Note: image not yet created


What have you been up to in October of 2006?


If you want to understand the blog and how it's laid out, go here.

Oct 31, 2006



Oct 31, 2006: I made my first animated cartoon, one that might be the world's biggest, most complex animated GIF. The high rate of recalling dreams also continues on - 10 more in just two weeks! I now have two songs that have passed the month-straight case. I otherwise haven't got much further into learning C.

#1 My first animated cartoon: My previous animations were short and tiny compared to the new animation I just did. On the main index, for a very long time, I had a note of saying of attempting to make my first cartoon and making an animation of at least 2000 frames. 2000? That's it!? Try 3634 frames. Yeah, that's right, 3634 frames. There's more to the animation than just a huge number of frames. There's much more realistic lighting than any of my other animations (limited in quality due to GIFs color limitations). There are more realistic 3D effects, clouds resembling my 2D game's first clouds, and there's even a very simple story in the form of text. You can also view the current speed and height during the animation. This might be the world's biggest, most complex animated GIF. The entire animation took a whole week to make, working at it for 80 to 90 hours! The first day, the basic plan was done, the spreadsheet. The second and third days were spent making the background scenery. Days 4, 5, and most of day 6 were spent making the foreground scenes, the new and seriously enhanced character. The last of the seventh day was compiling and finalizing the animation. Remember the "Mountain Fun" animation I made in the last update? This animation doesn't have the complex math used, but has 5 times the complexity of the last one. Rather than 17 layers, there are a monstrous 30 layers. The 30 layers comprise of 8 layers for fog, 1 city layer, 2 road layers, 3 hill layers, 4 mountain layers, and the last 8 being the clouds. There are 3634 base frames (3444 actual frames after optimization), 4 1/2 times as many frames (5 2/3 if comparing the two optimized versions). This may seem like 8 1/2 times the complexity, but the simpler math involved reduces it to more around 5. The animation really made it hard on my system using up a great deal of memory. I took a screenshot of it as I was very amazed at the memory usage needed and the number of read and write bytes (and I thought 13 GB from Winamp was high enough for read bytes, this is over 12 times that). GIMP crashed on me twice, not from the menu usage bug, but that it was just overwhelming for the program. The other was because it failed to allocate 166,000-something bytes. As a result, if GIMP didn't crash twice like it did (causing the 3 instances of script-fu as you're seeing), the read bytes should be 100 billion higher than this, page faults 18 million higher, write bytes 5 billion higher, and the CPU time another 2 or so hours longer.

GIMP uses 1,170,244 KB memory, caused 24,550,046 page faults, read 160,744,765,138 bytes, wrote 6,010,395,259 bytes, using 3 hours CPU time (against 5 1/2 hours idle time).


Believe it or not, I had this animation thought up since 2005 and back then, I only had the spreadsheet plans. I never got around to working on the animation because, at the time, it'd have taken 5 times longer. I originally just wanted an animation that demonstrated the various special abilities often depicted in my mind game. The time that was needed for the resulting animation to play was roughly the length of a Bugs Bunny cartoon. Due to the huge number of frames, lack of a BMP to GIF batch converter, even the frame-creation process otherwise taking 2 or so weeks to do for even making the scenes with just 2 background layers. It'd have also taken 12 hours to manually open each BMP file, save it as GIF, open the next BMP, and so on. I had triple-digit negative motive for doing it (in today's terms). Today, due to automating the frame-creation process, this animation was now possible not just with 2 or 3 layers, but even 60 if I could. I settled with 30 as I knew it would be pushing the limits of GIF's maximum color count of 256. I also added realistic lighting effects, very powerful 3D effects, even a speed and height stat. All this combined would've taken me 3 months using the old methods, likely even longer. Detail like this was unheard of back then and even this isn't my upper limit, although fairly close to it. The "Balloon Skydiving Without a Parachute" animation was the most complex I could stand using the old method, but it pales in comparison to my newest animations, let alone additional discoveries and new math formulas. It's the GIF file format that limits me from going any further. I have several ways of getting an animation. The table below compares each of my known methods as well as various aspects at their greatest potential (except file size).

Anim typeEase of viewingFile size*Cost to make†‡Max amount of detailMax colorsMax frame rateOther notes
GIF0.2 - ultra easy - just click a link and you're already seeing it4.08 MBFree1.7 - Low25610**GIF is the better of the two methods
AVI5.1 - easy, but very inconvienient - Download a ZIP file, unzip it, open the folder, unzip the second ZIP file, and view the AVI3.9 MB download, 1100 MB uncompressedFree8 - Unlimited16,777,21660***Although hard drives are very big nowadays and could easily cope with a 1.1 GB video, users could see it as a waste of space and the process to viewing them is very inconvienient.
MNG†0.2 - ultra easy - just click a link and you're already seeing itUnknownFree(?)8 - unlimited16,777,216UnknownThere is very little support for MNG files as very few browsers can even display them (Firefox is one).
Flash‡0.2 - ultra easy - just click a link and you're already seeing itUnknown$399Unknown16,777,216 (?)30(?)††The price needed to get the software is rediculously high, twice of what it should be and 4 times that of what I'm willing to spend.
Program and inputs2.7 - download the program (if not already downloaded), download an input file, run the animation player program, and open the downloaded file.0.6 MB‡†FreeUnlimited16,777,21660This is an even better method than GIF, but it has serious advantages.


Table footnotes:
* Based on my newest animation, which is 4,275,669 bytes. The others are only estimates.
** Although GIF supports up to 100 fps, browsers limit it to 10 fps.
*** I can't seem to use anything above 60 fps, but the file format could allow for frame rates much higher than this (I'd have to examine it - it likely uses milliseconds and is probably a 32-bit unsigned integer (unsigned because frame rate cannot go negative, unless played back in reverse) meaning nearly 4.3 million frames per second maximum).
† I don't have a way to save as MNG so I know very little about it. It is PNG's way of animated GIFs.
‡ I'm completely unfamiliar with the formats and design of flash objects, so I don't know much about it. I'm basing the details on what I do know, however, from what I've actually seen.
†† 30 fps for flash is the fastest I've seen. Due to being entirely unfamiliar with Flash's limitations, I don't know of anything higher.
‡† Not including downloading the program itself which is a one-time thing and the program may be about 200 KB or so.
†‡ Based on the cost to obtain the software to make something in this format - it is a one-time fee though.

The last item on this table may look interesting. So, where's the program? It doesn't exist yet. I first would have to learn C (as Gamestudio would be a pain to work with due to it's limitations and annoyances). Why such a small file size for the download? By compressing the files into a single file usable only by the program, and the type of detail I have in my newest animation, only one instance of the background scenery is needed to be saved. GIF, AVI, and possibly MNG all save it on a frame-by-frame basis. The scenery doesn't change it's shape or colors, just the positions are changed. This way, only one instance of the scenery is needed and this alone has a HUGE impact on the file size. By compressing this single instance, the file size is reduced even further. What about the change caused when using the flash attack? All that is is just a half-transparent all-white layer drawn on top of everything, and that would compress to about 8 to 20 bytes. It's only special foreground objects that would really take up more space, but the 170 used here would compress quite well anyway. The text would just be strings rather than images (including the on-screen data like the speed and height) would has a big impact on the file size. A spreadsheet wouldn't need to be downloaded, a simple file containing about 16 columns of detail is all that's really needed and this is further compressed. All the content is placed in a single downloadable file and you wouldn't need to unzip a ZIP file - it's already precompressed and the program would just expand it and the program would just place the objects accordingly. This method will easily allow rich detail. I may even create a program that'll allow you to create the animations in this fashion. The reader program will be free, but the writer version will not. The writer version would likely include a spreadsheet system done much in the same way I use Excel, only where you don't have to do the complex calculations manually - common things can be automated. At the moment, this is beyond my knowledge, especially the spreadsheet part, but it won't stop me from making the program once I actually get the knowledge of being able to write it. A pinball-like game then my 2D game in C would be the first step. I can already envision the interface in near-photo quality. Such a program would be one of my first non-game programs (in a sense, version 1 of my 2D game is a non-game program). I would actually love to make an animation at 60 fps - the one called "The Spinning Board" would be a great candidate for that. I took screenshots showing the complexity of the animation and what the spreadsheet actually looks like. Note how many columns and rows there are as well.

Spreadsheet for cartoon - great size complexity 1


Screenshot #1: This shows the area where the character lands on the cliff after the first of the glides and makes some puffs. The leftmost column is what the character's action is. 4 is gliding, 3 is stopping, and 1 is standing. There are 8 states. The second column is used for the direction the character faces, 1 for right and 0 for left. The third column tells what foreground object to use or 0 if none are used at all. The fourth column indicates whether or not the flash attack is used as it would just toggle the visibility of the flash layer. The fifth column is what story text to display or 0 if no text is to be displayed. Screen data (speed and height) are always shown. The eight columns to the right of this are the Y positions of the ground and those to the right for the X and Y positions for the other objects.

Spreadsheet for cartoon - great size complexity 2


Screenshot #2: Directly to the right of that in the first screenshot are where I have control over everything. The speed columns are in bold, X on the left and Y on the right. The two to the right are adjustments, used to get the speed stat to display properly. The far right is the notes column. Note my note for "cliff edge appears" as well as the "lands; starts stopping" note. Compare that to that in the first screenshot to what the first five columns show (A through E). Also note the indicator of text displays and the change shown in the spreadsheet in those first 5 columns. The 3 columns to the left of the X speed are the X position on the far left, the Y position to the right of it and the diagonal speed for the average speed, used for slopes. The empty column is a spacer to separate what I need to copy into a TXT file. This is what my program reads - the positions of the objects in the scene. The cell highlighted here is for the most distant clouds' Y position. With a scaling of 3000, they barely move, even at 200 mph. The column to the left of the separator is the height, in feet and the one to the left of that is the speed display value. Note the 110 mph top speed and the speed decreasing at 10 mph per frame while stopping and note the same thing occurring in the animation.

Spreadsheet for cartoon - great size complexity 3


Screenshot #3: Just to the right from the second screenshot is what this screenshot shows. Here you can see temporary notes and calculations. The column for "shockwave number" is used to help me gauge the size of the shockwave. The second shockwave effect and a more detailed approach which also includes opacity. Note 30 stages and that the shockwave remains for 3 seconds.

Spreadsheet for cartoon - great size complexity 4


Screenshot #4: Here's another area of the spreadsheet, near the end. After the very long glide, impacting the slope and at the area where the speed passes 999 mph, this is what the spreadsheet has. The 999 limit was indeed intentional. Note how the speed reaches 1316. That's 1316 mph as an average speed. The impact speed is actually 1320 - that's what the adjustment column is for. Note the column for the furthest clouds' Y position. See how it rapidly changes? Note the change of 0.4 where the speed is 1200. 1200/3000 is indeed 0.4. Note the area where the speed jumps to 50, but shortly after, the horizontal (X) speed increases. Can you see how the X position of the clouds change as the X speed increases? As you can see, there's a lot more to the planning and calculations. I can see the numbers and actually visualize how the end result would be without even needing to make the graphics. I can make fine adjustments as needed as well, such as the scaling for something being changed or the height of the clouds.

Spreadsheet for cartoon - great size complexity 5


Screenshot #5: To the far left from screenshot 4, you can see how the texts, foreground images, and other things change as well, just like it is in the first screenshot. Recall where the jump is and note how the Y positions are changing. The leftmost of the position columns is the closest of the ground fog, a scaling of 200. The next has a scaling of 600 then 1000, and so on to 3000, the edge of visibility being at 3200, but rounded off to get a more accurate fog effect. The full tutorial is here which gives a full background, history, and even the entire process I go through to make an animation.

I can imagine doing a google image search and someone finds this animation in there. The see the 4.08 MB file size but note the simplicity. The figure it's animated, but at a 320x240 size, something didn't seem right. Yet, it's even optimized! I was actually like that when I had my unoptimized "Balloon Skydiving Without a Parachute" animation.
#2 Dream recalling rate - a major new record: Since the end of September, I've had an unusually high rate of recalling dreams. Normally, I get one dream every 2 days on average, but I'm hovering more around 5 dreams every 6 days! Considering the time span, my old record was smashed. In my last update, I noted of recalling 15 dreams in 17 days. In this update, it's otherwise the same thing - 10 dreams in 13 days. Many of the new dreams take a page's worth in detail. In just one month, that's 25 dreams recalled! At that kind of rate, I'd have 800 by the end of 2007! None of the dreams are past ones, all dreams ocurring from the night. 500 by the end of this year now seems almost a certainty and it may even be 512 dreams by the end of the year. At the start of the year, I only had about 320 dreams or so. Now I'm clear up to 477. I only need 23 more dreams recalled and there's now 61 days to do it in. My average rate is 1 dream every 2 days so 512 dreams by the end of the year is indeed possible if I keep at my current rate. Of all my new dreams, there was not one case where I had two in one night.
#3 A month straight and still going: From September 24 to today, I've had FF1 World playing. It's all because of parts 4 and 7, especially the merger of part 7 into 1. The screenshots in my Oct 18, 2006 blog show the frequency distribution of the best spot, the best 24 beats (6 measures). The peak compatibility only crept up another 20 units to 2820, but otherwise is considered constant due to the margin of error of 10%. A difference of 280 is enough to counter the margin of error. I'm now only a week short of my record of 46 days straight. It would be interesting if I actually passed it and I'm now even starting to focus on breaking this seemingly ridiculous record since I'm quite close to it. 50 days straight would be something, but is not out of my reach. Unlike Battle Zone, this song has a broader speed range of good spots so that helps there. Battle Zone had 55 to 65% true speed as the best range, a span of 1.182 (after 65÷55). This song seems to range from 85 to 110% true speed, a span of 1.294.
#4 Little progress with C: I figured out that Windows uses a winmain() function, how to define a struct, and about bitwise operators. Gamestudio supports bitwise operators, but that's it. I haven't got around to trying them out yet. I still can't seem to figure out how scenes are to be displayed (as from reading a BMP file and displaying the output). I read about file reading and writing as well. I have yet to find out how to create waits, display images on screen, and to have a function called upon pressing a key. I haven't tested any of them and it seems that looking at the source from open source programs (like Audacity - I can't get it from GIMP, of which I'm more familiar with, as I can't open the .tar.bz2 files). I can open GZ files just fine. Once I learn these last 3 things, I can then transform my 2D game from Gamestudio's programming language to C. I've been losing motive as well so that's not going to help much. Both bookmarked guides don't seem to mention displaying images. The closest is OpenGL stuff, of which involves 3D and is not what I'm after. I'm after simple 2D items like a BMP image.
#5 New visions of my 2D game and a lighting math error: Recently, I've had numerous visions of the same scene of my 2D game. It had some fairly steep slopes on the left side and less steep slopes on the right. In the scene, I use the flash attack into the slope on the left to climb it and at the top is a platform that moves left and right. The scene is in the lower area, almost where you start in the current released version. I'm often going on the platform and riding it to the right. Another vision shows a completely new cloud system where I'm on a 2:1 slope a little below the lower parts of the clouds. In the vision, the clouds are much more like cumulus clouds. Unlike my previous versions, these have lighting and like the two closest cloud layers, smoothing effects. In the vision I'm just running around on the slopes, sometimes going a bit above the clouds, other times below them. I'm jumping around a lot as well and there's jump 20 being used rather than the current (and only available) jump 50. I can probably even give it a taste of George Game 13, but as a 2D game instead of 3D and without the huge fully explorable worlds. While making my first cartoon, I realized I made a big mathematical error with the lighting. As I have the lighting, the "Sun" is toward the top right and the far left of the mountains is fully unlit. The lighting formula says that the darkest part, being the ambient, is the cosine of the Sun angle to the slope. Head on, the value is 0 meaning 0 parts black and one part source color meaning that it exactly matches the source color. If at a 45° angle, it'd be the cosine of 45 degrees or the square root of a half, or about 0.70711. If the ambient is 1/3 brightness (rather than the current 1/4 (1/2 used for the current character), you'd mix one part of the source color with two parts the cosine of the sun offset angle. This gives about 0.80474 or basically adding 4 parts of the source color with one part black. With the way I have the Sun, the left half of the object would be in full shade. In full shade, 2 parts black and one part source color are mixed. I never noticed this until I was about to do a 4-step lighting system with the first of the hills in my animation. It was unmotivating so I went with the two-step method I've been using. My 2D game will use much greater detail than my animations. This means I have to redo the lighting. The only way the current lighting system could work is if the angle was divided by two causing the far left to be interpretted as 90°, but even here, there are mistakes as the "curve" isn't right.
#6 Old pinball game now envisioned: If you've been with my website since 2002 or so, you may recall a small little pinball-like game where you had a paddle, bumpers at the top, and some numbers in them. Today, I'm so much closer to making such a game. When I learn C, I'll likely make it my first game due to it's great simplicity. I still recall the rules and game mechanics quite well. Rather than the 4-bit color, I could use a better color palette than that. But still, there's the case of having to learn those 3 elements before I make such a game. The game works like Pong, but with a strong pinball feel to it. I recall asking others to make such a game. Back then, programming was well outside my range of knowledge, but, today, I can actually make it in Gamestudio, but it would be better as a C program instead.
#7 New animations planned: So okay, that "The Mega Race" cartoon was sure a massive undertaking, but I have plans for redoing some of my older animations. You may recall one called "The Spinning Board". At the time I made it, it was very simple. I originally had it as an AVI due to the fact that there was a huge number of frames and I didn't have the batch converter. Now what would you say for quality much like that of my first cartoon? It seems a bit much, but I don't have to redo the scenery saving two days right there alone. The math behind it is downright simple and I don't need much in the way of foreground elements. I could have it done in a half day or even as little as 3 hours! With about 800 base frames, my old method would've taken a day at the fastest with the level of detail I had before, but using the near-full automation I have now, the simplicity of the animation, and that the background scenery's base layers have already been done, 3 hours to completion is within reach, even with the 30 layers I've got. My old method would've needed even a week! That's just for starters though. You may recall other animations like "The Fall-glide", "The Speed Blast", and others that simply depict a special ability being demonstrated, I can use the scenery from my cartoon to form the end result. With more formulas now available, I can give a more decent and more realistic demonstration just what it's like to use these special abilities. Ever wonder what pitched gliding is actually like? How about using the bounce stomp bouncing at 800 mph followed by the pound stomp (and a massive shockwave upon landing)? Maybe float-running up a vertical wall or the bounce ability? With the scenery premade for each of these animations, even the character and the lighting, I could finish 2 or even 3 animations per day, even if 500+ frames! Recalling my animation prime time, during 2004, I made two animations in a month for a few months in a row giving something like eight 100-frame animations in 5 months, this is two animations per day with far greater complexity. I'll be spending more time making the plans than making the scenery (foreground scenes and objects mainly).
#8 No progress with story - lost motive: You may recall that I was eager to work on my story toward the end of September and my sudden addiction to video games. Well, I happened to have lost motive in doing it. Book form is just not for me. When I started writing stories clear back in early 1996 or so, more than ten years ago, I wrote them in the style of a play. The characters were on the left and their speech and actions were to the right. The speech and actions were all mixed up. The style used for movies is much cleaner than my old style of writing. Book style is unfamiliar territory to me as I don't have much skill. I have a major weakness with quotations. My play style never had this. The only trouble spot with the play style is all the tabbing I otherwise had to do. Eventually the motive wound down reaching negative territory and I stopped working on it and other things took place of it. If book format is not for me and movies are almost out of reach, I need another method. My original play style was a jumbled up mess and isn't an option. I'd have to use a different format style, likely combining the clean-look of movie scripts and my play style I originally used. I just now need my motive to get out of negative territory....
#9 2 months since even visiting the Gamestudio forums: I haven't visitted the Gamestudio forums in now 2 months. Why am I not bothering to go there? Because Gamestudio has too many limitations (and some annoyances) for my game development stuff, and that I figure Lite-C and the much-needed feature for my 3D game could take months to be released, I've lost motive for using Gamestudio and I'm now steering toward C to get around these limitations. What does this have to do with the Gamestudio forums? Because I'll unlikely be using it for much of anything, all I would be doing is just general conversations and I don't see much sense in posting such things. I barely even have a start on my 3D game and it hasn't been worked on for now 14 months, outside mere visions and modifying the design plans. It's all because of the much-needed feature of UV map scaling that is missing. By having to manually edit the positions of each and every skin vertex for 3000 times per model and hundreds of models taking 15 seconds per vertex, it adds a monster -100 affector to my motive, strong enough to turn a powerful +70 to -1.3. The other limitations (the var being the big one and that I won't be able to define camera and entity positions as a float or double and other things like that) further add negative affectors. At -1.8 motive, I have no intentions at all of working on my 3D game. I now have a worthy skill of making a 3D engine, but I have yet to actually find out how to render polygons and especially textures. I know how to compute Z-buffers just fine and that drawing the scene is done background to foreground. I also know how to compute lighting effects for basic lighting, but beyond that, I don't know. My 2D game is not worthy in Gamestudio due to Gamestudio's major limitations and restrictions for it. By learning C programming, I could easily meet my expectations for my 2D game - no DirectX or 3D hardware requirements (thus, it could work in old school computers). I could even add my most-wanted features - slopes, pitched gliding, and smooth movements, the smooth movements is not possible with Gamestudio at all. By smooth movements, I mean the sudden one-pixel jerks. At a scaling of 36, for example, and a speed of 8 mph, you'd see the object not move for 4 frames, then move a whole pixel, stop for 3 frames longer, then move another pixel. With full anti-aliased movement, of which Gamestudio does not offer, each frame that goes by would cause a change in the scenery as determined by my screwy formulas. Along with that, the fog color can be adjusted at free will and by changing the scaling, the fog intensity adjusted, both of which not possible with Gamestudio's 2D system, even though I know the algorithms for it. As a result, I have no intentions on visiting the Gamestudio forums for a while. I'll likely return at some point.
#10 Fun with t's and g's: I've been doing it for a long while lately. While recalling the day's events, I tend to doodle off the side, but not images, but entertainment with the t's having t-tails and the screwed up g's. Sometimes, I even have the t's with t-tails have extra-long t-tails within the text. It also occurs with dream details, but not as often. Sometimes, when I use a normal g within the details, I form a screwed up g in the space around the page (mainly the top). If the screwed up g looks good, I use it to form a word, either "good" if nice, or "great" when very decent. Those that don't look decent, about 90% of them, don't use it to form a word, except "gogo" at random rare times.

My notes used in my blog with t's having long t-tails and screwed up g's in the corner


Note the bottom left corner. Each row has four t's, mostly of the "dorky" type, some of the "stinky" type, and one screwed up g. Remember the article on how I update my website? I mentioned of a lot of scribbles and a mess. You can clearly see the scribbles and mess that I need to sort through. It's very easy in most cases, but sometimes, I need a close-up and use my very high-res vision and even there I sometimes can't tell. Almost all of it is what you see here, just with a few minor changes. Note the semicolons at the end of sentences. Can you see why I don't use periods? The only time I do is when an exclamation point or question mark is used, both of which shown at the first word of the fourth line on the page. The dot alone is often tricky to pick out. The squiggly equal sign means "approximately" in my cases. I can't use the symbol on webpages, short of using an image, so I use the ~ instead or the word (or a synonym of the word) "about" instead. Note the "±1" marked after the bottom time. Compare that to my daily entries.
#11 An insight to how I see lotteries and the like: As you probably know, I see everything as having some sort of numerical value or multiple numerical values. Lotteries are no different. On the HWW forums, a thread was started about a lottery-like system where you can win $1 million by voting in some state (Nevada?). I figured 10 million residents live there (I don't exactly know) and by voting, you can get a return of ten cents which doesn't seem like much. An argument/debate started about this. I further explained how I arrived at that and gave numerous examples. Pretend you have 50 points. You bet one point each time you roll a 6-sided die. You guess what you roll then roll the die. If you guess correctly, you get 5 points. If you had 46 points, the bet would drop you to 45 and the win would put you back at 50. Using the laws of probability, you can expect one in six correct guesses. After extended time, you'll soon run out of points. Another related thing was thought of for a coin toss. You bet 3 points, guess which side the coin will land on, flip it, and, if you win, you win 5 points. The same thing happens with that lottery system. I even compared the Powerball lottery and worked out the details. Although there are more variations in Powerball such as multipliers and multiple winning combinations, they all have to be factored in. Getting the multiplier has the better effect, a much better effect. The odds of winning don't change, but the prize can be multiplied by an average of 3.5 (assuming each multiplier value has the same probability of occurring). Another important thing to factor in are taxes. Taking the lump sum is a bad way to go as you pay 2/3 of that (or something) as income tax. By taking the installment method, you pay less in taxes and get the same amount of money in return adding to your advantage. It takes a while to work out the odds and decide when to get a ticket so that the returns are in your favor as the jackpot has to be up quite high. I don't have the exact value, but I suspect it's in the 200 to 300 million range. This applies to all lotteries, including the one originally focused on in the thread. Remember the little game with the die I mentioned just above? You'd be losing 1/6 point each time you roll the die. Where does that come from? Let's take many rolls into account. Let's say we roll 60 times. By the laws of probability, you'd have won about 10 times and gained back 50 points. From 60 rolls, you'd have lost 60 points. 50-60 is -10. -10 divided by the number of rolls, 60, gives -1/6 and thus you'd be losing 1/6 of a point. Sure you may pass 70 points initially or take 200 rolls, but to use up all those points, on average, you'd need 300 rolls (as 50 divided by one-sixth is 300). A similar thing would occur if you started with 500 points or even a trillion.
#12 function bad_thing() {(terminate_if_bad) function_calls();: Yet another derivative of the "function action()" case is the case of function bad_thing() and if it's bad, the function terminates. It goes something like this:

function war()
{
   if (war_is_bad == true)
   {
      return(0); // terminate the function
   }
   
   use_guns(); // basic function call
   use_explosives();
   set_traps(land_mines, covered_holes); // function call with some parameters
   enable_fist_fighting();
   intensity = 100; // 100%, the maximum // setting a variable
   allow_sneak_attacks();
   // and so on with function calls and other related things
}


This gives a good idea. There's another one for riots, thieves, and other things. If bad, function terminates so it doesn't execute and thus no war, riots, or thieves. This is the style of Gamestudio. The only difference with C is that I'd have "int war()" instead (or any variable type where "int" is, and without the quotes of course) as the first line. The rest is unchanged. I began doing this when there was a large amount of stuff about the war mentioned on the news and more elements were added when watching something about various disasters (like a riot springing out from a sports game) and eventually spread out to other things.

Oct 18, 2006



Oct 18, 2006: My first C program was made. A new animation similar to but twice as complex as my last was created and my mind game report was updated as an indirect result. I've, unexpectedly, been upgraded to 4 Mbps internet speed. An extensive experiment on color printing has been done to be able to print at higher print quality settings and have it print the way I see it on my monitor. FF1 World (my song), is now passing 3 weeks straight, halfway to my recently-set record. My story has made some progress as well, but not as much as I was hoping for. I got my new monitor, but with a few issues.

#1 My first C program: As you probably know by now, Gamestudio has just way too many limitations and annoyances for me, mostly limitations. I had a sudden urge to try and attempt converting my 2D game into a C program. Before I can do that, however, I have to learn the syntax, how to get windows like the Windows OS has, how to display images on the window's display, reading and writing files, and getting it so that, upon a key being pressed, something happens, without using scanf(). I've already got a good start on it. In fact, I made a fully functional C program that utilizes the scanf() and printf() function calls as well as a few other things. It's nice being able to work with floats, doubles, ints, and other variable types. So what does my first C program do? Well, you just start it up and select one of five options, to quit, or do a few side tasks involving basic input. The first one I did was a height test where you enter a value in coordinate units and I did a conversion then reported how good the height is. Just from this first part alone, I've already figured out how to do quite a few aspects of my 2D game. The second one I did was a spell power test, which does various conversions and computes the amount you'd pay in your electric bill to cast a spell. This was just a more complex practice run of the first one, the height test. The third one I did was a slope test, where the horizontal and vertical speeds are determined from a diagonal distance (sine and cosine functions used), where you choose what type of slope you have (whether in degrees or slope ratio). I needed to add the math.h include before doing this. The big thing that often throws me off with this is that sin, atan, and other such functions use radians rather than degrees so I always have to remember to add the "/180*pi" after a chosen angle to get radians so it behaves as I expect. I have almost no experience with radians, but I do understand what they are. A half-circle has pi radians. My fourth one was the use of the distance-acceleration formula. So what have I learned so far? See this table for a good overlook - I seem almost one quarter of the way there to having what I need for my 2D game.

What I learnedUses in my 2D game*Importance to me**Gamestudio limitations workaround effect***
Basic syntax style and formatEverything1 - very lowThings are considerably more confusing
printf()On-screen text and menus5 - moderately highI don't have to use several ifs to add variations, just a number or string
scanf()Custom inputs as for setting the scaling3 - moderately lowNone known
if, else if, and elseFaster processing of my numerous conditions5 - moderately highGamestudio has just if and else and excessive use of else causes missing or extra }'s and 20+ levels of indentation making the code messy. The "else if" instruction is of greatest use.
variablesAllowing for pitched gliding, slopes, obstacles, more realistic movement, and much more‡7 - very highThe var frequently gets in my due to its low range and accuracy and it adds frustration and great complexity†. I also have to remember to add a ".0" at the end of integers to get floating-point precision, of which I'm no where near used to.
for, while, do-whileWhile and for loops seem to be the most commonly used of the three, do-while with very few uses5 - highfor loops are one main thing that could be of use
switch-caseMy menus, ability usage, some special effects (puff sizes), and some other things.6 - highJust a whole bunch of if statements
functionsEverything4 - moderateI have to remember not to use "function" to define one, but what the variable type returned is (void if no returns are used)
includes and definesincluding headers and defining common things4 - moderateGamestudio supports both, but I have to remember to add the "#" before it, unlike Gamestudio
- - - - - - - - - - - - - - - - - - - - - -
What I have yet to learnUses in my 2D gameImportance to meGamestudio limitations workaround effect
Creating waitsProcessing functions one per frame4 - moderatePerhaps getting the frame rate set to the monitor's refresh rate for very smooth motion, but otherwise none that I know of.
Getting actual program windows in WindowsAn open/save dialog (to allow multiple save files) as well as fully customizable resolutions5 - moderately highGamestudio does not support this at all unless you manage to make your own file system, the closest possible to it, but you still can't get the Windows-like atmosphere.
StringsDisplaying text that manipulated by the userX - unknownGamestudio supports it, but, since I don't know what C has to offer, the certainty is very low.
Reading and writing filesThe save/load systemX - unknownI have no idea what C offers, so I have no idea
Displaying graphicsThe scenery and graphical interface, possibly allowing for ultra smooth movement from anti-aliasing effects6 - highIt's easy with Gamestudio, but I don't know how C compares.
Press a key, something happens, without scanf()Jumping, gliding, selecting menu commands, and the other game controlsX - unknownI don't know what C offers.
PointersReferring to the various graphic elements and possibly other things4 - moderateGamestudio supports it, but I don't know what C offers.
StructsUnknown (panels and text objects are structs)X - unknownGamestudio has them, but custom ones cannot be made.
- - - - - - - - - - - - - - - - - - - - - -
Other limitations and annoyancesUses in my 2D gameImportance to meGamestudio limitations workaround effect
3D hardware requirementsEven old computers without 3D hardware can run my 2D game.6 - high3D hardware is always required and cannot be disabled.
DirectX requirementsEven old computers barely with DirectX 6 could run my 2D game - I don't need DX.6 - highDirectX 9.0c or it will not run and there's no way around it.
Operating system requirementsBackward compatibility to Windows 98 is useful††5 - moderately highGamestudio supports 98, Me, 2000, and XP
Hardware requirementsCould work on old budget systems running Windows 98 (300 MHz processor, 32 MB memory, 400 MB hard drive, etc.) allowing for a broad range of audiences5 - moderately highGamestudio requires higher end hardware shrinking the audience.
Load timesJust double-click the EXE and you're already viewing the only logo6 - highYou don't have to wait for the "loading SOLYMPICSV24.WDL" notice to draw 2 1/3 lines' worth of dots before the game loads and you need to wait for two logos - this cuts the load time by 4 to 12 seconds, depending on the speed of your computer and whether or not you previously loaded it and have the full edition.
File namesI may be able to stick with my usual file naming system for everything.6 - highOnly 15 characters and spaces and other otherwise accepted characters are not allowed - annoying in that its so few!
Use of Molebox or other file packersThe problems Molebox poses should not pose any problems7 - very highA file packer is required because, without it, the WDL script can easily be read by anyone.


Table footnotes:
* This is what aspects of my 2D game are covered by. The printf() instruction is used for displaying text in my 2D game and is very useful in that sense.
** By learning C, the importance of this feature/instruction is compared to that of Gamestudio. It is based on a scale from 0 to 8 where 8 is of extreme importance and would easily steer me away from using Gamestudio and toward C. 4 means it could go either way and 0 means I'd lean toward Gamestudio much more than C programming.
*** Gamestudio has many limitations or annoyances that prevent me from doing much. This columns just gives a basic idea what limitations Gamestudio has and how C works around them.
† To explain what kind of complexity I need to improve the very low precision, I need to do all this:

var test_number; // global vars because other functions use these
var test_number_course;
var test_number_fine;

...

function add_pi()
{
   /* This fuction adds pi to its previous value */
   
   test_number_course += 3; // add 3 to the integerical value
   test_number_fine = 141592.6534; // add fractional value
   /* The fractional value must be rounded off to the nearest multiple of 1/1024
   with the fourth digit increased one unit to prevent it from rounding down -
   this allows for maximum accuracy */
   
   if (test_number_fine >= 1000000)
   {
      test_number_course += 1;
      test_number_fine -= 1000000;
   }
   
   /* like numeration, this uses the overflow like adding 1 to 99 to get 100 */
   
   test_number = test_number_course + test_number_fine/1000000;
}


This is unneccessary, but that's what it takes to get more accuracy, up to 9 decimal places. Without it, strange things happen as a result and they become noticable in a short time period, in the matter of a minute or two. For something like the position stat, I need a third variable (test_number_super in my example above as I'd call it) and a much more complex system twice as complex as this. With the double, it's as simple as "test_number += pi;" and I don't need to define those extra variables either. Floats have 7 significant figures while doubles have 15 significant figures. I wonder if there are 128-bit floating point variables available. Such a variable would only be useful for the position due to its extreme range. By 2 billion, you've only got 5 decimal places of precision which seems a little low. When 256x time is implemented along with slopes to get to extreme heights, reaching 100 billion is easily possible in a relatively short time period. This leaves only 3 decimal places of precision and is too low. Although I do see how it's possible to prevent this, utilizing my trick I use now, I could still get the positions set to avoid this issue, but defining a single 128-bit floating point variable would be better rather than 3 64-bit variables (I have to define 4 vars to get the position to work properly in my 2D game, 128 bits).
‡ See the "future plans" page of my 2D game's description and look for the 1, 2, and 3 notes for the various features. These indicate features I cannot add in Gamestudio. Just look at the motives for some of them - the mid triple digits even (200 to 500 - it's logarithmic rather than linear).
†† I could, possibly, port to Linux or some other non-Windows OS, but I'm unable to test it. I've only heard of Linux and seen very few screenshots of it, and others like Unix I've never seen at all. I've used Mac OS, but a much older version (whatever was around in 1999 and early 2000) and I can't test in this OS.

With what I know, I can make half of the programming of my 2D game already and that spans just 2 or 3 days of learning it! Once I get a standard window display and learn how to create waits, display graphics, use pointers, and getting it so when a key is pressed, a function is called, I can then, for real, make my 2D game in C. A prerequisit to this, however, would be my real very first program, what was version 1.0 of my 2D game which demonstrated what effect scaling has. Master this and I'm on my way to reprocessing my 2D game entirely in C. Oh how nice it would be for slopes and pitched gliding getting 300+ miles high.... Float-running up a mountain then heading back down faster than even 1000 mph, or just jumping for joy at the top of a peak just short of the clouds. Heh, I could almost make a 3D engine, but there's a few problems I don't yet know. The big one being how to get textures drawn onto polygons. I know how to place vertices now, but getting the faces (triangles) drawn as well is much more difficult as I don't know how it's done. Then there's the textures. All I can think of is a distance and size calculation for each individual pixel and this would seriously strain the CPU's time causing low frame rates. There's also lighting and fog as well. I know the formulas for simple white lights and that of any color fog (both of which are derrivatives of my color-averaging formula, of which I've known since 2002). Right now, 3D is at the bottom of my list of concerns. As a special enhancement, I'm making a quick reference guide that explains the various instructions, what they're for, how to use them, as well as the general formats. That way, just as I'd search the Gamestudio Manual for an instruction to learn what it's for, I'd just search my quick reference guide.
#2 No more video games: In my last blog, I've mentioned of playing my games for a very long time. However, I haven't played them at all. I did, however, add one thing related to my games - a small area you can build with debug mode in Sonic 3 & Knuckles to drown Tails the fox. Although not very thoroughly tested, it has been tested quite well. It was actually my very first fox drowner. I have another one related to this for Mushroom Hill Act 2 just after the boss, but I don't have any of the details and the fox doesn't drown. However, it lets you see a rather bizarre bug in the game with the battery blimp - it accelerates to well beyond 512 pixels per frame and strange things happen as a result of this. The screen's width is 320 pixels to put that in perspective and the fastest the screen can scroll is 24 pixels per frame, tiny in comparison. The fastest I've gotten Sonic is 128 pixels per frame, the absolute maximum, only possible in Sonic 2's Metropolis Zone Act 3 with debug mode enabled (and some tricks) in the area just short of the boss. In fact, it's the fastest I've seen in any 2D video game (except mine). I wanted to post some screenshots of what I mean by drowning Tails the fox, taken in 2002 or 2003, when I had my old TV capture card that was very staticy. I modified two of the screenshots long ago to get true color and clear up the mess (thanks to heavy repetition, it was relatively easy, although it still took a long time) but the other one is not causing the reds to get seriously amplified (or the greens and blues deamplified). I'm posting them in this update.

Drowning Tails - a drowned fox in Hydrocity Zone Act 1 near the end Drowning Tails - the number bubbles (a 2) with the fox about to drown
Drowning Tails - using debug mode and my first design concept for Hydrocity Zone Act 2, I made a custom area dedicated to drowning the fox Drowning Tails - debug at the start of Launch Base Zone Act 2 is where I built one area filled with spikes


The first two have been modified but the last two haven't. Note how the colors are seriously distorted in these. You can still make out what the various parts are. In the first screenshot of the unmodified ones (bottom left), some of the objects are missing due to the fact that there are too many things for the game to render and the later ones don't appear.
#3 New animation: You may recall the "Extreme Mountain Climbing" animation I made about a half year ago where the character in the animation ran up a mountain and landed on its peak. My new animation follows this, but in many different ways. The first is that I fixed the flaws I had. It was supposed to be a 4:1 slope, not an 8:1 slope. Also, I was to jump from a 1:4 slope when at the top to go straight up rather than the previous 1:2 slope. The fog is also more realistic and more mathematically accurate and there are 8 layers of mountains rather than 7 (17 total layers, not including 3 foreground layers). For enhancements, the biggest case is the very realistic scaling. The mountains have twice the pixel height as the previous one, but also have double the scaling making them peak just over 7000 feet above the clouds rather than the previous ones just under 1800 feet. This gives a much more realistic appearance. Icecaps were also added along with a change in the last of the 1:2 slopes near the very top further enhancing the realism. Unlike the previous animation, since I now figured out how to do it, the character also float-runs back down the mountain at a speed just shy of the speed of sound. He impacts a wall causing a shockwave, again with a more realistic feel than my previous attempt at it. Unlike the last animation, there isn't any frame with credits. The file size is about 40 KB larger and it is twice as complex as my previous most complex (yes, a record), and it's still nothing to my greatest potential. I'm even having thoughts of having a sky of broken clouds in an animation with a very strong 3D feel (now that I know the 3D formulas (yes, even the trig method is found, at least one way of doing it)). Perhaps an enhanced remake of the "The Speed Blast" animation with the lakes and some 3D-behaving flat-surfaced clouds above? Unlike my 2D game's clouds, this is a completely different design concept. Using my old method, such an animation would take weeks, even months to complete just 200 frames. With scripting it in Gamestudio, the entire thing can be processed at nearly 20 fps, done in the matter of only ten seconds, tiny compared to 4+ million. This animation I made using the old method would've taken four times longer, ignoring the fact it took a whole day to get the very complex calculations set up in Excel. I could do a 2000-frame animation in only two hours at the fastest (as I could just "borrow" the images from previous animations so I don't have to redo them). I've been wanting to do an animation featuring gliding from a full stop to 800 mph, but, even with one layer of objects with a high scaling, it would still take a half week just to process it my old way when it would take under an hour using Gamestudio. The next big upgrade my animations could use is the move to Flash or something besides GIF, but Flash's price is ridiculously high ($400 (four hundred)) and unjustified. My animations are very strongly suited for Flash, but it would take a motive higher than anything I've experienced to justify me getting that (think 3190 is high enough, try something in the mid 6000's!). For now, I'm stuck with animated GIFs and the gigantic file sizes they have. I just want to replicate what I do with GIFs in Flash - by setting positions determined by my very extensive spreadsheets (if it can read text files to set element positions, that's good enough). If you were to look at Flash animations and compare it to my animated GIFs, you couldn't see any real difference between the two. So what do you suppose my next animation is? Now that I know the 3D formulas (the trig-free method and the trig method), I'd like to make an animation where strong 3D effects are used. I've had so many animations I previously attempted such as gliding from stopped to the full 800 mph. Because this would take 1600 frames to do and without the batch processor, I found it unfeasible. Even with the BMP to GIF batch converter, I'd still have serious negative motive for animating it frame by frame. Today, such an animation can be done in a mere 4 hours, even with 50 layers of objects to get a strong 3D effect in it. With the 2-layer configuration, from start to finish, I could be done in only 4 hours. How about a in-depth demonstration of the various special abilities? I once formed the plan for a 3000-frame animation, and this was during the time I had the BMP to GIF batch converter. It'd have taken over a week just to process the frames one by one. Today, by writing a program, I can process all 3000 frames a smaller chance of error than I had with 20 frames in my old methods. Now the only errors would be those in the spreadsheet, creating the basic images, and bugs in the program (testing beforehand can help fix these bugs and as a result any bugs in the spreadsheets). My newest animation had 795 base frames, almost 600 unique ones. Yet, I only had one minor error when I'd expect about forty using my old method. If I could learn C programming to a full enough extent, I'd write a program to make animations using my techniques and form the animated GIF.
#4 I now have 4 Mbps DSL!?: The day I first got DSL, April 28, 2005, was a nice upgrade as I could at long last dump the slow dial-up connection. The 1 Mbps connection was amazingly fast. Before the changeover, downloading a 25 MB file would take just over 2 hours. After the changeover, it took about 4 minutes, a huge difference and I could actually watch the download progress until its completion since it was that fast when normally, I'd just go off and do other things for a few hours while I waited. Just recently, and unexpectedly, I discovered that the speed was upgraded to 4 Mbps. How did I find that out? When I was downloading Visual C++ 2005 Express, a 91 MB file, I was expecting for it take at most 728 seconds or just over 12 minutes, but when I saw the transfer speed at 200 KB/sec, I was very confused as that was more than twice as fast as it should be (then again, it could've been related to a timing bug as Firefox has). Rather than going down to the more expected 120 KB/sec, it actually rose and kept on rising until it was around 480 to 500 KB/sec which meant I had 4 Mbps DSL. Even the amount downloaded made sense. Rather than the expected 12 minutes maximum, it took almost 4 minutes and further proof of this comes in the file size. I thought hardware upgrades were needed for this as well (the DSL modem being upgraded, but as far as I know, it hasn't been changed at all). This happened during my time of my extreme involvement with playing my video games.
#5 Mind game report moderately updated: I've recently updated my mind game report as a result of my new animation. I was surprised to see that some pages haven't been updated since July 11, 2005! Back then, my 2D game was only a tiny fraction completed and likely as version 2.1 or even 2.0. As a result of this, often containing irrelevent things about my 2D game, I read through the entire report otherwise updating the entire document in parts and pieces. Some things were left as is though. I also made a collection of spelling, typo, and grammer fixes that I spotted.
#6 A bad day - a first in nearly a year: The last time I had a rather bad day was nearly a year ago, likely more. In fact, it's the first actual case in my blog (September 30) where the magentas are actually used! Although I've had a few cases where it gets into the blues, this is the worst in quite a long while.
#7 Massive printing and color experiment: When I print an image with the color 00FF00 on it (or ######), it appears as if it was actually the color 00A000 (or ######) thus considerably darker than it should be. The same goes for red, yellow, blue, cyan, and magenta. It also happens for colors between here, except hues from 0° to 60°, which spans red-orange, orange, and yellow-orange. 0° is red and 60° is yellow. 30° is orange. Orange, for some reason, prints out much closer to the true color, but is still slightly darker. I wanted a way to get colors as shown on my monitor to print out about as they appear on my monitor. A spectral fade from black to 00FF00 has very low contrast as a result, even on paper with a brightness of 92 (unlike the cheaper 84)! The printer printing darker than it should is the main cause. Thing is, if I use the lowest print quality setting, the colors appear fine and there's still little contrast, but the image looks very poor with the rather large specs. By printing at higher quality settings (normal and best), the print out is much darker than it should be, but the specs otherwise go away. I don't want the specs to be all that visible, I want the colors to not be darkened, and I don't want to use the more expensive paper. The solution to this was to run an experiment and look for patterns and indeed I found several of them. The shades of gray print exactly as they should. Pure hues (a color with one value being 255 and another at 0 (the extra one can be anything)) except oranges need to have 1:1 mix of the source color and white, the output determined by my frequently-used color-averaging formula. To get green to print out about what it should be, I need to change the color 00FF00 to 80FF80 (or ###### to ######). When I printed the 80FF80 color, it very closely matched that on my monitor. I first did the experiment starting by printing various spectra, often a fade from black to the target color. Red, green, and blue all seemed very linear. That is, halfway from black to the target color is where the color appears half as bright as the target color at its brightest rather than 70.711% as bright (as from a geometric scale). Then, with 3 test colors, red, green, and blue, I added progressively more white to the original color in large squares to see how much white was needed. I did the same for another set of 3 colors: cyan, magenta, and yellow and got the same results. I made the square bigger to include colors on the halfway point and orange was the only real oddball. Unlike the 1:1 mix, I needed a 7:1 mix of source:white to get orange correct. Then, I hypothesized that, if darkened to half the brightness, the amount of white is further cut in half giving a 3:1 mix. That is, 008000 should be 40A040 (###### and ######). The results I got were about as expected. With 3/4 the brightness, the amount of white, 0.5, is cut to 3/4 that or 0.375 (3/8) meaning a 5:3 source:white mix. For 00C000, I'd get 60D860 (###### and ######). I then tried mixing the source color with one part gray turning 00FF00 to 40C040 (###### and ######). Because of the gray mix-in and that gray printed as it should, I needed to reduce the amount of white to add since all shades of gray get none. Being halfway toward gray, the amount of white is cut in half meaning a 3:1 mix. This turns 40C040 into 70D070 (###### and ######). The result was as expected. Next was a blend of grays and brightness. Darkening 40C040 to half the brightness (206020; ###### and ######), I should expect the amount of white to be further halved. From the source, it's 0.5 for the white. From the gray, it's 0.25. From the brightness adjustment, it's 0.125. 0.125 is 1/8 meaning a 7:1 source:white mix ratio. This turns 206020 into 3C743C (###### and ######). Normally, I wouldn't use a color like 3C743C as a base, such colors are only used if calculations tell otherwise (as in this case). I've tested it with orange as well. FF8000 being darkened to 804000 yields 0.125 going to 0.0625 or a 15:1 source:white ratio. I tested it for all the colors eventually using 6 pieces of paper with 24 printed objects on them from all these tests. With the whole system confirmed, it was off to finding the algorithm. I had a strong feeling that the highest and lowest color values were used and thus many of my test formulas had these. None of them worked for all test cases. If it works for just 3 or 4, it isn't valid, but if it works well for 12 test cases with the first ones being easy and the later ones being complex mixes, it's quite promising. The test cases have known values and I use the formulas to obtain the known values. The closest I came was a formula that worked for all but 2 test cases and I eventually lost motive. It was until I went to bed that I thought of the most worthy solution. My feelings were right - the high and low values are indeed involved and they play a huge role. The formula I came up with is this:

orange_hue_offset = abs(30-hue)/34.2857143+0.125;
amount_of_white = (high-low)/510*orange_hue_offset;

The "orange_hue_offset" is only processed when the hue is between 0° and 60°. It has a value of 1 otherwise. The 34.2857143 is actually the exact value of 34 2/7. High is the highest of the three values in a given color. If the color was 4080FF, that of which I use for the sky in my images (not my 2D game though), high would be the 255 value (FF). Low is the same as high, only the lowest value, which is 64 (40). The hue, 200° is not within range so orange_offset is 1. I already know how to calculate hue, it's relatively simple. With this, I can just utilize Gamestudio and write a program to modify BMP files, reading from a source, then writing output in another output file leaving the original untouched (except from reading it). I'd then print the output file to get what I need and compared to the original, there are some differences, but the end result is much closer to what my monitor has.
#8 FF1 World - 3 weeks straight and still going: It is now the only song with over 32,000 plays, just past the halfway point from 10,000 to 100,000 on a logarithmic scale. Sep 24, 2006 is the starting point and since that day, I've been at the same song only changing the speed from about 86 to 108% of the true speed. This, however, is only halfway to the record of 46 days, but I have no intentions on going that far. The compatibility has risen slightly and I may last about 10 days longer at the maximum, unless the compatibility raises even more. It's been steady for the last 10 days so I doubt this is the case. Ever wonder what the best part looks like? While making an equalizer-enhanced version of the song to deamplify the bass and amplify the 1000 to 2250 Hz area, I recalled Audacity turning the waveform into a big mess of blues, magentas and whites. I spent some time trying to figure out how it done and I eventually found it. I later discovered just how it works which makes making EQ-enhanced versions much easier and faster since I could more easily see the frequency layouts. I took a few screenshots afterwords and I got the EQ-enhanced version shown in the screenshots. Around 96% true speed is one of the sweet spots. I also took a screenshot from true speed after using the hex edittor to change the sample rate and reloading it in Audacity as well as another of the most recent play counts, behind about 20 by the time the update is completed.

FF1 World best part - 96% true speed spanning 11 to 5000 Hz with cursor at the area of the highest spike


Screenshot #1: With the range set from 11 to 5000 Hz, this gives a good idea on the main view of the song. The position of the cursor is the about 16 pixels after the the part where the very highest spike in compatibility occurs, the merge from part 7 to part 1 of the next loop. Do note that this is 96% of the true speed and thus the pitch and duration are offset a bit (lower pitch, longer duration). My favorite range is from the 1000 to 2250 Hz area. You can clearly tell what the notes are like and their durations (see those quarter notes?). The scale is linear rather than the more expected logarithmic scale. This is 6 measures or 24 quarter notes. Yes, the screenshot is JPG-ed (pronounced as "jay pegged"), PNG does much worse in terms of file size.

FF1 World best part - 96% true speed spanning 11 to 20000 Hz with cursor at the area of the second-highest spike


Screenshot #2: From 11 to 5000 Hz is a rather narrow range. The human ear can detect up to 20,000 Hz. To help, I set the full range out to 20,000 Hz to cover the entire frequency band. There is a very high-pitched hiss in the song that I don't like, but then again, what are equalizers for? Much above 2000 is often strongly disliked. This is one of those songs with a rare exception. As you can tell, all the focus is below 2500 Hz. The cursor is at the second-highest spike in compatibility.

FF1 World best part - 96% true speed spanning 11 to 3000 Hz with cursor at the least liked area of the best part


Screenshot #3: With nothing in focus above 2500 Hz, I set the frequency span to a maximum of 3000 and got this screenshot. Here, the cursor is at the least-liked place of the best part, just before the very best part actually starts.

FF1 World best part - 96% true speed showing the waveform with cursor just after the area of the highest spike


Screenshot #4: The waveform is shown here with the cursor now 32 pixels to the right of the point of the highest spike. With the waveform, it's hard to tell what the song has in terms of frequencies.

FF1 World best part - true speed spanning 11 to 3000 Hz with cursor at the area of the second-highest spike


Screenshot #5: Wanting to show true speed, I closed the song in Audacity, opened my hex edittor, changed the sample rate, then reopened it in Audacity. The difference between true speed and the pitch and tempo to that of 96% true speed is miniscule, but compare this to screenshot number 3. The length is reduced to 24/25 of the original but the pitch is increased to 25/24 of the original. The cursor is just short of the area of the second-highest spike.

Ulillillia's Winamp play counts as of Oct 18 2006 - one-third of 100K plays on one song with 9 others past 10K plays


Screenshot #6: If you recall from my previous Winamp screenshots, the song being played (now at 72% true speed), was just past 22,000 plays. Since my latest case with listening to it, it's now just shy of 1/3 of 100,000 plays. This screenshot shows my latest play counts. Yep, 23 days straight and there's nearly 12,000 plays racked up on it, about 500 plays per day. This is normal to me. Since this is 72% true speed, the time is 100/72 longer than the actual true speed time.

For 3 weeks straight, this particular area is what is helping a lot to keeping me at the song even longer. Although I do change the speed, it's within a moderate range. The version on my computer has had a peak compatibility of 2800 and is generally more around 2730 for the peak for the normal version. The version on my MP3 player is only at 2590 which isn't that much of a difference. I should be able to go on for about 34 days at the maximum, about two weeks shy of my 46-day record. The song has 40 measures in all. For each part, there are 8, 8, 4, 4, 8, 4, then 4 measures. The screenshots above show measures 37 through 40 with 40 going seemlessly into measures 1 and 2. The very best areas are measures 22 through 24 and 38 to 40 going into 1 through 3. It was parts 2 and 5 (measures 9 through 16 and 25 through 32) that were the least liked, but now it's parts 3 and 6 and are least liked with the previous least-liked parts being a very close second. The last 3 measures of parts 2 and 5 are the least liked part. Part 1 is still second place though and is far from the first and third place.
#9 Story progress - slow but not steady: At times, I've been working on my story again, reaching what was scene 19 in my original version. At some areas, I had to make some major changes, but most of the story is left unchanged. The changes are things like the amount of spell power one has. One unit of spell power is 2 gigajoules of energy, a lot of energy. It is impossible to have as much as 1E800 spell power as there isn't that much energy in the universe. Other than this and a few enhanced descriptions, there's been little in the way changes outside simple typos and spelling errors. There have been days where I make a lot of progress in my story, days with no progress at all, and others with little progress. I'm now just short of 2/3 of the way done by page count (214 out of 324 from the script). The page count in the book form is only in the 80's, but I'm using Arial, size 12, and half-inch margins (I don't know what the typical size of a book is to help compare. I'm looking into Vantage Press, and once checked out their website for an E-mail and FAQ to see if my main question is answered - are color images supported? I couldn't find any E-mails and the FAQ only answered a few of the questions and I'm left with just this one lingering question. Graphics are supported, but I don't know about color graphics.
#10 New monitor, a few issues: 1920x1440 resolution at last! However, there's two things that bother me about it. In the horizontal center, the image on the monitor is blurred slightly because there's a 1/4 pixel shift upwards. The monitor also has somewhat lower contrast which makes my somewhat common black-text-on-a-dark-gray-background issue even harder to deal with than it needs to be. After all, even with size 9 of the font Arial, I can still read the text quite well. How do I know it's a quarter pixel? Mathematics and I can actually see it. From the top of what the pixels normally would be, I can see something above it as if mixing 3 parts the normal background with 1 part the color below it. The color 404040 when viewing white on a black background is extremely close to the color I was seeing and mathematically, this is 1/4 of a pixel shift and that's how I get this. How do I know it's upwards? Again, mathematics. If you shifted a white-colored pixel into black 1/4 of a pixel, the pixel that was black becomes the color 404040 and the pixel that was white becomes the color BFBFBF. I see this same effet which is how I know. Sure it may seem minute, but it bothers me. My old monitor didn't have this issue at all. I don't know if I'll be returning it though as I'm debating on it. From my previous new monitor episode, the difference was far more than this where I could see a darkened area and a reddish-colored area. That's like it was offset by 1.5 of a pixel (although I don't recall any offset, this is just an example). So, what kind of effect does have? Try my simulated examples here for the normal view and here for a simulated view. I view my monitor from 2 feet away and I can still clearly pick out this blurred effect. I have high resolution vision, so fine, that I can actually see the curve-shaped pattern which resembles something like this:

Monitor blur effect - vertical stretched 64x showing a curve with the high point in the center

Monitor blur effect - the true scale of the upwards curve


The vertical part on this image has been heavily stretched, but it gives the rough idea. It simulates something 3 pixels high in a band of white pixels 2 pixels high. The entire thing's width is the unstretched, at 768 pixels wide. It is white on a black background. The image below it is what the true size would be and it very closely matches the very effect I see. Unlike these objects, it would stretch 1920 pixels, but this just simulates the effect I'm seeing.

If your monitor has the "balanced" option, which allows for the left and right to create a curve similar to the images above, this is the exact effect I'm seeing. Too bad my monitor doesn't have a "vertical balanced" or "vertical pincushion" otherwise I could just make a few minor adjustments to the vertical balanced and I wouldn't have it. Also, in the far top left corner, I can see a darkened area that has a slight greenish tint. The dark patch is about 5/6 as bright as what it should be, something more around 4/5 as bright for the red and blue parts. In the top right corner is this same effect, but only 9/10 as bright, 5/6 for red and green. Although I do have speakers near the monitor, I move them away and degauss the monitor, but they're still present.
#11 MP3 player time syncronization experiment, again: Almost right after I first got my MP3 player in February or March or something of 2006, I've encountered a bug in it. In some cases, I'd play the songs at 8000 Hz sample rate, especially those of the slower speeds. I even called the company to see what the problem is. The player supposedly only truly supports 44,100 Hz sample rates and I recall telling them of the strange noises when 8000 Hz was used. I told them that 11,025 Hz played back without any trouble, but slower than it should be. Since the firmware was entirely up to date, I mentioned that my best option would be to run a time syncronization experiment to figure out how much I need to speed up the song in a WAV edittor to get the MP3 player to play back at true speed. From there, I started a long list of experiments and my first time syncronization experiment lasted only about 5 hours. From my experiment results, I got the value of 1.37804 with a margin of error of 0.00001. That is so precise, it's off by at most 2/3 of a second per day. I ran battery consumption experiments and found that the battery lasts in excess of 18 1/4 hours with great consistancy, nearly 30% more than the stated 14 hours. I also saw what effect the bit rate had, what effect WAV had over MP3, what effect the equalizer, contrast, volume, and other things had. The longest time recorded is just shy of 19 hours, something past 18 5/6 hours. Then, wanting to be even more sure on the adjustment, I reran the time syncronization experiment just recently (Oct 2 into Oct 3) and instead of 5 hours, clear up to 16 hours. The results confirm my original experimentation values, only 3 times as precise. I only got 2 reliable results. When my MP3 player's clock changed to 11:40:00, 16 hours, 4 minutes, 38 seconds on the dot have gone by (±0.1 second). This gives 1.37804762... for the adjustment. For the 11:43:00 on my MP3 player, my other most reliable result, I had 16 hours, 8 minutes, 45.9 seconds (±0.1 second) of actual elapsed time. This result is 1.37804409 for the adjustment factor. Thus, I can say that the adjustment factor I need is 1.378046 now with a margin of error of about 0.000003 (although 0.000002 is shown). This is accurate to 1/4 second every day. That is, if 4 days pass, there is an accuracy loss of up to one full second and it'd be almost 8 months before being off by a whole minute. The battery lasts about 3/4 of a day so the time loss is only about 1/5 of a second. It takes twice as long to say "twice" than for this 1/5 second to pass. Any longer is not possible as Audacity won't let me go any longer for one (as I've reached the 4 GB file size limit of the file system and the battery won't go on much longer than 18 hours as it is.).
#12 Those t's and g's: My obsession with t's having long t-tails as well as the "screwed-up" g's seems to be becoming more intense, especially the screwed-up g's. The t's are mainly paper-related but the g's are both paper and vocal-related. I pronounce the screwed-up G in a rather odd way and t's in different ways depending on the length of the t-tail. Under normal means, I pronounce both normally, but for entertainment, I do it this odd way. I've had a few cases where, on my papers containing the day's events and details about my dreams, I have a little thing that looks like this:

The various types of t's and g's


Note how, the longer the t-tail, the less "normal" it is. The t on the far right is a custom-made one since I don't have any fonts with that. The "classification" is below it. I sometimes "rate" fonts based entirely on the t's and g's. Century Gothic is a perfect 5. Courier New gets a perfect for the G but 2 stars for the t giving a 3.5 rating. Times New Roman gets 4 stars for the t but 3 stars for the g giving 3.5 stars. A font rated one-star, the lowest, would be one with the "dorky" t and the "seriously screwed up" g. From left to right, it's 5, 4, 3, 2, and 1-star for the t's and 5, 3, and 1 for the g's. I even recorded an audio file to explain just how I pronounce words differently. The words "singing" and "singing" have 3 different ways of being said, the second, the one with the screwed-up g's, having two variations. Even the length of the t-tails have an effect. The longer, the more exaggerated the sound becomes. For how I'd pronounce the t's and g's in the image above, from left to right, top to bottom, it'd be something like this (MP3, 11,025 Hz, 24Kbps, mono, 20 seconds, 60 KB). The screwed-up g's, however, have 3 main variants, all of which given.
#13 Games - way too many T and M-rated ones: On Oct 7, 2006, I went into the Wal-mart Supercenter to get a mouse as a gift (a computer mouse, not the furry kind). When at the electronics section where games are, I looked to see what was there. I was surprised to find that 3/4 of the games were rated T or M, half of these being M. The other 1/4 are the other two main types present: E10+ and E. There's also an eC rating and Ao rating, but I never saw such games. It's this sort of thing, games with way too much adult content, that caused me to lose interest in playing games, except classics. If you recall from my previous sudden addiction to playing my games, they'll classics from the Playstation (the first release, not the second) and the Sega Genesis. I've played about 1/4 the games with the Gamecube (Super Monkey Ball 2 being my most played followed by Sonic Mega Collections, Spyro the Dragon: Enter the Dragonfly (this game was downright slow - 10 to 12 fps slow), and Super Mario Sunshine, all classic-style stuff. Around 2001 or so, I saw three times as many E-rated games as I do today. Around 1998, there were probably more E-rated games than M-rated games. The next-generation consoles I have no intent on getting as a result of this. There's other reasons such as the controllers being wireless. Wireless means batteries and batteries die. When the battery is about dead and you're in the heat of battle, well, you know what'll happen - control will become almost impossible or very difficult. As is, my probability of getting even one of these systems is about 2% per month starting when first released. At $500, way too high, this drops the probability even further, but the costs will soon go down and would increase the probability and one year from now, that's about a 30% chance, still somewhat small. If there were several more E-rated games, the probability would be much higher. But, then again, if I can make my own games, why bother? There is that case of having to build levels, program it, fix the bugs, create characters, enemies, puzzles, and all that.
#14 15 new dreams in just 17 days: Normally, for 15 dreams, it'd take 30 days, but I seemed to have had a recent spike in my recall rate and I may just make the 500 mark by the end of the year. Oddly enough, when rechecking the total dream count, I came up two units short for some reason. In the last update, I actually had 452 dreams, not 450. I also checked the total dream counts within each category against the total shown and they were all correct. Oh well, from this, and rechecking my addition, both using my mind and the calculator (calculator twice), the true total dream count is 467. It's still a bit too close to tell, but at the moment, I'm ahead of schedule. There are 75 days left in the year and I only need another 33 dreams and I seem to average 1 dream every 2 days. I was surprised that I haven't had a mud dream since January, well, unless such a dream was moved to the fun or weird categories as mud dreams tend to be quite fun, but also rather weird as who'd bother playing in gooey, squishy mud neck deep for plain fun anyway? Then, while updating my website, just before the update, I recall two dreams in one night.... It's an impressive bump in the frequency of the dreams spanning a number of days!



Daily Events



Oct 1: It's the first day of October, a neutral-luck month to me. I didn't recall a dream so it's back to zero in a row. I first checked my E-mail and the forums. I then continued on with my story completing what was scene 14, another one of my favorites. I ate afterwards watching TV. I then did various tasks. Afterwards, I returned to my story and continued on. I ran into the issue of the acceleration rates of the glide ability when pitching is involved. I spent quite a while trying to find the formula and eventually did. It turned out to be a lot more complicated than I thought. There are multiple parts as well and I found them all. A side benefit to this is that, when I add the pitched gliding feature in my 2D game, I'd have already got the algorithm figured out and I'd just need to include that in the programming. I continued my story to partway through what was scene 16, my top favorite of all scenes. I watched "Discovery Atlas" as I've been looking forward to it since I saw the ads for it (provided they weren't muted from a bad ad preceeding it, one with whistling, just very annoying, or is played too much - all ads after it starting with the bad ad are muted until the show returns and the ads on the next ad session are not muted until another bad one comes on). I returned to my story and went at it for 75 minutes still in scene 16. I just now crossed the halfway point in terms of the page count on my printouts, just starting page 163 out of 324. Just when I went to bed, mom got me involved with hooking up connectors for her laptop since she finally got a longer cable. I'm still hooked very well with FF1 World. The one on my MP3 player has too strong of a bass - I may make an EQ-enhanced version amplifying my favorite parts and deamplifying my least liked parts - it's my primary use of equalizers. Woke up at 6:05 AM and went to bed at 11:19 PM.
Oct 2: Nice, another dream. It takes a whole page in detail and is quite weird. Due to not recalling a dream yesterday, the streak was broken and I'm back at one in a row. I ordered the new monitor. I next continued my story reaching what was scene 17, one of the two very longest scenes. I ate afterwards watching "The Death Star" for the 14th time. It is my top favorite of all such shows. Only those classic cartoons have more higher play counts. I began processing an equalizer-enhanced version of FF1 World and found that from 1100 to 2250 Hz is the very best area. I had to deamplify the area from 200 to 600 Hz. I also prepared for another time synchronization test of 16 hours. I watched Monday Night Football during the conversions since there was absolutely nothing else on. The conversions were going 5 1/2 times faster than before - could it be the new hard drive? The EQ-enhanced version on my computer is so much better, spiking at 2800. Yet, that's without the equalizer enabled. With it enabled, there is a sharp drop to 1730, almost a 40% drop! Too much of a good thing can be bad for you and can pose serious problems. This is a good case in point of this. I'll use the equalizers on the original source using my new-found knowledge of the song where 1KHz to 2KHz are the best part with 600 and under needing to be deamplified as they are too strong. When done forming the equalizer-enhanced version of the song, I recalled a mess of magentas, blues, and whites. See news item #8 above for further details on what I mean by this (including screenshots). I found 2 bugs with this as well - man is Audacity just packed with bugs. Woke up at 8:03 AM and went to bed at 12:02 AM.
Oct 3: That's 2 dreams in a row! I first checked my forums and E-mail. I then assisted my mother in getting icons on the desktop. I had a hard time figuring out where something was taking 12 minutes to find it. I then ate. When done, I was going to work on my story again then got side tracked by something with higher motive - finding the algorithm needed for something to be printed in true color with color settings higher than the lowest. I got quite far. For details on this, see news item #7 above. I finished gathering the data needed and began forming an algorithm. I spent 2 hours testing various formulas, but it didn't work with all test cases. All but 2 test cases were working. Once successful, the ultimate test is to print out a photo and compare it to what is actually printed out. After this, I processed a sampler version of FF1 World trying out my 24-step binary logarithmic system. I watched some TV during this time and once done, I continued my story, getting about 1/4 of the way through what was scene 17, one of the two longest scenes. I began my super time syncronization experiment at 1:00 AM on the dot (the player started playing at that very moment the clock changed). I'm now at 94% true speed on my song, FF1 World, still. Woke up at 9:43 AM and went to bed at 1:04 AM.
Oct 4: Nope, no dream. I first helped dad with getting an axel installed. The axel had nuts so tight and rusted out, they, all 8 of them together, took almost an hour to get off so adjustments can be made. I was cold and hungry and went in 85 minutes later. I ate afterwards. When done eating, I watched TV for about 2 hours (until 6:00 PM). Mom wanted me to make games. I thought of one using numbers, simple ones, with very simple rules, barely even 1/3 of a page (and third graders could play). I spent 2 hours trying to resolve a puzzle - getting 47 of something in a worthy pattern that would give equalateral triangles. I then spent 3 hours making the game board and 90 minutes making the spinner needed for it. Just as I finished, I showed mom it and she liked it. She told me she wanted puzzles, not games. I began making a maze using my usual four-step process: shape, answer, dead ends, then walls. I reached the last phase just completing the third. I didn't get any time to work on my story. My experiment results are in. For details, see news item #11. I'm still at FF1 World. Woke up at 11:31 AM and went to bed at 2:09 AM.
Oct 5: Goodie! One more dream recalled, a rather bizarre one. I first continued on with making those puzzles and games. Two were puzzles, two were games. The maze I worked on yesterday was one of them. The other is a secret message that is easily decoded. The two games were 9 tic-tac-toe setups and the other is a game where you have to form boxes by making lines connected from dots in a grid. When done, I printed it out and mom liked it. Dad was rushing me to do things. One was to get a transmission into a trailer and the other was to rake up twigs on the ground in a rather small area. I saw little meaning in doing so as more would just fall anyway in a short time span. I even delayed doing this for a full day due to negative motive. I ate when I was done with it. I watched the news while eating. Continuing on with my story was next. While working on my story, my new monitor arrived. I quickly got to installing it. I set it up and it worked okay, but had two problems. The first was that the display was messed up and had to be adjusted. I spent an hour adjusting it getting rid of the trapezoid effect, curves, and other things like that. I settled with 144 dpi, 6/5 as high as my previous 120 dpi (but then again, I have 6/5 as many pixels on the width and height). See news item #10 for further details on this. When I started up Word to get going on my story again, even at 100% zoom, Word was showing 2 pages at once side by side. I had to set the zoom to 150% to fix this as, for some reason, when I press the up arrow at the top of the current page to supposedly to the bottom line of the previous page, it didn't do anything - bug! While working on my story later on, I had to clarify some facts about Volleyball as what I had in my older script was messed up, wasn't making sense, and was just confusing me. Good ol' Wikipedia! I'm still on what was scene 17 though. It's one of the longest scenes. I'm still at FF1 World, 94% true speed (93 1/3% on my MP3 player). Woke up at 11:14 AM and went to bed at 3:13 AM.
Oct 6: Wow! 2 dreams in one night, a total of 1 1/4 pages. My record for this is 3 pages. I first worked on my story finally finishing scene 17. My dad mentioned of going grocery shopping right when I was about to eat. I ate then took a shower finally leaving 3 hours after first being mentioned. I went to the Wal-Mart Super Center to get a mouse for mom's laptop and goy my usual. When done, I saw my dad waiting in the parking lot for mom. I joined the wait with him, but he soon went in after her and it was 15 minutes extra of this then, after loading groceries, it was a 10 minute wait for mom to get through the checkouts. I finally returned home and put stuff away. I installed the mouse then tried resuming my story. However, out of nowhere, my motive for learning C programming shot up. I took notes while following tutorials and this was the day I also returned to only the HWW forums after being absent from them for so long. For further details on the start of my quest to learn C programming as well as other details, see news item #1. I soon ran into an unexplained behavior with my first function. It was returning 0 when I expected 88 when the input was 600 (600 coordinate units in my 2D game is 88 feet) and it didn't make sense. FF1World is still running, but at 98% true speed. Woke up at 12:39 PM and went to bed at 4:46 AM.
Oct 7: 4 dreams recalled in just 3 days. Yep, I recalled another one, taking a full page of detail. Between floor stuff and getting on with my programming stuff, I ate nearly 7 hours later. I've now got switch-case on hand. See news item #1 for further details. One user on the HWW forums seemed quite impressed. At the moment, I'm now 5 times further than I was 3 years ago during my last attempt at C. I'm, at the moment, only 15% of the way there for program complexity compared to my first program (version 1 of my 2D game). For the last 4 hours of the day, I worked on my story getting almost done with what was scene 18. I'm halfway through the best part. I'm still at FF1 World, now 86% true speed (96 2/3% on my MP3 player). I also had a nosebleed. Woke up at 3:08 PM and went to bed at 7:53 AM.
Oct 9: I had a dream, but couldn't recall any bit of it when I officially woke up. Back to zero in a row again. I first worked on my story and, 2 hours later, I ate watching TV. One time, while watching TV, 6 1/2 minutes of nonstop TV ads were on. I returned to my story then watched Discovery Atlas - Italy. When done, I returned to my story finishing what was scene 18 and act 5. I added yet another function to my small test C program - using my distance-acceleration formula. As to the printing color-change algorithm, it's saturation that matters most. Now, if I could verify the saturation of the color A05028, brown, I have found the algorithm. See news item #7 for further details on this. Now I just need to verify it, but motive went negative again.... Once verified, a real test would be a photo or screenshot manipulated by programming it. I don't yet know how to work with files in C, let alone the familiar Windows dialogs, so I can't use C at the moment. I do, however, with Gamestudio. I'm still at FF1 World, at 86% true speed. I had 2 nosebleeds. Woke up at 7:04 PM and went to bed at 8:42 AM.
Oct 10: Goodie! That should be dream number 460 by now. I slept for about 11 hours, rather long for me. I first checked my E-mail and checked the HWW forums. I then reported a bug with XVI32. I ate afterwards watching TV for 2 hours. I rechecked HWW and my forums. I may need to replace my new monitor. For the reasons, see news item #10. I then had a stronger urge to remake the "Extreme Mountain Climbing" animation fixing a few flaws it had. I then recalled the one where I was to charge down the mountain. Earlier, I didn't know how to do it. For further details, see news item #3. I formed the main plan, of which I spent 2 hours testing the accuracy of the values. I just set the main values, but I have yet to set character positional values and what foreground is used. I got about 1/3 of the way done with the closest mountain layer with 7 to go after this. I also finished the cloud fog layers. I'm still going strong with FF1 World, currently at 86% true speed. Woke up at 8:42 PM and went to bed at 12:17 PM (±1 minute).
Oct 11: Two dreams in a row, this one with very few details. I first watched TV then ate. When done, I checked my E-mail, my forums, and the HWW forums. I continued on with the animation. I worked on the scenery and special effects, now finalizing things. At some point, mom interrupted me to do that puzzle stuff again. The maze I did a few days ago was too small and those who wanted it wanted a full page version with another puzzle to replace it. I made a word search in a 10x9 grid of letters. I fit eighteen 3 to 7-letter words into it maxing out the available space so I couldn't add any more. 5/6 of the entire grid of letters was used up in findable words. I thought it'd take an hour, but it actually took 35 minutes, 15 of which just in trying to think of at least 12 words of things common in all or most sports games (ball, foul, win, lose, team, player, etc.). Another 10 minutes were needed to figure out what grid size I wanted as well as measuring to balance grid size, text size, and spacing from the other puzzles, all calculated. Woke up at 9:57 PM and went to bed at 3:06 PM (±2 minutes).
Oct 12: 3 dreams in a row, this one just 5 full lines or 1/5 page. With nothing on on TV for a bit, I first continued on with my animation. I'm now at the scripting phase to automate the entire process. There are 805 frames in all, including duplicates meaning 80 1/2 seconds. See news item #3 for more details about it. I ate after finishing defining the bmaps. I'm eager to know what the result is and how it compares to my visions. Upon finishing eating, I finished the animation, but, there was a problem. In one mountain range, the third-closest, the problem was that, rather than the 1/4 highest part, the icecaps were noticably off. I found that the problem was that the top 1/3 was used. The base height is 40 pixels at the very tallest. The icecaps should be within 10 pixels of the highest possible, not 13 1/3 as I had. I had to redo that entire mountain range. Using my old method of going back to the first frame at fault, I could've just wasted 2 days doing this, given the great complexity. Considering I had the finished GIF before encountering this, using my old method meant spending 4 hours of having to resave the modified BMP files as GIF. I only spent one hour making a completely new set of mountains with the fix, reanimating all 795 base frames, then saving all 795 frames as GIF then optimizing the animated GIF getting 578 unique frames. I'm also envisioning another animation with a 3D-like heavily clouded sky or a seriously enhanced version of the speed blast animation. When done with my animation, dad got me at fixing cheap doors, of which my motive was slightly into positive territory, even without them begging. I spent about 2 2/3 hours at it. I made a zoom test to see what the animation was like when magnified. I'm still at FF1 World right next to 31,000 plays in Winamp and likely 4000 more from my MP3 player. Woke up at 11:54 PM and went to bed at 2:31 PM (±2 minutes).
Oct 13: 2 dreams recalled in one night, a very weird one, and a fun video game one featuring a near-repeat of a past Bubsy-in-the-forest dream where I was trying to get up super high. I first ate watching TV, including something about the US Civil War on Assignment Discovery. I cleaned out HWW and my forums then went on updating my website. For details about the mind game report being updated, see news item #5. I got at replacing one of the images of the menus since, originally, I had 1E800 spell power for Knuckles and even a half billion for humans. 1E800 spell power, or 2E809 joules, is far more energy than the universe contains. It's more around 2.5E32, about that of 8 sphere-shaped asteroids 5 miles in diameter with a density of about 2.5 grams per cubic centimeter converted to pure energy. This is based on the maximum potential though. Ultima6 takes 88 million spell power, 8.8E07 for comparison (based on maximum efficiency). While researching this, I got sidetracked with Wikipedia looking at space stuff first, then stuff about PNG and GIF, then stuff for the various years and decades around the current time and in the future, mostly the future. This lasted nearly 3 hours. I finally returned to updating the mind game report finishing it. The new animation has been added to the list. I modified the list to move the title column to the specs field and deleting the entire title column. This prevents the table from stretching excessively. Other than that, nothing else happened. I'm still at FF1 World at 86% true speed. I've still got a strong motive to continue on (no signs of getting bored of it). Woke up at 2:08 AM and went to bed at 7:34 PM.
Oct 14: Wouldn't you know it? That's 5 dreams in a row! 1/3 page though.... I first checked my E-mail, forums, and barely began the headlines of my blog. I ate afterwards watching TV. I took a shower when done eating so I could go to the grocery store. Before leaving, I was to figure out how much pipe was needed to hook up a furnace. I left but it was first to Menard's for pipe. It was then the grocery store at Wal-Mart Supercenter. Before getting groceries, I went off to get earmuffs so I can work in noisier environments without having to plug my ears all the time seriously hindering my ability to assist. I returned home, put stuff away, then went to help install the pipes, which took 3 1/2 hours. I had a fairly high motive for this, almost 2.8. When done, I began editting my blog. The headlines are about 2/3 done. Still at FF1 World.... Woke up at 5:00 AM and went to bed at 9:14 PM.
Oct 15: I had another dream, but forgot every bit of it - back to zero again. I first checked E-mail and forums, then started updating my website. The tertiary status page was updated as well as a few more headlines. I ate after this and resumed. I added the hydrocity fox drowner and finished the known headlines. I, in the middle, had to take an air conditioner out to the dumpster. I continued on finishing 1 of about 13 dreams then watched Discovery Atlas. I resumed working on the headlines, since I expanded on one, having recalled screenshots. Other than that, nothing else happened outside basic checks on HWW. Woke up at 7:25 AM and went to bed at 10:49 PM.
Oct 16: Goodie! Add one more to the dream count, 1/3 page details though taking 42 minutes to write due to my mind drifting off for long periods of time and frequently. I don't recall what I did first after this, but it was most likely working on updating my website. I checked forums as E-mail as well. I ate watching space-related stuff on the History Channel. I returned to updating my website until I noticed odd distortions on my monitor in the top left and later the top right corner. That's when I finally made up my mind. I called Philips, the manufacturer, to see what I could do about it. I got one of those machines, as expected, only this one had voice-recognition, something I almost never see or hear about in such a system. I've long known about related programs since like 2002 or so, but not in use on one of those machines you get when you call a store/manufacturer. So, with figuring it out, I got assistance. I didn't get anywhere and once asked for the serial number, they found it to be defective. I was to send a fax showing proof of purchase. That went okay, except they never got it for some reason. I was told to call back in an hour so I waited during this time, mainly reading over Wikipedia. It was during this time the defects in the top left and top right corners were analyzed in greater detail to get the numbers I needed to accurately describe it. It is very noticable in the "factory defaults" are used. I resent the fax and they got it after having to hold for 30 minutes. I'm now going to get some postal thing so I can ship it back for free. For further details, see news item #10. After this, I resumed updating my website adding a few things to the headlines, clearing all but one of my notes, and eventually finishing my dream journal where I noticed the official dream count was behind by 2. All 13 dreams have been added. I started on the daily entries completing September 30 and setting up the template for October. I have one note yet to clear and all but one of the daily entries. I did, however, upload my animation and posted it on the GIMPtalk forums. I'm still gazing at it's wonder. This is adding a 1.4 affector to updating my website. I still have FF1 World playing, but at 90% true speed this time. Woke up at 8:36 AM and went to bed at 12:48 AM.
Oct 17: I woke up at 9:27 (±1 minute) with a somewhat weird, but good dream using 4/5 page in details. I was too tired so I went back to bed. I woke up with a second dream, a video game dream involving Mario 3. The first snow fell as well. I slept for about 13 hours, very high for me, but barely half of my record of 25 hours. My website has been updated. I spent a rather large amount of time on the GIMPTalk forums. I also cleared away spam on my forums. That's the second time a second warning has been given. I first worked on my website update until 4:20 PM. I ate while watching the news. I returned to the website update since nothing was on on TV at all. I continued with the update and finished. I went to bed after this. FF1 World now at 72% true speed. Woke up at 1:25 PM and went to bed at 4:22 AM (±1 minute).
Oct 18: Great! Another dream, only vaguely recalled. Just 1/5 page of detail. I first checked the forums and E-mail. I then browsed through the tutorials for C programming eating at about 5:12 PM. With a lot of good stuff on on TV, spanning several hours, I watched TV clear until 1:00 AM. After that, I recleaned the forums then started working on a true record holder animation - 30 layers of objects and 3750 or so base frames, about the length of one of those classic Bugs Bunny cartoons. For further details on it, see news item #1. It was an animation I wanted to do earlier, but had serious negative motive due to my old methods. Now I can use my latest methods and have it done in just 3 days or so instead of 2 weeks or more. I even want to have higher quality scenery. Sure it may have a 3 MB file size, but I want to go to the very extreme. The base plan is pretty much done. I need to finalize the plans, make the scenery, program it, and finalize the result. FF1 World is still going, swapping speeds often on my computer: 64, 120, then 108% true speed in that order. Woke up at 2:28 PM and went to bed at 6:55 AM.
Oct 19: Boohoo! Zero dreams recalled - back to zero in a row! I first checked the forums and E-mail. I then began working on my animation - lots of errors found. I ate around 6:00 PM. I watched TV until about 9:00 PM, from which I rechecked the forums. I made several other fixes, then finalized the spreadsheet, except character movements, as well as text and foreground scene changes. I began making the scenery, finishing the fox, sky, city, city roads, highway, and 2 layers of hills starting the third and final layers of hills. There's 4 layers of mountains then 8 layers for clouds yet to do. Next would be the character, foregrounds, and other special effects. Forming the final spreadsheet, writing the program, and finalizing it remain after that. It's just around the 4000 frames mark. See news item #1 for more details. While working on the scenery, I checked over the GIMP forums. I'm still at FF1 world - 104% true speed (105% for MP3 player). I'm now only getting minor signs of getting bored of it. I also discovered a math error in my 2D game's lighting - see news item #5 for more details on it. Woke up at 3:05 PM and went to bed at 4:39 AM.
Oct 20: Yeehah! One more dream recalled, a travel dream that uses 3/4 of a page in details. I first checked E-mail and forums. I then finished the hills layers. My dad got me at getting twigs for Halloween decorations. I was very cold and my right ear was starting to burn. I was only out for 20 minutes when it felt like 5 or 10°F. I ate afterwards watching TV. I watched something about rats while eating. I continued on with my super animation. I finished all hills, all mountains, and got a brief start on the first layer of clouds. I want the scene to be of decent quality. The mountains and hills use up about 45 or so colors alone. With the fog, highways, city, clouds, sky, character, and other foreground stuff, I could end up with just over 100 colors. The pound stomp shockwave effect, puffs, and especially the flash attack could cause me to have more colors than GIF supports. Any more detail and I might just end up passing the limit. Duplicate colors, however, are not taken out. FF1 World still plays at 104% true speed. Even after the 46-day record (and 107), I've had Battle Zone stuck in my head since, especially the record-liked area. Even though FF1 World has played for 4 weeks straight, the song played in my mind (and rarely hummed), is Battle Zone, almost always the record-liked spot. Woke up at 3:09 PM and went to bed at 5:05 AM.
Oct 22: Goodie! Although I recalled a dream, it actually occurred the first time I woke up, around 10:00 AM. I went to bed again and woke up around 3:00 PM. I was still tired and went back to sleep. I finally woke up after this. Only the first time I woke up was a dream recalled, a water one. I first checked the forums and E-mail. I then made enhancements to the clouds for my animation but ate after redoing the first and only cloud. I finished all clouds and began the textual parts. The entire story is done, except saving the images. The speed and height indicators have yet to be done. I have the character and foreground scenery yet. It's now going into the fourth day at this and it may be the end of tomorrow or the start of the day after that I finish. I'm still at FF1 World, now going into week 4. I'm still at 104% true speed though. Woke up at 6:01 PM and went to bed at 8:07 AM.
Oct 23: Yeehah! That's the third dream in a row! This one takes a whole page in detail and took 50 minutes to write the details. I first worked on flooring stuff - parents. I ate watching a lot of TV afterwards. I watched TV until 2:00 AM. From there, I checked the forums and E-mail then continued my cartoon. All text is done except notations in the spreadsheet. The character is also done, both the lit and unlit versions and both the actual images and spreadsheet details. The lit versions are very nice-looking. The "stand right" and "fall-glide right" ones are the two best-looking, the latter being the best of them all. I, still, however, have the foreground scenes and special effect decals. There's also the programming and finalization as well. I'm also getting intentions on redoing the "The Spinning Board" animation. See news item #7 for details. I'm still at FF1 World, 104% on my computer, 110% on my MP3 player. I plan on updating my website after the cartoon and remaking the "The Spinning Board" animation beforehand if there's enough time before the start of November. Woke up at 6:31 PM and went to bed at 10:19 AM.
Oct 24: Four dreams in a row - this one takes a good 1 1/2 pages in detail and is rather weird. Man, if this keeps up, I'm almost guarenteed to have 500 by the end of the year. I first ate, but Hamburger Helper without the Hamburger as I've always done with such meals. I checked the forums and E-mail then resumed working on my cartoon. The first 21 of the foreground scenes are done. I, especially, like the very start. I ate pizza afterwards as Hamburger Helper is not filling. I continued on with my cartoon. I reached the point in the foreground scenes where the speed blast's effects are involved. Having not heard back from Philips about the monitor replacement stuff, I contacted them and the shipping details stuff was not present (on their end) and they had to recheck details - another day's wait. I returned to the cartoon finishing the speed blast effect and left off on the first of the impact shockwaves. I may be at it all day tomorrow and even the next day. I'm more around 2/3 done with the foreground layers from what I can tell. I also ran a test to find out how many colors I had - there were much fewer than expected - 83 instead of my thought of 110. The flash attack jacks it to double that. Although 90 spare colors seems like a bit much, the special effects use them up quickly (especially the shockwaves). Other foreground objects (race stuff) also add to it. Woke up at 9:15 PM and went to bed at 1:13 PM.
Oct 25: Rats! No dream recalled - back to zero in a row. I first checked E-mail and forums then ate watching TV. I returned to the forums and tried starting on my cartoon but was cold. I spent 30 minutes trying to warm up via the heat duct. Once warmed up, I didn't get as much progress as I hoped for. I kept making mistakes due to missing details. I had major confusion lasting an hour, all occurring at the moment the character moves to the right after jumping up the cliff just before the bounce stomp is used. I cleared the confusion an hour later and continued on to the point where the character lands and jumps. When starting the shockwave part, I encountered a problem. The details of the shockwave have not been established or standardized. It seems to be the speed of impact raised to the two-thirds power to get the height in pixels of the first frame. The spreadsheet screenshots in news item #1 explain the after effects. This means I have an error in my "Mountain Fun" animation I did very recently. While all this happened, there was a debate/argument going on on HWW about odds and returns. Details are in news item #11. Combined with mistakes I made in the animation and this, I didn't have that good of a day. I'm still at FF1 World with the same speeds as yesterday - 104% on my computer, 110% on my MP3 player. Woke up at 10:50 PM and went to bed at 2:55 PM.
Oct 26: A video game dream involving the classic Bubsy in the forest level with the fast falling again. That's like the third time I got this theme in a short time span.... What does it mean? I finally finished the foreground scenes making a few corrections to previous ones. I also finalized the spreadsheet. I've not only got the programming and finalization left. 3600 base frames is a lot to test. Due to taking much longer than I was hoping for, almost twice as long, I'm not able to get the remake of the "The Spinning Board" animation. After that is updating my website, before November starts. I'm still at FF1 World at the same 104 and 110% true speeds and it now passed 37,500 plays in Winamp. Woke up at 1:32 AM and went to bed at 4:50 PM.
Oct 27: What's with recalling dreams so much during the last month? Yep, I recalled another dream. 500 by the end of the year now seems a certainty. I watched Assignment Discovery while writing details. I checked the forums and E-mail then ate watching more TV. I spent almost the entire day working on that cartoon and found a major error in the spreadsheet and the result took nearly 2 hours to fix due to unexpected mistakes and inconsistancies in the spreadsheet. I fixed the mistakes and soon ran into 2 more mistakes. First, the shadow from the float run was wrong and offset as well as too short. This was an easy fix though. The other was that the wall being crashed into after the speed blast is just used was incorrectly done. A faulty shadow and incorrect lighting. That fix took a while longer. I soon got to writing the program in Gamestudio which took 2 hours. It was only until 7:00 PM that I actually got to see the end result. I spent 2 2/3 hours fixing bugs and testing it. It was as decent as I envisioned it to be, even a bit better. I also ran across another limitation of Gamestudio. Truetype fonts (TTF files) are rendered in Gamestudio as fully anti-aliased. This is dangerous for animated GIFs and there wasn't a way to disable it. The fixed width font system wouldn't work so I had to study the font's numbers in greater depth and write a complex algorithm to work around the anti-aliasing issue. It took 1 1/2 hours to work around this. I finally got around to extensively testing it. Testing went well and I began compilation. It was taking way too long as GIMP is super slow, and I was too tired to continue on with finishing it. 770 MB in 3633 BMPs - Yow! I should have 20,000 BMP files on my hard drive now. I'm still at FF1 World, 33 days straight. I only have minor signs of getting bored of it. But then again, it could be that I haven't changed the speed in a few days. I have a feeling that the animation will be too much for my hardware. It's really pushing it, even 1/3 of the way in. There's swapfile as well so that could help. Woke up at 3:43 AM and went to bed at 10:57 PM.
Oct 28: Man! Another recalled dream!? This one's a half page in detail. I first started the batch conversion. Since I figured it'd take over 2 hours, and that I needed to go to the grocery store, I took a shower while the conversion was running. Although by then, 2/3 of the frames have been done, it was taking much longer to process a frame. I left for the grocery store and went first to get a freezer from a restaurant that was closing down. I arrived at the grocery store, the Wal-Mart Supercenter. I first got CD-R's as I was almost out. I got my groceries then left for Domino's arriving home afterwards. I quickly put the frozen and refrigerated stuff away then ate my pizza from Domino's. As always, I degrease it... using 3/4 of a roll of paper towels, 2/3 of a roll before I even start eating. This is normal that I do this. Why? If you haven't noticed the "fat removal trick" in the tips and tricks category on the main index, you'd see why. Such pizzas often come with a large amount of it and by using paper towels until very little comes up, the amount of fat can be radically reduced, to even 5% of what it was, more the longer you keep at it and the speed increases when the pizza is warm. I do this every time I eat pizzas. The box may say 30% of the daily value per serving (and there's 6 servings in the box), but by using this trick, it drops to a much less value. I've used this trick for several years, probably 4 years by now. Anyway, when I returned home, the conversions were done. I got around to optimizing my animated GIF. That went okay, but it wouldn't let me close the old window as GIMP crashed. Another 2 2/3 hours wait for another conversion and I was back in business this time saving backup XCF's. I soon encountered an error in the alignment of the foreground layers and 2 other alignment errors. I fixed them taking 90 minutes to do so, and spent another 2 2/3 hours at the conversion again. I soon found more mistakes and I was tired of having to wait for so long. So, I figured a way around it. The changes only affect a few frames, only 5 adjustments were needed. I set timings as needed and made two other optimizations. One final GIF, and nearly 400 MB's worth of XCF backups, the first GIF had too many colors and colors were altered unexpectedly as a result. I had to get rid of the natural half-transparent flash attack flash and use an all-white flash attack instead which uses 1 color, a color already in use (hint: the starting line). This gets rid of the transparency and the color changes made seriously jacked up the color count. I reprocessed and replaced 3 frames. I finalized it and released it and checked the forums and E-mail. While waiting for the conversions, I took out Sonic and started drowning the fox again. I only did that since nothing was on on TV. At the end, I even created a fox remover for Mushroom Hill Zone Act 2, after recalling an amazing bug with the flying battery blimp after defeating the boss. I posted my design plan on my website as well, based off the concept of my first successful version. I'm still at FF1 World and still at 104% true speed on my computer. My MP3 player, however, now has 90% true speed. Woke up at 8:51 AM and went to bed at 4:43 AM.
Oct 29: The streak of recalling dreams continues on. I could have 500 by the end of November if this keeps up! I first checked my E-mail and the forums. I started the headlines of my blog completing 2 of them. I ate at this point trying a new brand of pizza. I spent nearly 3 times longer than usual degreasing it. Even after using two paper towels on top of each other 4 singles and high pressure, it was still soaking through unusually well. That's even worse than pizzas from Pizza Hut, Domino's, and other such places. That had a big negative affector. The crust was okay, an affector of 1 (no change) and the sauce was decent. The cheese has an affector of -2.8 and the sauce at +1.9. The compatibility offset is -1.5 to Freschetta and -1.8 to Digiorno. Freschetta has the best sauce and crust while Digiorno has the best cheese. Digiorno has the wheat crust version which is healthier which adds a positive affector. The difference in compatibility between the two is a minscule 1.2. When done eating, I continued on with adding the headlines to my blog as well as browsing the HWW and GIMPTalk forums. Barely 11 hours awake, in contrast to my last 2 days, I went to bed unexpectedly early. Woke up at 1:24 PM and went to bed at 12:58 AM.
Oct 30: Rats! No dream this time. I first checked E-mail and forums. With still no word on the monitor replacement details, I contacted Philips again. By this time, FF1 World passed 40,000 plays in Winamp. They have supposedly sent the replacement. Now, however long that'll take, who knows. But, if another week goes by and I don't get it (by Friday), it's round 4 at the support. I finished the headlines to what is known. I added the Mushroom Hill fox remover details and 2 of 9 dreams. I've got at least 7 more dreams and all daily entries. I spent too much time on HWW. Indirectly, I was also updating my blog at the same time. I need to finish before November starts otherwise the blog would move to November instead. It's Halloween, my third-favorite holiday and is liked slightly better than neutral (1.2). November is next, a neutral-luck month. I'm still going at FF1 World. My computer now plays at 107% true speed and my MP3 player is now at 73 1/3% true speed. Woke up at 10:37 AM and went to bed at 12:24 AM.
Oct 31: A very vaguely recalled dream was recalled - just 4 measely lines. I first ate watching "A Haunting". I watched 4 episodes of it then went to check my E-mail and only my forums. I continued updating my website finishing 30 mins before the month changed. Only one group of 4 trick-or-treaters came, the same scenario the previous 2 years. I never saw them. When done with the website update, I checked the HWW and GIMPTalk forums. I was too tired to do much more and realized I forgot to crosscheck the news item numbers in the daily entries. I realized it when I went to bed. FF1 World still plays. Computer is now at true speed and MP3 player still at 73 1/3% true speed. Woke up at 10:13 AM and went to bed at 1:26 AM (±2 minutes).



Go to another time:
Earlier than 2005: Earlier than Nov 2004 | Nov 2004 (and late Oct 2004) | Dec 2004

Year 2005: January | February | March | April | May | June | July | August | September | October | November | December

Year 2006: January | February | March | April | May | June | July | August | September | October | November | December

Year 2007: January | February | March | April | May | June | July | August | September | October | November | December

Year 2008: January | February | March | April | May | June | July | August | September | October | November | December

Year 2009: January | February | March | April | May | June | July | August | September | October | November | December

Year 2010: January | February | March | April | May | June | July | August | September | October | November | December

Year 2011: January | February | March | April | May | June | July | August | September | October | November | December

Year 2012: Yet to arrive....

The news entry headlines in bold are color-coded for a reason. They tell how big the event was for the headlines and how good my day was for the daily entries. From May 2006 onward, a slightly different system is used, of which has triple the precision. For dates prior to this that use color coding, from the white color, skipping two to red and so on is used. Here's what each color represents (for updates to before May of 2006):

Extremely major event / Very good day
Major event / Good day
Somewhat major event / Somewhat good day
Normal event / Average day
Somewhat minor event / Somewhat bad day
Minor event / Bad day
Extremely minor event / Very bad day


Footnotes:
None