counters that display the amount of pieces that each player has captured. I have allowed the iPhone to determine when a player has won, but this code does not yet take into account all possible ways to win, just the simplest.
Nevertheless, the game logic is essentially done and I now need to start working on a feature that I have used for only one application, and then did not fully understand: data persistence. I will also have to start implementing the three other views that the application will have when it is done: a new game panel for setting game preferences and beginning a new game, the change appearance panel that will allow the appearance of the pieces and board to be changed, and the information panel, which will display game instructions in a simple, readable format. I will most likely work on these in reverse order so that I will gradually become more comfortable with the code involved.
By the way, please do not judge my ability at CINGO from the video at the top of this post; it is intended to demonstrate all of the features I have currently implemented as quickly as possible, and not to demonstrate my strategic skill.

iPhone Journal 0xC

I have made immeasurable progress today! I planned to write the code for tapping a piece to select it and then tapping a different spot to move it there. In fact, I was able to finish this in the morning and early afternoon. I then limited piece selection to pieces of the correct color based on whose turn it is in the game. I implemented all of the code necessary to allow a piece to make multiple jumps but not multiple moves and then expanded the code to limit movement options based on piece type. Finally, I added the ability to capture enemy pieces by jumping over them and learned how to use Xcode's integrated debugger when my application froze and complained about memory access errors (I was trying to deallocate memory to free up RAM and ended up releasing several pointers that pointed to memory I actually needed). In addition to this movement code, I also added a button that switches the current player after a turn is complete to allow multiple jumps while not giving the user a hint as to jumps being available. A direct screenshot from my physical iPhone is presented below (Note that the white, circular pieces appear more distinct when viewed on the iPhone screen):

When I get back to work at a more reasonable time of day, I will try to finish the game-logic by allowing the "CINGO" move to be made (capturing a surrounded piece), will allow diamond pieces to become mixed pieces by reaching the other side of the board, will add code to keep track of captured pieces, and will make certain that it is actually possible to win.
After that, I will begin to work on a captured piece indicator, a current player indicator, and data persistence so that the game will be saved if quit, but I will certainly not have time to finish implementing these features. I also hope to add a short movie demonstration of moving pieces that I will add to my next journal.

Wednesday, March 4, 2009

iPhone Journal 0xB

As planned, I worked on movement animation today. I gained a better understanding of how Quartz 2D works and that at no point does the programmer actually refresh the screen; it is refreshed after the function that must contain all of an applications drawing commands. This made it incredibly difficult to create a framework for piece movement, but I at last succeeded by creating a special function that fires 24 times/second. I managed to make the drawing code incredibly efficient such that it only redraws the graphics for the region directly affected by moving a piece. I also determined how to detect where the user presses on the screen and wrote pseudocode for the function that will handle this information. The video clip below is a demonstration of me clicking a test button that is tied to smoothly animating the motion of a piece on the board. While it may not seem very significant, the true changes are visible only in the code; the same animation function that performs the movement in the video can be applied to any piece movement on the screen, even jumps.

I am very glad that I began working on this project earlier than I had intended because of all of the unforeseen problems I have encountered already and will encounter in the future. The experience of producing the application has also taught me many aspects of the Objective-C language that are not even mentioned in my textbook. Programming forums are a great reference for what commands can be used to achieve specific effects. I am getting used to using Apple's technical API documentation as well.
Tomorrow, I plan to implement most of the selection and movement code that is inherent in every move of Cingo. I shall also expand on the foundational code I wrote today to start building some of the logic functions.

Monday, March 2, 2009

iPhone Journal 0xA

After a short break on the weekend, I was able to continue working on my iPhone project with a fresh mindset. As planned, I worked on one of the graphics chapters in my textbook. To my delight, Quartz 2D, Apple's simplified graphics system for drawing on the iPhone seems capable of handling the drawing required for a board game such as my final project, Cingo. I went through the code of setting up the drawing canvas and progressed towards drawing lines of various colors, as shown below, yet I believed that I could better manage my time by skipping the repetitive sections of the chapter and looking in detail at the ones I will be using in my application.
At one point in coding the example project, I was given a ridiculous error by the compiler, something along the lines of "Unable to read symbols for "/System/Library/Frameworks/UIKit.framework/UIKit" (file not found)." This is completely meaningless as the UIKit framework was certainly included in my project. After over an hour of searching online, I was able to determine that this error results when the interface is set up incorrectly, but rebuilding the interface did not alleviate the problem. Suddenly, I glanced at a seemingly harmless warning message that I got when compiling. The entire problem was caused by forgetting an @ symbol before the quotes of a string constant.

I became frustrated with the sample application and so began to experiment with the pre-made example to see what I could do with images in Quartz 2D. Some simple tests confirmed that the API supports transparency, but that a static image may not be placed below the drawing view (canvas).
After I had sufficient knowledge of Quartz 2D for my current purpose, I began to lay out in my mind the features and structure that the Cingo application will have. Thereafter, I began to code the board drawing functionality and to design the board and pieces. A screenshot of the applications I used in this process is shown below:
On the left is the current simulation of the interface after compilation, Xcode is running in the background, and Interface Builder consists of the black window and the lower right window underneath that containing the board. I used Blender 3D to quickly render the pieces and board, and then modified these images slightly using GIMP and Preview to make them fit on the iPhone screen in a presentable format.

Tomorrow, I intend to start studying/implementing several of the following topics:
Data Persistence (save on exit), moving a piece from square to square smoothly, using a series of sprites for animation, detecting finger taps/slides, coding application logic, and creating a useable interface. I cannot meet with Mr. Collias or any of the people in his team on Wednesday, so I plan to do so only on Friday; doing so tomorrow or on Thursday is pointless. Next week, I plan to have learned most of the basic concepts I will need and will then finalize my project along with a presentation and a final essay combining my experiences.