Last ditch effort

Summary:

Well, most of what I planned for this competition has fallen by the way side. As a last ditch effort...

Last ditch effort

PiscesMike 18 Aug 2016 18:49

Greetings and salutations fellow programming enthusiasts!

As you may or may not know, what I had planned for this competition pretty much collapsed. It appears I'm not the only one who has had this happen, or people just gone off the tracks in a completely different area. That's a little comforting actually, just knowing I'm not alone in my abject failures. Honestly, my work may have been doomed from the beginning. I just can't keep myself interested in doing Breakout again - I probably should have realized how futile it was earlier, and moved on to the second (or third, or fourth) game I had been planning to do. That being said, I did learn some neat things reworking the framework behind the scenes, and who knows… maybe one of these days I'll actually finish up on that project. It's not like it was a complete waste with the refactoring, and since I had the incident where I lost all of my work from the previous couple years, I do need a working 'complete' version again at some point.

Now, after all of that has been said, I've decided to move in a new direction. There are still 17 days, 5 hours left for this competition. So, I've decided to do something simple and make a 'saving throw' or 'hail mary' final try at this thing. I'm going to attempt a simple checkers game. My first thought was to go ahead and make this like my previous XNA games, with a simple board and checker pieces. However, after working through some of the object layout last night and getting some basics set up in the project I've come to the conclusion that I actually need to do something a little simpler.

RBWhitaker has been an incredible asset in all of these competitions and collaborations, and one thing that he has taught us is separating out the rendering portion of the code from the internal workings of the game itself. This is a great idea, a great plan, and I'm sure the 'professional' way to do things. However, coming from a 'google search and copy the code' background mostly, it's been a little hard for me to implement. I've been used to just making the game objects, and each object tracks its own textures, positions, etc. Not that that is a bad or improper thing, its just different, and tends to lead to problems that could be avoided if things were done the 'correct' way. One of the big issues doing things my way is the ability to pop in a new renderer for a new platform, recompile, and release. So, taking that into consideration I've decided to focus on structure and reforming my horrible code practices so it's a little closer to something I would be proud to release.

So, my plan for the moment is to create a checkers game - a console game, at that. I figure if I can get things working the way they're supposed to using a console, adding in a 'graphics' renderer near the end should be a breeze. The whole point of these exercises is to move forward, learn new things, and grow as a developer. I think this approach will work very well for that.

Well that's it for the old PiscesMike at the moment. Hope you all are making decent progress on each of your projects, and we will see you in the chat room soon!

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Re: Last ditch effort

PiscesMike 19 Aug 2016 01:43

Evening all!

Well, I had planned to get in 3 hours last night on this project, but things came up. On the bright side though, since I was off today I was able to get some things done. Actually more than 3 hours, so I'm happy with the progress.

I got things laid out and the project set up, got a simple renderer going, and the console box now displays the 'board', the 'checkers', and accepts a line of input from the user for a 'move'.

So I'm actually ready to start working on the game play which means some simple AI. I'm deciding now how to achieve that actually. Kind of getting a little worn out for today so I don't know how much more I'll get done, but I'm off to a pretty decent start, and once I made some final decisions on how I want the AI to work, things should move along quickly. Updates tomorrow, with probably a screen cap.

Have a great evening!

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Re: Last ditch effort

rbwhitaker 19 Aug 2016 02:23

We talked about this in the chat room, but I would venture a guess that doing two human players in a hot seat style game would be easier to implement than AI. Definitely, it would let you get rolling with the gameplay without getting bogged down in AI. A good Checkers AI will probably take some time, and (my personal opinion) I think the basic gameplay is more important than AI.

At any rate, I'm really happy to see this last ditch effort from you. I'm in the same category as you, that I had grand plans at the beginning of the 100 days, and then somehow, something manages to pop up and make a huge mess of things. Scaling the original plan back (or trashing it and starting fresh like you and I both did) is a very reasonable thing to do. It's definitely way better than the alternative of "give up and do nothing for the competition"!

So good luck on this new submission to the competition! Looking forward to seeing where it goes!

Re: Last ditch effort

Brett208 19 Aug 2016 07:28

PiscesMike,

What do you mean by console game? I usually think of console game being an XBOX or Play Station game, but it sounds like you are talking about something else.

Re: Last ditch effort

PiscesMike 19 Aug 2016 09:33

Hey Brett208! When I speak of a console app, I'm talking about the old DOS text mode window you can use in Visual Studio. You can make your XNA game display a console window by changing a setting in it's properties. It's useful for displaying debug info and other stuff. Anyways, here's a screen cap of what it looks like if it helps:

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Re: Last ditch effort

PiscesMike 19 Aug 2016 10:02

Hey RBWhitaker!

I think you're right about implementing a hot seat version with two (human) players in that it would be easier, and therefore quicker. My thoughts so far have been actually how to do that and create a list of 'valid moves'. In other words, I can approach it in a couple different ways, and the first two that I thought of were: (1) Get the 'move' from the player, and do a calculation to see if it's valid, or (2) Before the round is made, create a list of 'moves' which are 'valid' for each piece and then check the user's input to see if it's in the list and valid. I've been kind of leaning towards number 2 since that would make implementing the AI fairly straight forward. My big concern at the moment is how to actually make those calculations, I see a long list of if statements in there and it seems like there should be a better way to do things.

Interestingly enough, just writing that out has helped a little bit. I have things set up to work on a grid that has letters for the columns, and numbers for the rows (see picture here above or below somewhere of the console screen cap) so moves are made like: B3-B4. Thats all fine and good for displaying things and getting moves from the user, but I've never actually implemented a (somewhat) complex set of rules for movement around the board, and capturing pieces. In the text renderer, I convert things over to a string[9,9] array. It may be easier to implement the logic using the string array since the output move(s) would be something like 2,3 - 2,4. Anyways, just rambling at this point, which is where I was when I stopped last night so I'm going to give it a little more time and see if inspirations hits.

Unfortunately, work starts at 5pm this evening, and I won't be home until late at night so I won't be around chat unless it's 4-8am in the morning here. Hopefully we'll get a chance to catch up sometime this weekend!

Take care, and good luck all!

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Re: Last ditch effort

Brett208 20 Aug 2016 06:04

Thanks PiscesMike,

I didn't know you could switch XNA into a Console Application!

Re: Last ditch effort

PiscesMike 20 Aug 2016 09:22

Technically speaking, you can't! That's still a problem actually. I hijacked the Draw method and rerouted it to a Console.Readline();. It took a little tinkering to actually get the console window to have focus. Anyways, the actual program is a little bit of a mess, I just stole a nice looking picture from it.

You can output stuff to the console window in XNA though, and I'd imagine in Monogame and anything else. It's just a useful tool to check things out without having to add spritefont assets and such. A lot of programs use it for various things, and usually it's suppressed but you can display it if you like.

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Re: Last ditch effort

PiscesMike 19 Aug 2016 10:44

As a side note, it's been so long since I've played checkers I had to look up the rules. It's a bit different than what I recall and the logic stuff might be a bit easier that what I was expecting. I didn't realize the pieces had to move diagonally - I thought they moved forward when not making a 'jump'. I also didn't realize you HAVE to take the jump if one is available, or that you HAVE to take all the jumps available. Anyways, just rambling again.

Here's the rule set I'm referencing:

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Last edited on 19 Aug 2016 10:46 by PiscesMike Show more

Re: Last ditch effort

PiscesMike 29 Aug 2016 19:55

Greetings all!

My project is progressing slowly. I seem to be stuck on an error where the game logic is returning the incorrect piece (and data from the location the piece is supposed to move to). Other than that, things are going swimmingly.

Anyways, this post is mostly a question for RBWhitaker since I know I won't be able to catch up to him for a few days in chat. We were discussing my issue last night before I had to go for chores and couldn't quite nail down the issue. So, here goes:

RBWhitaker, here's the code I use to initialize the board data:

        public PieceType[,] BoardData = new PieceType[8, 8]
         { {PieceType.Invalid, PieceType.Black, PieceType.Invalid, PieceType.Black, PieceType.Invalid, PieceType.Black, PieceType.Invalid, PieceType.Black},
            {  PieceType.Black, PieceType.Invalid, PieceType.Black, PieceType.Invalid, PieceType.Black, PieceType.Invalid, PieceType.Black, PieceType.Invalid},
            { PieceType.Invalid, PieceType.Black, PieceType.Invalid, PieceType.Black, PieceType.Invalid, PieceType.Black, PieceType.Invalid, PieceType.Black},
            { PieceType.None, PieceType.Invalid,PieceType.None,  PieceType.Invalid,PieceType.None, PieceType.Invalid, PieceType.None,  PieceType.Invalid},
            { PieceType.Invalid, PieceType.None, PieceType.Invalid, PieceType.None, PieceType.Invalid,PieceType.None, PieceType.Invalid, PieceType.None },
            { PieceType.Red, PieceType.Invalid, PieceType.Red, PieceType.Invalid, PieceType.Red, PieceType.Invalid, PieceType.Red, PieceType.Invalid },
            { PieceType.Invalid, PieceType.Red, PieceType.Invalid, PieceType.Red, PieceType.Invalid, PieceType.Red, PieceType.Invalid,PieceType.Red },
            { PieceType.Red, PieceType.Invalid, PieceType.Red, PieceType.Invalid, PieceType.Red, PieceType.Invalid, PieceType.Red, PieceType.Invalid } };

Piecetype is just an enum that contains 6 items: Red, Black, RedKing, BlackKing, None (empty space), and Invalid (the black spaces on the board are never used). So, my question is, is there something going on there that I'm missing? From what I thought I understood, that statement should fill the BoardData array in the order specified - that is, with the black pieces at the top of the board, and the red pieces at the bottom of the board. So, for instance, an entry from the user of A6 would get down converted to an int index of 0,5 - which should be the 0 (first) column, and the 5th entry down…. a red piece. I did notice that my rendering function seems to be off somehow… instead of scanning through the columns adding up the values and converting them to a string for output, and then moving down through each row… the function scans through each row and then the columns….. and displays the correct results. The only way that seems possible is if the data itself is somehow turned 90 degrees? Am I missing something with how the data is being set up? Shouldn't the above statement fill the array from left to right, and then from top to bottom? Or for some reason, does that statement fill in the first column and the rest of the columns?

Thanks for any insight! Hopefully I can get it fixed tonight.

Take care all and good luck!

PS My description of the issue is probably horrible, but something seems to be happening with the way the data is stored in the array.

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Last edited on 30 Aug 2016 07:27 by PiscesMike Show more

Re: Last ditch effort

PiscesMike 29 Aug 2016 20:03

If it helps, this is my rendering function. Taking a look at it now, I may be mistaken about how it progresses. Either way, this function *does* work, in that it displays things correctly..

(That being, it goes through each entry, one at a time from left to right to build a display string, displays that, and then moves down to the next line)

using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;

namespace Checkers
{
    class R2_Console
    {

        public void Draw(Board board)
        {
            Console.WriteLine("        A  B  C  D  E  F  G  H");

            for (int column = 0; column <= 7; column++)
            {
                int zeroBasedOffset = column + 1;                               // Need to shift the display text so it starts at 1 instead of zero
                string sTemp = "     " + zeroBasedOffset.ToString() + "  ";     // Create a string to hold our display stuff, since this is the beginning of                                                                                
                                                                                // a new display line, we include some spacing stuff and grid locations
                for (int row = 0; row <= 7; row++)                              // Construct a line of text based on the board data
                {
                    if (board.BoardData[column, row] == PieceType.Black) sTemp += "B  ";
                    if (board.BoardData[column, row] == PieceType.BlackKing) sTemp += "Bk ";
                    if (board.BoardData[column, row] == PieceType.Red) sTemp += "R  ";
                    if (board.BoardData[column, row] == PieceType.RedKing) sTemp += "Rk ";

                    if (board.BoardData[column, row] == PieceType.Invalid || board.BoardData[column, row] == PieceType.None) sTemp += "   ";
                }

                Console.WriteLine(sTemp);                       // Write the line of text to the console
            }
        }
    }
}

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Last edited on 29 Aug 2016 20:05 by PiscesMike Show more

Re: Last ditch effort

rbwhitaker 30 Aug 2016 02:09

I've always considered the first index in a multi-directional array to be the row, not the column, as you've done. Either one can work, as long as you're consistent.

I'm following the conventions of matrices, where people will talk about a matrix M, and then do M_2,3 to mean row 2 and column 3. The row is specified first. (Though to be fair, in the math world, when people talk about row 2, it is actually the 2nd row, not the third, because what self-respecting mathematician uses 0-based indexing?! So doing it "the math way" is probably not a completely fair comparison in the first place.)

Anyway, as I said, you can treat it as row-first indexing or column-first indexing. What you can't do, and where I think you're getting tied up is with mixing them up. Your code to populate the 2D array is filling it in, row by row. Your first series of numbers has black and invalid cells, which implies going across the top of the board. But elsewhere, all of your code seems to assume that it is column first, rather than row first.

I think you'll need to straighten that out, either by changing the way you populate the initial board so that it does things column by column, or change the rest of your code so that it does row first, column second. (I'd switch everything else to row first, and leave the initial state generation alone, but that's a personal preference for a row-first matrix, and based on what you've said so far, probably the harder of the two.)

Re: Last ditch effort

PiscesMike 30 Aug 2016 07:22

Thanks RBWhitaker! Invaluable advice always.

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Re: Last ditch effort (update)

PiscesMike 30 Aug 2016 08:21

Well, you hit that nail right on the head RBWhitaker. Changed the order of the indexes in three functions (fixing the functions instead of changing the data per your suggestion), and it worked on the first try! Thanks a million!

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Re: Last ditch effort (update)

rbwhitaker 30 Aug 2016 12:12

Awesome! Glad I could help. It looks like I missed you in the chat room this morning by only about 15 minutes…

Re: Last ditch effort (update)

PiscesMike 31 Aug 2016 07:32

Isn't that always the way!?

Yeah, I was thinking you were going to pop in and I was trying to wait it out, but alas the chores were calling. And, still are! At any rate, I'm going to do chores first tonight, and dev later. Maybe even just hold out till Wednesday when I wake up for that. It would give me the benefit of being well rested, having a clear mind, and having most of the day to work on the game. Except of course for a lot of house stuff I have to get done now that we're going to be having a live-in guest for a bit.

Anyways, I get to work on the logic module today! It already supports (kind of) two players, in that any move that doesn't land on an invalid square it will perform. So, you can take turns as two players if you like, but there's nothing keeping you from just moving black or red every turn if you like. So, tonight or today that is, I'm going to try and make it detect actual 'checkers-legal' moves and enforce those… including making you take all the jumps available. If I get that in, and make it actually remove piece that have been 'jumped' I should be in good shape going into the end of the competition.

Take care all, and see you soon!

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Re: Last ditch effort 'Final' update

PiscesMike 06 Sep 2016 10:27

Hey all! Thought I'd throw up my 'final' update. Since I'm not finished, it really isn't final, but more of an end-of-period update. Seems like there's a couple other people still continuing work, so I guess I'm not alone in that aspect.

So. How did things go? As stated above, the original competition plans kind of fell through and I attempted a last minute, last ditch effort. Still didn't get finished with that, but did make some decent progress, and learned a couple things. As it stands right now, I'm actually only about one method away from finishing. Kind of got caught up on the logic module.

So, for my checkers program: First I learned about sparse matrix, and when it's inappropriate to use them. (Basically the only time to use a sparse matrix is with a huge array.) What is a sparse matrix? It's basically a way to represent a few objects when you have a large matrix, and it's more efficient to use instead of a matrix that might be highly populated. Turns out in my case, the array I needed is better iterated through than using a sparse matrix. So, lesson learned.

Also, I started picking up on some more object oriented programming techniques, which apparently is something I still need to concentrate on. Found a couple spots in the game where I had code or methods mixed in with stuff where it didn't really belong. For instance, I have some methods in my logic routine that take care of formatting some text data for figuring out the moves. Turns out, that should be in the Keyboard Controller interface, and doesn't sit well in the Logic Module. So even though I spent some time figuring out the design, there were and are, areas for improvement.

I also learned about 'pure' methods. I have a bad habit of making most of my variables public in nature, and that's not a good way to do things. So, I've been concentrating on making sure the correct parameters are passed into methods that need them, and the return data fits the needs of the method being called. The idea of 'pure' methods means that the method doesn't rely on class wide scope variables. It just accepts whatever data it is supposed to work on, and returns the results.

Going along those same lines, I learned a little more about the general structure of variable in classes, and how they should be protected from outside interference. This involves a little bit of extra code, as you must set up a method to return a read only copy of the data for outside use if needed, and structure the class itself to perform any logic on that data instead of allowing outside code to do so. This is an area where I need to do a little more learning, and this project was a great time to do that. Still need to update some things, but we're off to a good start.

Another thing I learned is that a simple checkers game is sometimes not-so-simple. There turned out to be quite a few complex rules for figuring out valid moves and such. That being said, after getting the logic module done, I realized it's basically backwards. I started with a bool array to represent the valid moves which sets all the default values to false, and then scans to set valid moves to true. Turns out, determining a false result is much simpler than doing all the calculations to determine a positive result. Therefore, creating a bool array that uses a true-first default value would have been much simpler. Once I finish the current logic module I'm going to redo it to function with the true-first array. At least this way, I can compare and contrast the two way of doing things.

Anyways, got some company so I should run. Take care all!

---

"May the mercy of His Divine Shadow fall upon you." - Stanley H. Tweedle, Security Guard class IV, The League of 20,000 planets

Last edited on 06 Sep 2016 10:50 by PiscesMike Show more

Re: Last ditch effort 'Final' update

Brett208 09 Sep 2016 14:36

PiscesMike,

It sounds like checkers turned into a productive project. I haven't heard of the term pure methods before, but what you are describing makes sense. I tend to encapsulate my classes fairly well, but I do rely on a lot of private class variables. It can be a chore to debug though when a lot of methods are manipulating the class level variables.

I'm interested to hear how work continues on the project.