And it is Done.

Blocku is done. I have built and packaged the final release and submitted it to the sugarlabs activities page. It feels good to be done. I have accomplished all my goals from the tentative schedule from the beginning of the quarter, and have managed to add some extras into it. As such, this project is complete as I see it, and it is also the end of the quarter, which means the end of my co-op. What this means is that this project will no longer be under active development. I will update the wiki with the current status, and will most likely stop updating this blog for a very long time.

And with that, here’s me signing off.

http://wiki.sugarlabs.org/go/Activities/Blocku

http://activities.sugarlabs.org/en-US/sugar/addon/4399

http://gitorious.org/blocku/

Seriously almost done

From last time when I thought I was almost done, the only thing left to do was to build an xo release. I have done so, and discovered that there’s countless glitches (as expected), but what’s even worse is that the glitches were specific to Linux. So since then I’ve been working on Blocku only in ubunutu, hopefully to eliminate any other OS specific glitches. Since then, I’ve had to do major work on the file saving on loading format, as everything there was broken. After that, I realized I had accidentally left an entire feature unfinished. Time Attack mode, when racing against the time, never actually stops you and says “Out of Time” when the timer reaches zero. That’s fixed now. Lastly, the part that I feel truly finished the game, is localization. I have now converted every string in the game to a dictionary object that reads in the appropriate strings from a text file with whatever language the user selects. English is done, and I can do German as well. I have found a spanish translator, but still need a french translator. Hmm, where to find you.

http://gitorious.org/blocku/blocku/commit/408d2d596b3908b076a3a7d4f5ee426b9f07f474

Almost done

As week 8 is coming to a close, so is Blocku. The game is almost complete. There are only a few finishing pieces left and some overall polish to be applied. For specifics, audio needs to be included, as well as a slightly better visual cue system as to what is happening. Most notably, highlighted buttons and blocks. Apart from that, I need to figure out if i can have it so that the current block being dragged gets put on top of the other blocks. Oh, there’s also localization that needs to be completed.

Apart from work that needs to be completed, Custom Game mode is finally finished, including the ability to create, save, and delete custom boards. All custom boards come with their own list of recent completion times.

http://gitorious.org/blocku/blocku/commit/5faaf19f5b8e65d152ee16b7f2eab417a215226b

In other news, a developer release has been published on the sugarlabs activities page http://activities.sugarlabs.org/en-US/sugar/addon/4399 . This is a pre-release developer build with not everything working to its full extent. It already has 20 downloads, which is 20 more than I expected. But maybe I’m just being pessimistic. Blocku should see a final release on February 24th.

Overdue post

This week has been a week of bug fixes. Most notably are two fixes that have existed since the beginning. The first is that the mouse cursor is now transparent where it should be. HOORAY. this has been bugging me and everyone else that tested Blocku forever. Mouse transparency. The best part about it was the fix too. From every forum post or whatnot that I read about getting image transparency, more and more broke. They all required several methods and other whatnots. I eventually just decided to look at the documentation of loading in image files, and I saw that there was a convert_alpha as opposed to a simple convert() that also retained alpha channels. As simple as that…

Secondly was the fact that random sprites just decided to reorder themselves on the Z axis every time the game restarted. Sometime you’d get These blocks on top, other times it would be those blocks. This also included the Win text, objective text and the mouse cursor. So the cursor would disappear behind certain blocks and in front of others. But it was never consistent which ones, and how many blocks it would be in front of. The fix was a simple matter of using a layered update instead of a batch render. Why are the fixes always the simplest things?

Aside from that, I’ve started implementing a pseudo-scoring system. Puzzle mode has a target number of rotations to shoot for, and this is displayed at the side. Time Attack mode will display the 5 most recent times set by the player(s) and record these in a text file that it reads from and writes to. Upon puzzle completion, the game asks for your name in an input field in a pop-up.

http://gitorious.org/blocku/blocku/commit/b42327f25451b72dd263070127dce66b30251347

Create a board

Create a custom board is now fully integrated and working. It may not be the most polished create a custom board in existence, but it works, and it’s there. They way it works is in the Story Mode, there is a button to take you to the create a board screen. It’s  simply a really large board where all blocks have 0’s on all sides. Click a block to edit it. Numbers go up to 6 digits should you ever need to go that high. If a block has all 0’s, then it won’t be in the final board. The user then clicks the save button and enters the board name, and the objective text. The board is then saved in the proper format for reading in later on.

I decided to make the create a board where the user is presented with a full board to begin with, because the other method – where the user has an empty board and hits add a block – would still require a snap to grid for smooth board creation. Now the way the grid is programmed, is that it gets generated after the blocks do. So this way was a lot easier, and it also serves a design purpose. When presented with a full board, the user can instantly see the extent of the maximum size of the board possible.

http://gitorious.org/blocku/blocku/commit/0a56e0393ee9331e8c5322aa1c22b026fa87df23

STORY MODE!

This has been trouble. Story mode, or loading in custom created boards, has been an exceptional amount of trouble to get working. One would think if the text files were already written properly, and there’s already algorithms in place to read them properly, that it would go smoothly from there. No. From my experience, File IO has always been the trickiest thing to get working. In this case, the biggest problem was cross-platform compatibility. I’m developing on a windows machine, and certain things work differently from Linux. Most notably the path separator. In windows, the separator is \, but Linux its the other way around /. I had forgotten this in the beginning, so my check to see if the file exists always returned false in Linux, therefore I spent numerous hours looking for how to do the file checks in Linux. This turned into multiple rewrites of the code, all of which failed. I eventually happened upon the answer when I saw that new files were created in my Linux VM that were called boards\testFile.txt. That’s not right. This is when I remembered the different path separators, and after some research found out I could have simply used os.sep to go into the boards folder and then test the file. Next up, finish up the level creator.

http://gitorious.org/blocku/blocku/commit/7f9825d6b6f8969dd40965310de1c9dea84a1c9b

TV Crew

Yesterday, WXXI TV crew showed up at the FOSS Box. I was summoned to answer some Blocku related questions, which ultimately turned into me getting my five minutes of fame on television. I did a quick demo of blocku, and talked about the built in features and game modes available to the players.

As for the development itself, Story Mode is slowly coming along. I have finally figured out how to implement a text box, something that I had to program for the first time myself. Basically it’s a display rect with a pygame font that changes in a while loop and is assigned to and assigns a string to keep track of it. The concept may have been simple, but it was hard enough to create. Tomorrow should see an actual use of the textbox for loading in a user specified custom board, and ultimately, the textbox will be reused for name entry in high scores and such.

http://gitorious.org/blocku/blocku/commit/352e9c4acf1a502d376cbfe8ffd127365c28233e

End of week 4

End of week 4. Puzzle mode has been completed. How this works, is that the grid is already completed with the blocks, but the block rotation isn’t. What the player then has to do, is to rotate each individual block until the numbers satisfy the equation again. The timer at the top is gone, and has been replaced by a rotation counter. It counts how many times the player has rotated a block. This will eventually be used to determine scores, as the game also saves the number of times it rotated itself to achieve the scrambled state.

Overall, puzzle mode was relatively simple to integrate, as all the code was already there. All this required was a rework of the scoring system, and a couple of checks of which game mode was selected. There was the possibility of still being able to drag blocks in puzzle mode, something that shouldn’t work, but that was also fixed with the addition of a single if statement. All the options are also the same. That is, being able to choose difficulty, number range, and addition or subtraction. So not much had to be changed there either. There was a small tweak, as puzzle mode only has two difficulties compared to four. Easy mode is a 3×3 grid, and medium is a 4×4 grid.

Something I should have talked about sooner, but over break I was also play testing blocku with a group of friends, and their younger siblings. One of them was even the target audience, so that worked out nicely. Play testing  was simply “Time how long it takes you to complete this puzzle”. However, the results were interesting. These were all time attack. In easy mode, it took friends in the college age group between 5 and 10 minutes to complete. Medium took considerably longer, and no one managed to finish one that involved rotation. The 4th grader required 20 minutes to solve an easy puzzle. Now this game was originally targeted at the 4th grade audience, but after seeing these results, I’m not too sure about that anymore. During development I already realized that rotation makes the game exponentially harder. Maybe if this game was worked on as a group project it might still be 4th grade stuff. But apart from test results, everyone seemed to enjoy the game, and that is a good thing at least.

http://gitorious.org/blocku/blocku/commit/a2b294df78a0f190a466627e24ee430fb8f5c9e0

back from break

Its good to be back in school. Anyway, I have no added the option of selecting between addition and subtraction. This option allows the player to choose whether they want to solve the given problem by adding the adjacent numbers to the answer, or by subtracting the adjacent numbers to the answer. In the case of subtraction, order of operations is ignored and the larger number is always used first. Or you could think of it as always using absolute numbers.

Secondly, There’s also been several minor tweaks. Not necessarily bug fixes, as nothing was broken, but more like memory optimization. I managed to reduce the number of objects by reusing similar ones for different cases.

Plan for tomorrow and probably the week is to get puzzle mode implemented.

http://gitorious.org/blocku/blocku/commit/c1c38189760579e714f84c1b90b8b72bf949de2a

End of week 3

The end of week 3 is upon us. And with the end of week 3 of winter quarter, also comes the end of a decade. As we move into a brand new year, along with it comes tidings of the next great step in development. The beginning stages past prototype, and full functionality.

Enough of that. The past week has been a rough one concerning development. I started on getting Game Mode selection implemented, and I tried to do this by importing a widget package for pygame. At first I could only find the GTK widgets, and those don’t work here with pygame, but then I found pygame specific widgets. I spent a good amount of time trying to get those to work in various ways, but simply couldn’t figure it out. I then looked up just general widgets written in plain python. I found some, but couldn’t get any of them to work either. So i ultimately had to settle on writing my own. Before I began this, I redesigned how the game mode selection screen and options would look like, so that I could minimize the amount of widgets needed. This had two benefits. The first being that I wouldn’t have to code as much, and the second being a simpler UI. Always important.

Now I have Time Attack mode mostly implemented, and the user can select number generation range as well as difficulty level. There’s 3 ranges, low medium and high, and 4 difficulty settings. The difficulties are 3×3 grid no rotation, 4×4 grid no rotation, and then those two again but with rotation of each blocks. The number generation range goes as follows, but is subject to change: Low is 15-30 , Medium is 20-99, and High is 100-999. The timer for Time Attack currently counts up, and I will have to extensively playtest to figure out good starting times for each mode. I will be doing a lot of this over break, and I’m also going to continue to code over break.

http://gitorious.org/blocku/blocku/commit/2ce68c41580a2093907f110acf944b6848fe0399