Worklog for Sauer
Crystal Conquest
Return to Worklogs
| ||
| I am sad to announce that this project is cancelled. The reason I am forced to do this is because of bugs in BlitzPlus regarding the graphics drivers. Initially, the game was running slowly so I decided to implement the OpenGL driver. Even though the performance was more than enough, there is an annoying problem where the program crashes after I exit it. I was then directed to a DX9 lib, but this did not function properly. My next route was to devise a way to get the DirectDraw driver to run fast enough. My solution was fullscreen mode with scaling for widescreen monitors (see last post). This worked fine until I wrote a function that needed transparency using MaskImage, which did not work correctly in the DirectDraw driver. It works fine in the integrated driver and OGL, but, as we established, that doesn't work. Now, I could change my button drawing routine to work with solid squares instead of nicely bordered, resizable boxes, but I don't feel like I should have to settle for less to fit B+'s limitations, especially when a routine like this works in the other two available drivers. I then took some advice and tried to make the game run in a B+ canvas. Still crashes on close, and has horrible performance. I am so fed up with this. I have spent a lot of time and energy on this game and was making some great progress. At least I can say canceling it was not on my accord. I will spend a little bit looking for graphics bottlenecks in hopes I can make it run on the integrated card. If I don't find anything, and BRL doesn't step in to fix these bugs, I'll will be saying goodbye to Crystal Conquest, as well as BlitzBasic as a whole. -Sauer |
| ||
|
Yesterday I added widescreen support. Basically the user can decide
how much they want the screen stretched. If there is no stretching, it
puts black bars around the edges. I did it simply messing around with
Origin and Viewport. Its not the greatest but it'll do. If I find an easy to implement way to add better support, I'll do so, but for now I'm pleased with the result. The one downside is that it messed up my button rendering function. I half fixed it but there's a nasty bug with what looks to be transparency issues. Still sorting that one out. -Sauer |
| ||
|
I've switched onto my other project lately, so not much progress here.
This is healthy though, and seeing as I've only invested my time in two
projects for the past year and a half means I must be doing something
right. But today I was thinking about Crystal Conquest a little, and came to the realization that it is nearly done. When I decide to come back to it, it should be in its final stages. Something concerns me though: I'm not entirely sure this game is fun to play. Sounds a bit silly, being this far into it and wondering this, but I guess I never even really thought about it. I had a vision of what I wanted the game to be, made it, and here we are. Personally, I think its a bit suspenseful and certainly challenging, but its definitely difficult to judge your own creations. I guess we'll find out. If its not fun, its not fun; there was no investment lost or dream shattered. But it will be finished, it will be polished, and it will be a professional production. After all, that is what I set out to do. -Sauer |
| ||
|
I finished the shop where you can buy items to help you solve the
puzzles. There are currently 6 items, as I decided that having an item
to convert to every color tile would be redundant and offer little to
the solving of the puzzle. I added the credits aspect to the game, and now I just have to do the end of level resolution screen next to calculate how many credits the player gets from solving the puzzle. I did run into a problem with one of the items, the Neutralizer, which converts the block value to 0. The problem is, because I'm using arrays, 0 is registered as no-pick. The solutions are simple; I can think of another item or go into the architecture of the picks and switch the "no pick" number to -10 or something. I haven't quite decided yet, but what I do know is that it would be really nice to have dynamic arrays like in python, of a C++ vector. I could have used a linked list I guess... Oh well, I'm going outside now. -Sauer |
| ||
|
Today I took some time to edit the map screen so you can choose your
stage. I also added level descriptions that are comprised of a brief
paragraph about the planet and the time/goal numbers. I still have to go through and pick specific values for the time and goals, but I'll do that tomorrow. I also want to combine some elements of levels with others, like having null blocks on stage 3 of the fog level or have the colorless level have a closing in effect. That's all game design stuff, which I'll do next time. -Sauer |
| ||
|
Today was one of those times. I had been looking at optimization
techniques, frustrated with my fog level slowing to a crawl because of
the tile rendering. I had narrowed it down this far, but couldn't find a
viable solution. Obviously I had my memory intensive method, but I wanted to render the tiles like I was. So I started looking into the render code, which fortunately is only in 2-3 places (once for the main and the other two in "special case" functions) What I saw horrified me. My code was littered with "*32" arithmetic commands! What the heck was I thinking? In my other game, I bit shifted like a madman to avoid optimization problems, but in this project (most of which was written around the same time) I did nothing of the sort. Why this is I cannot say, but I went through and fixed all the multiplication in the rendering as well as throughout the code where I deemed it unnecessary or could go with a power of 2 and bit shift. Still a problem with a rendering. Puzzling at first, but then I looked a bit closer; I know Blitz is supposed to have a fast 2D graphics, and 9 tiles on a screen (even including lookup) should not cause a problem. Then I looked at my previous statement: no, 9 SPRITES draw quickly. Is my tile rendering spritebased? Yeah. Sort of. Half way. The colored tiles are sprites, but the numbers are written on the fly using the text command. Facepalm. So I quickly generated a list of prerendered tiles, threw them in, and imagine that. Hums at a cheery 60 FPS, particle system and all, no problem. I can't believe I didn't spot it before, but I'm proud of myself for figuring it out using a systematic approach. After a day of "what was I thinking when I did this" and many facepalms, the level looks beautiful and runs great. Can't feel too bad about that. -Sauer |
| ||
| Thought I would go ahead and make a new entry; doesn't hurt anything. I fixed the bottleneck. Not entirely sure why THAT was causing the slowdown, especially since it always happened after a minute or two of runtime, but it doesn't matter. What I did was change the item selection from clicking on buttons to simply pressing down the mouse wheel. I figure its easier on the users, to make faster picks and using items quicker. Also, its less ugly clutter at the bottom of the screen, and less stuff a artist would have to do. I also went through and looked for for loops that could be broken out of, and found a few. If I just keep it in my head to be vigilant for optimizations by the time I'm complete I should have a lot of them worked out. In all reality, the code is not massive by any standard (just over 1000 lines) so it shouldn't be too difficult. There's still a slowdown on the fog level, and I think this is because of how I have to access the array, analyze the contents, and draw the contents nine times per frame (one for every tile). There has to be a better way. One idea I had was to make a 320x320 sprite that lays on top of the grid, and the different frames are the different "windows" that you can see through depending on where the mouse is on the grid. This would probably be pretty fast, but take a ton of memory. I like to try to rationalize and say that the game currently takes little memory and computers are so powerful, but I still can't bring myself to cope with a 100 frame, 320x320 sprite. Its insanity. So I must be creative. These are the kinds of things I like to think about before I go to bed though. I do my best thinking in the shower, so maybe tomorrow morning I'll come up with something brilliant. The good news is that the last level I have to do (the closing in level) will be really easy and not computationally expensive, and after that I can move on. I'm getting more and more inspired to finish the game as I get closer and closer. I'm a little curious as to what the reception will be. I must be sure to go above and beyond so I don't get railed when I decide to release the alpha. EDIT: Thought I would go ahead and post an updated in game shot, without the ugly buttons or tiles. This is of planet 2, Arok, and those bluish purple things are the null blocks.  -Sauer |
| ||
|
I was working on the fog level, but am currently running into a huge
FPS loss. Trying to think of creative ways to overcome this. The
closing in level will be easy so I'm going to take a lot of time to get
this level working properly without cutting corners. EDIT: Alright well I found out that the FPS decrease has nothing to do with the fog level code; its actually some bottleneck in the code. After letting the game run for a minute or two it slows to about 29 FPS, which is about half of what it is supposed to run at. First I checked to see if it was in the picking mechanism and redrawing the screen (which is very computationally expensive in relation to the rest of the game). I got two major optimizations in, one that breaks a for loop when deemed unnecessary and the other a function that only redraws the tiles that changed, not the entire grid. Still slows to a crawl. Now I'm going through and commenting out just about every function to isolate the problem. More to come. EDIT 2: After commenting out a good portion of my code, I found the problem to be in the item selection. I don't know why it has the effect of slowing it down after running for a few minutes, but now it doesn't slow down anymore. Now I have to implement a new item selection function. I don't mind too much because I never liked how the old one looked, and this gives me an excuse to fix it. I expect small slow downs throughout the game, but as long as it stays over 45 FPS it will be fine. The code is probably getting to a point where it could use some tidying, and I'll get to that ASAP. -Sauer |
| ||
|
Today I redid my little particle system that is used every time a
combination of 11 is found. I had the traditional "throwing particles
around" style emitter but to make it look good, it was bogging down my
frame rate. Also, I'm no expert on particles anyway so it didn't look
that great. I replaced it with a simple sprite system that shows an explosion. This is simple yet effective, and any artist would be able to go in and make a better suited sprite. I also added the "dummy" blocks for level 2, and the 'dead' world of level 5. This leaves shape shifting, fog, and closing in levels. For the closing in levels I'll probably convert them into dummy blocks like in level 2. Today I took the time to get some images uploaded. I went through and found all the images I got from google and found all of their sources. Keep in mind that these are probably all copyrighted, so I cannot use them for the final game without the author's permission. I more or less just wanted everyone to see what effect I have in mind. Title screen:  And a map screen, showing the shop (the space station). This screen will have level details over the planet, save button, and back to main menu button somwhere too.  Thanks for reading. -Sauer |
| ||
|
I went ahead and made credits, instructions and story pages for the
main menu. Then, I took the map screen and added smooth transitioning
from level to level. I also added a little space station that will
serve as the shop. Next up is to change the data structures to match the 6 planet, three stage per planet style. Logically, after that would be the displaying this information on the map screen, and then reflecting it to the main loop. Also, as far as game design goes, I'm considering making a "challenge" mode which is just a jumble of strange puzzle combinations. I think this would be good, as the scenario portion of the game will be quite small. Again, I'm still sticking to the ultimate goal of having an artist come in and replace nearly all of the images and make a high quality finished product with lots of polish. -Sauer |
| ||
|
Tonight I went through the game and cleaned up the code a bit. I was
seeing some inconsistency when it came to the way variables were being
treated. I found my code to be well commented. I redid the graphics, again, and this time, its it. No more graphics for me. I stated at the beginning of this project that I wanted them to be done by an outside source, so I'm going to treat them as placeholders and just get on with the project. Tonight I wrote a little button function which will be the main source of the program's UI. I also made a main menu with some comments depicting a 'road map' of the pages I need. I'm going to continue cleaning up the code and getting rid of some redundancies. Also, I need to change the level data structure, as I want to implement three stages per planet, for a total of 18 stages. The difference in the stages will be the goal number, goal score, and goal time to create different levels of difficulty. Each planet will have to have a certain special feature to the puzzle. Here are them listed briefly: 1. No special feature, just the first level so nothing fancy (earth like planet) 2. A 'block' that has no value and cannot be elimated (ice planet) 3. A shape shifting level where some values change every now and then (red, rocky, mars like planet) 4. "Fog" feature where you have a limited sight around the mouse (jupiter like planet) 5. No color bonuses (dead, gray planet) 6. Level that slowly closes in on itself (lava planet) There. This post has turned into a bit of a design document. Lets keep going shall we? Save/Load will just save progress in the game, what levels have been beaten, what items you have, time logged, etc. I may even do a high score thing for fastest solutions Story will just be a couple of static picture story boards, which I'll find some placeholders for. Nothing big there. The shop will have numerous items. You buy them with gold you get from beating older levels at fast times; the harder the level the more gold. They are: -Block Bomb: creates a block of ones of 1 color -Randomizer: makes a new grid -Machine Coolant : gives you an extra 5 seconds -Jackhammer: breaks the ice blocks in level 2 -Flare: breaks the fog in level 4 -ColorConvert(Red): converts a block to red, one for each color -Zero Block: converts the block value to 0 -others as I think, but thats a good base. Blarg I'm tired, can't wait to work on this a little more tomorrow. -Sauer |
| ||
|
Created a video and a youtube account; here's a basic demo. Shows a
rudimentary title screen, level select screen, and some of the items. Comments are welcome on youtube or via email: http://www.youtube.com/watch?v=c5_8GaVh30g -Sauer |
| ||
| Yep, had a strange urge to do them today, and redid both the tiles and the backgrounds. I'm not entirely done with graphics... I still need to do effects for menu's, transitions, and the 'specials' on each level. For now though, I'm happy with what I have and can move forth designing the levels and putting the game together. Here's a screenshot; if you feel compelled to give feedback, it would be greatly appreciated. My email is in my signature.  -Sauer |
| ||
| I'm making little progress with this game. Still have to do the graphics. I'm making no progress anywhere when it comes to programming, as my job and other interests are getting in the way of quality programming time. Next chance I get to sit down, I want to knock up a nice screenshot of what I'm doing... I may even use some nice marble textures for the tiles as well. -Sauer |
| ||
|
I've decided to dump the majority of the graphics I got from the
graphics guy. Although they looked nice, there was an inconsistency
with planets and backgrounds. After re-looking at a post in the forums, I downloaded the Genetixa 3 Viewer (or something like that) from spiralgraphics.biz and found they have some nice textures (as well as some really bad ones) and also nice planet images. With the viewer, I can resize and change the color of them, as well as edit some of the procedural aspects. What does this mean? Well I'm set back a bit because I have to redo graphics. I've also forgotten all the special things for each level except for 1 from the previous post. Even though its a slight setback, I still am not bored with this project (something that rarely happens with me) and it is going along nicely with my other project. I've switched my focus mostly to the other project, but what I am doing is rotating projects to keep my ideas fresh and prevent burnout. That is all! -Sauer |
| ||
|
I've been talking to my graphics guy about game design and we came up
with a couple of items to make the game more interesting but not
overbearing. First there are six levels, each with three sub levels. The difference in the sub levels is the number to get increases per level, and the time or goal may be adjusted. This makes a difficulty gradient in the three sub levels along with the six planets. Second, each planet has a special challenge based on the nature of that planet. They are to add challenge and variation to the game without overloading it. with gaudy features. Small changes are some items need revision, and upon every match a little time is added to the clock. As far as coding goes I've implemented the map screen with little difficulty and changed some graphics. Next up is implementing the three zones in each planet. Then its back to the list set up before. -Sauer |
| ||
|
I went ahead and added those images my buddy made me, and they look
great. I believe he got a lot of them from some other online space
game, but I'm not too picky and I'll definitely give credit, and if
somehow it takes off I'll get proper permission. Least of my worries
right now though. I've been so busy with school lately I haven't done much on this game. I've been working on DOF2 a little but not as much as I would like to. Next week is spring break though, so I'll have plenty of time to get some work done. I think the next step is to get some alpha testing done and get some feedback on gameplay enhancements. I don't want to add too many features (as the game is pretty fun as it is) but I would like a couple more things to spice it up. Then after I get some ideas I'll go into level design. Thats a status update, my time is limited and split between two games, but I'm no where near burnt out or finished with Crystal Conquest. -Sauer |
| ||
| I have a buddy that has photoshop and claims to know how to use it well, so I assigned him the task of making the map screen. I had already made a map screen, but then when tweaking it ruined it. Instead of doing it again, I found someone who could probably do a better job and had better software. If he does alright on this job I may ask him to do other backgrounds or screens. -Sauer |
| ||
|
So now I changed my mind about changing the theme. I really hate the
map screen and the theme I changed it too. I can't really describe what
makes it different but I absolutely can't stand it. Also, when looking at my project goals from the first post, the goal is to make something of the highest quality of the casual games industry, with no specification on the sales of the game. I think that changing to another theme was in attempt to ensure better sales, but that's not what this project is all about is it? Furthermore, as life gets more expensive the idea of hiring an artist becomes less and less a possibility. Well, let me rephrase that... the idea of hiring a really good artist. There are always people that will work for cheap or royalties, but just don't produce the best work. I'll probably have to take that route, so I can have at least an improvement on graphics, but not necessarily sell worthy graphics. Again, because sales are not a priority, mostly just proof that a quality product is obtainable. I'm having fun finding pictures of planets, stars, nebulas, and comets and implementing them into my game... and I'm always saying how fun is the important part. So space theme it stays! -Sauer |
| ||
|
After reading some articles online, as well as some input from the
members of this forum, I've decided to change the theme of Crystal
Conquest from space to 'explorer' or 'adventure' theme. It seems that
the space theme does not sit well with the portals. I just started working on the map screen, which is the only thing that is really affected. Since the game was going to take place on each planet, I can just say its on different parts of the world. It fits in to. I plan on changing the story to something like this: The world's power reactors are on the fritz! Powered by the power crystals, the reactors need to be unjammed so energy can be restored to the planet. Travel the world to unjam each of the (x) elemental crystals and save humanity! Why something like this would sit better with the casual crowd is beyond me, but that's what I'm hearing. I'm currently working on this game and another application, so I get a good mix and don't get burnt out. When I get frustrated with a particular aspect not working, I just switch gears and feel better. I'm off to go work on my new map screen, which will look like some sort of global map. Maybe I'll use Earth? -Sauer |
| ||
|
Just after posting that I wouldn't have enough time to work on this
project, I had one of my more productive days of late. I basically
finished items, the only thing left to do is create more of them. The
mechanics for how items are selected and used has been completed. I noticed after playing the game a few times that you are allowed to select diagonally. At first I thought I should hurry up and fix it but after a couple plays it really opens up more opportunities for matches and forces the user to plan moves by observing all of the possibilities. I like it! Another 'mistake' that worked is that I realized you are able to hold down the mouse button and drag across to pick numbers. This makes play more fast paced and makes larger selections easier. Everything is going swimmingly except I noticed my game runs pretty slow... a whopping 25 FPS. I'm not sure if this is acceptable (the game plays fine for the most part) but I definitely need to do some streamlining to get that up a bit. I ditched my junky particle system and that helped a bit but something else is slowing it down... So far the code clocks in at 635 lines, and has been in production for 22 days. I'm pretty happy so far with my productivity. So next item (along with streamlining) is creating the map screen. -Sauer |
| ||
| Been really busy with school, so work on this project has been designated to weekends. Obviously less productivity is expected. -Sauer |
| ||
|
Yes step by step is definitely the way I am taking this project,
inching along piece by piece and trying to make everything as flexible
as possible for the artist and myself. As I have never worked with an
artist before I don't know what to expect, but am trying my best to make
it as easy as possible for him/her. Well I finished designing the HUD (or at least a basic version of it) as well as began working on the item mechanism. I got four basic items and their functions all worked out, but have done little more than tested how they will display on the screen. In the following days I will continue working on the items. I have pretty much laid out a general outline of what needs to be done. This is strictly gameplay wise, having nothing to do with final processing (resolutions, compatibility, testing etc.) 1. Items 2. Map Screen 2.5 Mini Games, Shop, Story Development 3. Title Screen 4. Effects and loose ends Granted this is pretty general but I believe that having a plan, no matter how broad, will be useful. Also, to specifically write down everything that needs to be done would be difficult and wasteful, as casual game production is extremely fine tuned and sometimes ideas spawn from simple testing. Well enough of my blabbering, here's a screenshot. Keep in mind that my art is awful and will all be replaced. This can however, give a general idea of what the game will look like. Comments and feedback are appreciated (as if anyone was reading this thing anyway) and my e-mail is in my profile. EDIT: IMAGE REMOVED. -Sauer |
| ||
|
Today I started designing my HUD for my game, and my idea was a
"machine" that wraps the bottom and right side of the grid. The bottom
would be for displaying picked items, and the right side would have two
bars, one displaying time limit and the other displaying current score.
Because the game revolves around 'powering crystals' the bars should lead to a crystal, and when full, the level is complete. The problem is that I don't know how to create proper place holders and implement the bars, which should be flexible because the artist will draw both the machine and the bars. I guess in that sense it doesn't really matter what I draw... I'm just going to have to re-program it anyway. I also need to begin designing the map screen as well as an improved title screen and menu. -Sauer |
| ||
|
I didn't have a lot of time to work on the game today, but I did find
some time after work to find a nice freeware font on the internet and
find a title screen background. The font is sci-fi but easy to read. I may use it in the final product if everything blends well, but everything graphically is still officially temporary. The background for the title screen is actually the command deck of Starship Enterprise from StarTrek, so that is definitely temporary, but I do want the command deck of a ship as the title screen, and so far I like the atmosphere it gives. As far as gameplay goes I noticed that as the game progresses it gets extremely difficult to make matches, and the reason was that the grid never reset per level, meaning the grid was parched of potential matches. I added a function to reset the grid values so this was taken care of. -Sauer |
| ||
|
I had written a prototype of the game using simple text based GUI and
display to see if the design was up to speed. After taking some advice
from the forums, I believe the game now has a sufficient design plan to
begin creating a version with placeholder graphics. Today I spent some time commenting and reorganizing my code. I feel that I work more efficiently when I mark off sections of my code and label all functions, so I took the time to do that. I also when online to find some landscape images of each of the eight planets, for backgrounds. My goal is to find/create placeholder graphics that are as similar as possible to what I want the final product to look like, that way it'll be easier on the artist when hired. So far it appears that the main mechanics of the game are running smoothly, which is a good sign. -Sauer |
| ||
|
Description: A casual game where you match adjacent tiles to add to
eleven. If the tiles are of the same color, then a bonus is added to
the score. Each level has a score goal to reach in a set amount of
time; both vary from level to level. The game also includes items to
use during play which can alter the time limit, goal, and numbers on the
board. Items can be 'purchased' with credits earned during play, or
won by completing certain achievements. Goals for Project: Create a product worthy of portals or self publication. Product should meet all of the highest quality standards of the casual game industry. Whether the game sells well is not a priority, but simply the practice and pride of creating a noteworthy game. Notes: Game should be fairly short, yet provide significant challenge toward the end of the game. Graphics should be crisp and professional, done by an outside source. Story is interesting, but not incredibly deep. Story: Many light-years away lies a solar system of eight planets that orbit a single sun. As the sun grew older, it became apparent that its rate of growth must be controlled to avoid supernova. To do this, StarFleet Command harvested the energy of the power crystals found on every planet with specially made reactors. With all eight reactors working at full capacity, the sun could be controlled and it's longevity increased. But trouble strikes when the reactors are jammed by a mysterious figure, holding the planets hostage to their own sun. It is up to you, a StarFleet cadet, to visit each of the eight planets to restore the crystal reactors and save the solar system. -Sauer |