Jul 29, 2009

Jul 29, 2009: Platform Masters - my new, simple, all-in-one game, has more progress than TSO... in 3 weeks! New bonuses with TSO but its too early for 2.5D. SNES obtained - got 6 games. Old computer died - got new one and a video. 42 new dreams recalled, 3 in one night, and dream journal redesign underway. Major construction going on and still ongoing. Better fireworks show, despite less variety. Is Firefox itself the culprit to my download troubles? JPG quality enhanced. Book should be out any moment now (big notice on main index will be posted when it happens). Various other ideas for current and future events.

#1 Platform Masters - my new, simple, all-in-one game: Okay, I've got 4 or 5 games in the works already, but another one!? This game is different in many regards. It all started on March 27, when my computer was down - it was active in my mind. It uses all common elements in video games - lives, health, levels, worlds, bosses, etc.. If you recall George Game 13, think of that game. Platform Masters is based heavily on GG13, but, instead of 3D, it's 2D. I've noticed that, with TSO, I was focusing far too much on the graphics - tone them down! That's what spawned my motivation for starting Platform Masters. In addition, the "getting high" videos for Jumping Flash 2 started it. In JF2, there are clouds drifting above. I had views where I was looking straight across as I kept on climbing higher and higher, until I maxed the height... from ground level (I don't have a GameShark for the PS1 so I can't use such codes for it). This inspired me to make 3D-like clouds. It turns out that, after fixing all the bugs, I got quality far beyond what I was expecting (and I mean far beyond). When you've got a flat ground rendered entirely in 3D, and the camera was looking straight across, this is the exact same effect (or so close to it that you can't tell the difference).

First simulation of Platform Masters' looks

These are simulation screenshots as to how the game will look. Of course, some things will change. This area is known as "Keveran Forest". It is the first world that uses bottomless pits and, knowing the screenshot, it's world 6 (of now 16). The top screenshot shows what things are like during actual game play and the bottom screenshot shows what things are like upon completing a level, and how the score really adds up.

Originally, I intended on having 15 worlds with 21 levels each (10 required regular ones, 1 required boss level, 7 optional jumbo levels, and 3 optional supersize levels). Now I intend on having 16 worlds with just 16 levels in each (10 required regular levels, 1 required boss level, 4 optional jumbo levels, and 1 optional supersize level). A typical level is expected to take about a minute to complete, so short that checkpoints are not needed. Jumbo levels are 4 times bigger but expected to take 3 minutes to complete (extra space for shortcut opportunities and treasures like extra lives) and supersize levels are 8 times bigger and are expected to take 5 minutes to complete - they have much more open space and more shortcut opportunities. The time limit is based on difficulty 6 without the use of shortcuts. Difficulty 0 puts this at 4 minutes, 12 minutes, and 20 minutes respectively. The largest of levels may have 1500 seconds for completion at difficulty 0.

Strangely enough, I lost my motive for programming it right as I'm at the point where I can actually play the game. The biggest problem with programming, for a bit, was malloc. I couldn't get it to work for some unknown reason as I either got a bunch of errors or a bunch of warnings, no matter how I used the * and & icons. I got closer when I had the pointer in a struct, but after spending hours of converting to using malloc, I found that the whole program was glitching. Without even touching the code for window creation or windowclass creation, errors were occurring here. I'm thinking that it may be because I used an array of structs, each with the pointer in it. I just had an all-black display. Once I converted back to using global arrays, all the problems went away. I've pretty much given up on malloc. I'll need to find another alternative for dynamic memory allocation. Then again, when you think about it, when was the first time a 64 MB video card was released? I'm getting 2000 as the year... 9 years ago. This means that mid-range systems from 2003 and budget systems from 2006 should also have such a video card - a wide range of computers (considering that if I have the game released in 2011 (there's no guarantee of a release date), even more powerful computers are available. Thus, if I make wise use of the alloting of the global arrays, I should be fine with the global array method. The only down side is that the program will constantly be using that much memory. It also limits quality settings, of which I was hoping for quite a bit with the clouds. Instead of the high quality clouds that I use, I could have a medium-quality and a low-quality version. While I could still do this, the memory is still being hogged and there's no benefit for the draw count (yes, I still have that limitation too).

I've made a lot of progress already: medium-density clouds, low-density clouds, the character (the important stuff for playing levels), the platforms (a new method is available to open up more sizes and colors and use a lot less memory but I don't know if AlphaBlend works with palettized images), the scenery for Keveran Forest, most of the scenery for Keveran Reef (world 7, the ocean), and most of the scenery for world 8 (theme is classified outside it being at night) are the major elements done (those that take the most time). Some of the foreground objects (sprites) are done - hazards (spikes, cacti, etc.) and most on-screen info stuff (the 3D, colored text along with their icons - some still need icons) - are done, but enemies (such as the snapping, carniverous flower) are not done. I currently need to figure out how to build a worthy level creator so I can make levels easily and efficiently. I can make levels right now, but manually inserting positions is a bit tedious since it's hard to visualize. The game uses "blocks" and "coordinate units" as the units. CU is 11/75 of a foot, 1.76 inches, or 4.4704 centimeters. A block is simply 32 CU (after the texture of the platforms), which is ~4.693 feet (56.32 inches), ~143.05 centimeters. The overworld, used to move between worlds and other special areas, has a great deal of progress as well, though it isn't completely finished. I still have the mountains and decals (forests, lakes, cities, etc.) to add yet.

I have written a document detailing the game's design, for what isn't classified. The development history section is the most frequently updated. There are also plenty of YouTube videos as well, showing you how the clouds behave, along with the test level, and how the game is intended to look using two simulation screenshots (shown directly above).

#2 TSO - it's too early for 2.5D and Platform Masters bonuses: In my previous update, I mentioned of needing to convert TSO into a 2.5D format - render things in 3D but make it play like a 2D game. I'm jumping ahead of myself. 2.5D might be worthy for a version 5 but certainly not version 3 and probably not even version 4. I still have the draw count issue to deal with - 2000 for a 2005-era computer is extremely low when today's computers are far, far faster. 3D games show a lot of detail, but this game doesn't even get close. I really need something faster than AlphaBlend and it seems like SDL is probably my only option.

Trouble is, I don't like SDL's license terms for commercial use. The game may only be a 15 MB download, but having to include the source and stuff could easily triple this (I don't know the actual download size). If I could use static library linking and only need to post a URL to SDL in the game's about section, I won't have a problem with that and I would easily take up using SDL not only for TSO but also Platform Masters. There's still dynamic memory allocation too that's problematic and TSO could use it, since I'm likely to go over 64 MB memory usage, going to 128 instead (you don't find 96 MB video cards - it's always a power of 2). If I could get the drawing speed I see in GameStudio with my own drawing function (I know how to do this, but I just don't get the speed benefits that even the slow DrawDibDraw had when I used it). I'd need hardware acceleration for this and that's where SDL comes in, I hope. Either that, or it's DirectX.

When I was working on Platform Masters, I also, indirectly, made progress toward TSO and other future projects as well (such as my 2D RPG). Take the text-drawing function. Word wrap with TSO wasn't working properly under all conditions. I needed it redesigned so that it did work under all conditions, or at least as many as I could think of. I had a 10-test process that the code needed to work with. If it didn't pass the 8 major ones (which are individualized tests) and the 2 combos (which combine many of the 8 tests into one (the latter contains all of them)), I'd have to make a change and try again. It took 9 hours to do it and it wasn't until I thought of logging the point at which the end of the line is reached that I've finally solved the puzzle. Now I just need to copy the code over to TSO and TSO will be fixed (after reformatting the calls, since the function header has changed).

My number-formatting function was also faulty. Negative numbers didn't work properly with TSO. When they dropped to -10,000 and below, the point where commas would be used, the thousands data wasn't being found. The function also didn't have any support for integer-based numbers (like the score would be). This entire function is now fully complete, clear out into the quintillions (the limit of a 64-bit integer). Yep - just copy the code over and readjust the content in the calls and TSO has more progress.

Another bonus is a new texturing method. I discovered this while doing the lava hazard for Platform Masters. Without it, the lava was grade C+, barely acceptable for TSO. With it, the lava is now grade A-, easily acceptable. There's a huge difference in quality. This should, in a way, help enhance the texture for the forests, grass, and other such things.

One further bonus, a big one, is a major enhancement to the Urusu Valley area. Originally, it was intended to be rather plain-looking with little in the way of textures beyond basic grass, water, and forest. Now I can throw in farm fields and cities at no extra memory cost. The method involves the very same one used with the clouds in Platform Masters. This area is expected to take about 8 MB of memory, depending on how I optimize it and what scaling/size I use. Currently, it's now planned to be 50 feet per pixel though I may use 60 or 55 instead. Given that 1920x1200 is a fairly common high-end resolution in today's computer shops, the 50-foot spacing is about perfect. Pixel duplication isn't seen until more than 630 pixels below the screen's bottom (for a resolution's height of 1260). Urusu Valley, essentially, starts at a scaling of 314.627 (for 31,462.7 feet, or 5.959 miles) and ends at a scaling of 530.627 (for 53,062.7 feet, or 10.05 miles). The first pixel would be based on a scaling of 315 and the last would be 530.5. This means the image must be 432 pixels tall. The width, 90 miles or 475,200 feet, would be 9504 pixels. For optimization, the image is likely to get segmented into 5 or so segments and cropped. It means more work, but it pays off with reduced memory usage. Without this, Urusu Valley would need 12,317,184 bytes (11.75 MB) of memory, rather high.

What about this duplicate pixel thing? That comes from the way the texture "expands" as it heads toward the bottom of the screen. When near the horizon, a perfectly flat field appears to form sharp angles with the horizon. With increased height, the angles get closer to being a right angle. The clouds in Platform Masters have 3 duplicate pixels when right next to them (it's hard to see due to the smooth gradients used). See the last few paragraphs for the math behind it, if you're curious to know (it's simple stuff, really).

The method now to be used with Urusu Valley can also be used in the foreground as well, but needs help. The method with Urusu Valley, as is, means needing so much memory for just one lake, that you'll have to have a 64-bit system, even with optimization. Just Lake Windias alone, being 10 2/3 miles across, needs 3 MB of memory per row of pixels! Given the kind of spacing needed, an inch out per row, you're looking at 30 GB of memory for just that lake (maybe 1.5 GB with optimization). It's 629 1/4 times closer at the closest edge so 50 feet divided by 629 1/4 gives about 0.9535 of an inch for the spacing (1 for simplicity). My solution comes from how I did the clouds in Platform Masters - make it a tileable texture. However, the lakes are not perfect squares in shape. The fix is to merge the vertex-mapping system with the texture-tiling system. The vertices will allow any shape to be used. By mapping the tileable texture in within the bounds set by the vertices, the memory usage may only be about 2 MB, easily acceptable (though it depends on the texture size (width mainly) and quality that I decide to use). The grass, desert sand, rivers, roads, and other such things could also benefit from doing it this.

Another optimization method, of which I could also use on Platform Masters' clouds, is using the intended quality but, for distances 1.25, 1.6, and double (or just double) that of the base, I could, instead, just increase the spacing. This is kind of what I currently do with TSO's current method for drawing the ground, only it's a crude method used solely to reduce the draw count significantly. Regardless of the spacing in my new method, the draw count is essentially unchanged and, for 1024x768, will always be up to 384 (ignoring any tiling). This method allows for maximal quality and a realistic shape put in one without affecting the draw count (it may be slightly improved), and improves on memory usage as well. Of course, this method has yet to be tested.

These methods, though it may seem more drawing intensive, will actually reduce the draw count. This is because special effects such as fog and texture-blending are preprocessed in the image itself (fog is processed upon loading) so that frees up about 400 draws. Then again, with SDL, I don't know how effective it'll be since I don't know the draw count limitation with it.

One new idea I have, for the landscape, is to have better-designed slopes. Originally, I was using steps of 5° and alternating between two different slope grades to get a different type of slope (such as alternating 15's and 20's to get a 17.5). You might be thinking of the vertex-based system I had planned. Well, it's the draw count again (either that, or I don't recall how I worked it out, more like the case of not recalling the details)). Nope. Now I'm thinking of using slopes based on frequency of occurrence. Most of the land is quite level so slopes of 1° change (basically 1 pixel on something 60 pixels wide) are present until about 15 or 20°. From 15 or 20 to 45 or 60°, since these are less common, but still fairly common (especially in the hills and mountains), they step by 2.5°. From 45 or 60 until the steepest possible, they step by the now-current 5°. This mostly eliminates the need to alternate the landscape. For more speed, having not thought of it before, I need a quick look-up table (basically, it's an array of values) to determine what the intensity of lighting is for that slope. This avoids the need of using CPU-intensive atan. Another case of this useful look-up table (a different one), would be figuring out the sine and cosine values of each slope. Only a few slope types are used (69 or 81 total) so this can improve the efficiency (though cost a few more bytes of memory, 648 bytes at most per table). Gliding, as a side note, could be the same way, though I don't yet know.

Another new idea I have is to make the hills and mountains in the background individualized. That is, instead of the flat bands of them, I'd have each hill or mountain separate, with their own separate scaling. For this, I'll need to build a map. I can reuse quite a bit of what I've done right now. At least this way, the grass behind the hills becomes much harder to see and the scene looks and behaves more realistic. It'll mean a few more draws though, but it shouldn't affect memory usage, at least not much. If I do it right, I may actually save memory, sparing it for the cirrostratus clouds! Speaking of clouds, I may use clouds like Platform Masters for the cumulus humilis clouds, but I think I'll stick to my original plan (but make the clouds individualized instead to make the greatest use of memory and have the most realism along with it).

As you can tell, there's a lot of new changes for TSO from working on Platform Masters, for the better. Now, if only I can resolve dynamic memory allocation and the draw count limitation, I'll be cruising along!

One common, new vision I've been having is thanks to Speeders. Oh, a hint already huh? In the background, there is a cop car with the lights flashing (uh oh!) on the far left. A little to the right and much closer is a speed limit sign (hmm), which states that the speed limit is 65 mph (it's a highway ya know, expected, until the city is near where it lowers some). My character is either float-running at 214 mph or gliding at 173 mph (give or take some), obviously speeding. I find it kind of funny in a way. The cop, however, is actually after a car that is speeding.

As to how the duplicate pixels come about, let's consider the 1920x1200 resolution, to help you understand. The math behind it is something on the fourth grade level, basic multiplication and division, with no need to know the order of operations. This resolution means that ground occupies 600 pixels of space. First, we need to find the height, based on the point where the closest part of Urusu Valley is at the very bottom of the screen. The formula, used frequently with my artwork and especially animation is this:

Pixels*Scaling = CU;

600 pixels times a scaling of 315 means the distance from the horizon (ignoring depth) to where it's at is 189,000 CU (27,720 feet or 5 1/4 miles). That's the height the "camera" must below in order for that part to be on screen (exactly this and it'll move one pixel off screen). For the next higher further row, you need to know how to find what the scaling would be for a given row and this uses another simple formula, a lesser known one that results from solving for "scaling" (with variable names changed for reference in the situation):

Scaling = HeightCU/Row;

"Row" is originally "pixels" and "HeightCU" is originally "CU". The row directly below the horizon is row 1. This is because closer objects are drawn in front of the more distant objects and since this whole row covers the true horizon until ever-so-slightly a bit less 1 pixel below that, it's considered as row 1 in the calculations. This would make the scaling of that part 189,000, that of HeightCU. The row below, row 2 (spanning from 1 <= x < 2 in the mathematical sense), would use 94,500. The bottommost row is actually row 600 (note that 189,000/600 = 315). The one above it is 599. 189,000/599 is about 315.5259. The row at 315.5 will, thus, be drawn here. Now let's consider the highest commercially available resolution (that I'm aware of), 2560x1600. This means 800 is used instead. Running the same process, you get 252,000 for HeightCU, and row 799 would give 315.3942 as the scaling. It hasn't crossed 315.5 so the row of pixels for 315 will be duplicated here (since the closest parts are drawn in front (you draw back to front), the result is actually rounded down). Row 798, being 315.7895, will have the 315.5 row drawn (the 316 row follows then two 316.5's).

#3 Old computer died, back better than before: Maybe that squealing power supply episode was a warning of something more disastrous to come? Only a week later, on March 25, 2009, my old computer died. When I turned it on, I could hear the fans running, but nothing was showing up on the monitor. I didn't hear the familiar beep from POST either (POST stands for "power on self test"). If there was a hardware problem, then you'd hear something other than a short, single beep at the start. The pattern varies with the BIOS used. The lack of beeps indicates that either the motherboard or the processor died.

For about 2 years or so, I've been wanting to upgrade my processor because it was supposedly the slow component of my computer. However, to do that, I needed to upgrade the motherboard as well. The dual core processors I was looking into don't support DDR memory so I also had to get new memory as well. DDR2 is what is used with such processors. I was surprised to find that memory was so dirt cheap. I can max the memory out on a 32-bit OS for just $35! I thought 3 GB was going to cost $70, but wow! Since I didn't really like the design of my old case (rather small, and the power supply was in a spot directly above the processor which limited the availability of fans).

So, what did I get then? I got the E8500 processor (Intel Core 2 Duo, 3.16 GHz, 6 MB L2 cache). I got G. Skill DDR2 800 MHz RAM, 4 GB (the maximum the motherboard can take, but I only get to use 3 GB of it - a 32-bit OS limitation (I was originally thinking 3 1/4 GB, but never the whole 4 (unless I had a 64-bit OS, which has weak software support))). The G. Skill brand is new to me. The motherboard I got is G31M-ES2L, the model of it anyway. In addition, I also got an "Arctic Cooling" CPU fan. I've heard that stock fans are inferior - they do the job, but aren't as effective. I also got a new case as well, some Cooler Master brand item. Its usage is rather awkward when it comes to installing hardware.

On March 31, I spent pretty much all day putting it together, recording a lengthy video with it. I thought I was going to have to go through about 16 batteries for the length of video I had. Fortunately, I picked up those "Ultimate Lithium" batteries from Energizer. They advertise that they last up to 8 times longer. They should advertise it as lasting at least 8 times longer instead because, in my case, they're going on for 10 times longer and my camera still showed 3 of 3 marks (they actually went on for about about 13 times longer). Normally, I'd have problems after 15 minutes of video, but instead, I've got 2 hours of video! I've got another take on that "Hamburger Helper without hamburger preparation", another 20 minutes, and a few other photos along with it. When they'll run out, I have no idea yet....

Anyway, after I hooked up everything, I encountered a big problem. I hadn't run backup in 2 months and I really needed to. Trouble is, with a new motherboard, processor, etc., Windows was completely unable to start, even into safe mode. If I wiped out my hard drive, I wouldn't, then, be able to run backup at all. Luckily, I had spare hard drives lying around (I intended on selling them too!). I wiped one of them out, formatted it, installed Windows and bare essentials like Winzip (of which was backed up on other disks (not the hard drives, but CD's and my external hard drive)). I ran backup, put them on my external hard drive, put in my main 250 GB hard drive, installed everything, including nonessentials, then put my back up on, and pretty much went from there.

My new computer has numerous advantages. When it came to processing videos, I noticed a big difference in speed. Normally, for 640x480, I'd see 20 fps as the processing rate in Virtual Dub. I was now getting 50 fps, meaning that my new computer is 2 1/2 times faster. Everything loads up faster as well. A long time later, over a month into it, I got to processing songs, that of Gnasty's Loot stage in Spyro 1 was the first. I'm used to the 16 KHz speed (for the resulting MP3) going at 50x CPU/play speed, but now I'm getting 120! That actually shocked me when I saw that.

Another big advantage comes with my software development. Whenever I get infinite loops, I don't have to wait 5 minutes for the program to finally close because everything becomes so slow. This is where the dual core processor comes in handy. Instead of 5 minutes to close a program in an infinite loop (GIMP included), it's now 3 seconds, the normal amount of time it takes for me to switch to Windows Task Manager, choose the program at fault, click "end process", and click the "OK" button. It becomes 5 to 10 seconds if I have to search for the culprit (this usually applies with my own programs).

#4 SNES obtained - started playing after 6 years without: I've been wanting it for about a year and a half by now, for YouTube, but the SNES is who knows where. I haven't seen it for 6 years roughly. So, around May 5 (give or take a day - see the daily entries for a more precise date), I got one off Ebay. I originally got it so I can test a game for my sister's birthday. I wanted to finish my site's update beforehand, but I wasn't quick enough and I had too little motive for it. I got several games for the SNES. Interestingly enough, the original Bubsy wasn't the first one I got, though it was later obtained not too long after. What are the games? I've listed them below with four aspects mentioned.

Difficulty: how easy or hard the game was. The harder the game upon the first play, the lower the score and the more likely I needed to consult GameFAQs to get help with.
Game play: how well I liked the game play. The more I liked the game play, the higher the score.
Compatibility: this, if you know me, should be obvious - it's how much I liked the game. Obviously, the higher, the better. I mention a low-precision compatibility value (the true average) after it.
Addictiveness: the more addicting the game is, the more likely I am to log hundreds or thousands of hours on it (now, what do you suppose Bubsy is going to be?).

Game #1: Super Metroid

Difficulty:        Hard
Game play:         High
Compatibility:     Above neutral (about 2.5)
Addictiveness:     Low

The first game I was involved with was Super Metroid. I rented this game, but got stuck (like I am with the original). I was supposedly to be exploring Norfair, but I didn't have the Varia Suit (and was unaware of its existence). Even when I played it again, I still got stuck in the very same spot. I often did sequence skipping, getting power bombs, and back tracking when you normally wouldn't be able to (thanks to bomb jumping, of which I recalled doing when I rented the game). I had to use GameFAQs to get unstuck. Things went well again, but more problems arose. After having explored Norfair, getting a game over from "my favorite boss" because I didn't know how to defeat it, I ended up getting stuck again, needing the speed booster. Upon getting it, I was able to go through the rest of the game without any further problems outside two tough bosses. That 4-eyed one in Norfair with the strong lava below, my favorite boss, was one case I got 2 game overs with because I didn't know how to defeat it. Later on, in Maridia, the lobster-like boss was also troublesome because it was so fast at the end and it took so many hits to defeat. I got game over 7 times from this thing and barely made it on the 8th attempt. That boss was the hardest one of all, and yet, I had 1099 in energy with 200 in reserve. I finally completed the game for the first time in 13 1/3 hours or so. I then went through the whole game again, clearing it in just under 5 hours without a game over occurring. From this, I recorded a full game video for YouTube, where I clear it in 3.4 hours. In this video, you can see just how much trouble I have with the lobster-like boss in Maridia. Finally, on the fourth run, I got it to within just 2 1/4 hours.

Game #2: Final Fantasy Mystic Quest

Difficulty:        Medium-hard
Game play:         Medium-low
Compatibility:     Semi-favorite (about 6)
Addictiveness:     Medium

The second game I got was Final Fantasy Mystic Quest. Knowing the music with the Final Fantasy series, I thought this game would rate high. It's not quite all that decent, but it does have it's up sides. I found this game a bit challenging, mostly because of the fact you only have 2 characters. If one, say, gets petrified (of which protection against it comes very late), and the enemy goes before the heal potion is used to fix it, petrifying that character, I'm doomed. With other FF games, this problem is not present. I was only level 36 when I finished. I'm thinking of using the Level Forest to reach level 50 or something. In the duration it takes me to reach level 99 in FF7 (3 days), I could probably reach level 50 this way. As of this update, I only got to 21....

Game #3: Super Mario World

Difficulty:        Very easy
Game play:         Medium-high
Compatibility:     Above neutral (about 2.5)
Addictiveness:     Low

The third game I got was Super Mario World. I actually owned this one beforehand, but, like with the old SNES, it's at who knows what location. Yep, I got that video of that bug I wanted, maxing the score in a short time span (8 minutes in my case). I also got to see a new set of rather difficult levels, where I see "special" formed from animated star textures, from the center of those in star road. I never knew of their existence until then. I still have another 4 or 5 exits to find though before all 96 levels are completed....

Game #4: Bubsy

Difficulty:        Easy-medium
Game play:         Extreme
Compatibility:     High favorite (about 150)
Addictiveness:     Extreme

The original SNES version of Bubsy was the fourth game. When I got it, I recorded a video right away, so I can experience my first ever play of the SNES version. During this, I take note of any differences I found. Many of the bugs are still present, though a few appear to be fixed. When I played, I hadn't played the game in several months so I didn't do quite as well. Eventually, I also got a video on the 80-lives glitch and, discovered on June 14, a new, better, and faster way to reach 32 pixels per frame. I was also focused toward this game's music too. In some ways, I was disappointed, but in others, I was more into it. Here's a comparison between the Genesis and SNES versions (compatibility is based on best known speed, using the real average (as opposed to the spike-weighted average which is often higher):

Song*	Genesis	SNES
Village Zone	Semi-favorite (about 10)	Semi-favorite (about 12)
Carnival**	Neutral (about 1.2)***	Semi-favorite (about 7)
Desert Train	Below neutral (about -2)***	Favorite (about 60)
Desert Zone	Favorite (about 70)	Dislike (about -60)
Out Where the Lake Is	High favorite (about 100)	Favorite (about 80)
Tree Climb	Semi-favorite (about 8)	Neutral (about 1.5)
Space**	Below neutral (about -2)***	Below neutral (about -2)***

Table footnotes:
* I don't know the names of these, so I reference these using my own names or the type of world used.
** This song is unnamed - the theme of the world or area is given instead to distinguish.
*** This song is quite uncertain since I rarely listen to it or vaguely recall it. Given the law of new neutrality, the rating is expected to be at or close to neutral.

While some songs got worst, others got better or remained unchanged. Note the huge change for the worse with Desert Zone, the song with 350,000 to 650,000 life time plays (probably should now be 400,000 to 700,000). It went from a favorite to a high dislike. It's all because of the extremely irritating instrument that, ironically, makes the best part of the song in the Genesis version. But, note that Desert Train went up quite a bit. On average, the music is quite a bit worse - 5.578 for the Genesis versus 3.38 for the SNES for the logarithmic average (aka the geometric mean of each set of numbers).

Game #5: Final Fantasy 4

Difficulty:        Easy
Game play:         Low
Compatibility:     Above neutral (about 3)
Addictiveness:     Medium-high

The fifth game is Final Fantasy 4 (called FF2 on the system, but I'll call FF4 since it's the fourth in the series). At first, I thought I wasn't going to be able to play this game because it wasn't on a system I had. I soon found that the PS1 had it and even the Gamecube, under "FF Chronicals" and "FF Crystal Chronicals". I avoided these because it also included "Chrono Trigger". Seeing the word "trigger" in it, I thought it was a shooter... until GameFAQs told me otherwise (oh how useful GameFAQs is!). I've played FF1, FF2, FF5, FF6, FF7, FF8, FF9 (incomplete), and FF12 of the main series at the time I was considering getting it. Since, when I thought of getting FF4, I was involved with the SNES, I thought I'd get it for the SNES. I did. I was both satisfied and disappointed. The story was good, but the music wasn't all that decent (the music for boss fights is the best). The gameplay is a bit annoying when it comes to menu usage - you can't just press L or R to cycle through characters to view their stats and you have to repeatedly go in and out of the item list just to use a bunch of items. However, I really love the fact you can get more than not just 99 or even 255 of an item, but even thousands. My fourth leveling ground video in YouTube (of my second play-through) shows you what I mean - I have over 700 "cure2" items. In my first play through, I originally was after my usual - high levels early on. I didn't get any videos because I had no idea what there is in the game. I completed it without the need of consulting GameFAQs, which is a plus. I almost had to once though. Since level 31 Dark Knight Cecil had 600 HP upon changing and going back to level 1 (level 31 is unusually high), I thought that 600 was ridiculously high. Thinking of an experiment that results in starting a new game, I found that my system completely erased all the game data out of nowhere (probably from bad or weak contact). I ended up starting a new game anyway. I intended on going to level 50 as Dark Knight Cecil but was starting to get bored of it by 45 so I grabbed a video for YouTube, recorded one of the process that is. After realizing that Cecil still had 600 HP upon the change, all the effort was essentially wasted. After all, it was just an experiment. I then got him all the way into the 50's before resuming the quest - everyone started on 50 and Yang, of which is a very useful character for me, oddly didn't get any more HP after his first level up (from 60 to 61), though his other stats increased.

Game #6: Final Fantasy 6

Difficulty:        Very easy
Game play:         Medium
Compatibility:     Semi-favorite (about 12)
Addictiveness:     High

The sixth and most recent game I got for the SNES is Final Fantasy 6 (called FF3 on the system, but I'll call it FF6 since it's the sixth in the series). This game is actually nothing new to me. It's only a year newer than Bubsy and I've had Bubsy since 1994. It started when I watched two friends struggling with this very easy game (to me, it essentially had no challenge at all, so little that, in the second world (the World of Ruin), upon getting the airship around level 33, I can already go for meteor in that ancient castle without much trouble from the rather powerful enemies probably 10 levels above my characters and without the use of vanish or protective spells (aside from the basic cure and cure 2 spells)). This was during the Lete River and house fire in Thamasa events that I watched them. I was probably about 3 or 4 levels below what they had (below, yes, because I was after making the most of the stat bonuses) and it was still very easy. Late in the game, I just go around casting meteor on everything, doing 9999 damage every time. I was tempted to get FF6 because of meteor and wanting to see the game's true speed (unlike the familiar PS1 version's frame rate problems). The enemy names are sometimes different too, longer. Instead of "chicklip", it's "chickenlip". Instead of "tyrano", it's "tyranosaur". I also found Atma Weapon (the equippable sword, not the enemy) for the first time which, late in the game made quick work of Bracosaur - gone in two turns instead of about 3 or 4). This was useful when it came to leveling Umaro up. Yep, I finished the game this time, defeating Kefka, who really had no challenge either, and I noticed a few differences in the ending too.

I'm now thinking of getting one of those Zelda games. Zelda isn't foreign either - yep, those two friends that I associate FF6 with are involved again. I recall them playing the NES version where I saw hearts in the top right corner and other details at the top. I never got to play it. I'm more familiar with DirectX programming than I am with Zelda, considering that I don't even know how to set up DirectX in a program (I only know the basics, such as what it is (provides direct access to the hardware, often for making processes go faster, one thing my own games could use)). You might be wondering why I haven't gotten into Zelda. It's mostly because I forget about it and when I do remember, I'm not involved with my games. Beyond this, I really don't have any further plans for SNES games.

#5 Fireworks 2009 - better show, little variety: Fireworks - I've been shooting them off for about 18 years by now (since I was 7, using my parents and very simple stuff). When I posted videos of the entire showcase, the intro and the main show, I was surprised to find how others handle them - very unsafely. Take, for example, lighting a bottle rocket. Some YouTube comments state of the user holding it in their hand while lighting it, against that of the warning label. I use rocks for my launch base and design it in a way so that it is structurally stable and strong, to prevent the rockets from going off the sides or going backwards. When it comes to combos, grouping multiple fireworks together to create new effects, I take safety into consideration, as demonstarted in those videos (particularly in part 5). My fast running speed (12 mph when it comes to the really big stuff) helps as well.

Now, if I had a camera that didn't get that pinkish effect (which is only present outdoors - too bad I can't set the sensitivity down (-1 1/3 eV would be good, perhaps even the bottom at -2)), I could do a lot better. In fact, one thing I would like is a camera that records 1024x768 video at the typical 30 fps, or a 640x480 video at 100 fps (maybe 120 or 90).

I had too many jumping jacks and smoke balls but too few fire crackers (particularly the singles). I wished I could've gotten a few others as well, but dad was pushing me because of an unexpected dinner being ready. When I'm not in a rush, I do a lot better.

#6 Got on radio... sort of: In my home town, there's this "Wheeler Dealer" thing going on. While doing construction stuff one day, I heard that they had a coupon for the 20-inch jumbo pizza. I knew I wanted to see if I can get it. I grabbed the phone and began calling. I called, and I called. On try 27, I got a bunch of rings, but no one answered. 40 tries went by and I was still calling. Sometime around call 47 (the number isn't recalled), I got rings again, but no one answered. 50 tries in and still nothing. It was reaching the end of the show around 60 tries in. I was starting to lose interest by 70 tries in. On call number 74, I got rings, again. About 25 rings in, I was on the air. I asked if they still had Nite Train but wasn't specific enough since they had another one. I asked about the 20-inch pizza thing and they still had them. I grabbed it then essentially signed off as usual. I was quick with it, so not to hog up air time (I'm otherwise quite talkative....). It's the second time I've been on the air (the first was one of CNN's viewer polls about an internet police force (yep, I still recall that quite well - you can even find it in my blog somewhere, likely around 2007 (to quickly find it, use control+F to bring up the "find" feature (most, if not all, browsers have it) and search for "internet police force". If no results are found, try another month and repeat the find.).

#7 Super Monkey Ball 1 moon jump at last: Thanks for finding the code! Upon getting the code (and debug), then tweaking it for hover and faster climbs, I got a world tours video of the game just as I had done with Super Monkey Ball 2 about a year or so earlier. As a bonus, I also posted the codes (some modified (using GCNcrypt and my programming knowledge to do this)). It made the game more entertaining. I especially like fiddling around with the city world.

#8 Dream journal reformat - slowly but surely: It's long overdue. The dream journal has dreams with its ratings bars inconsistent. The "travel" and "other" categories are getting overwhelmed. Plus, it's hard to tell noteworthy dreams or record-holders apart from other dreams without having to check the list of noteworthy dreams at the top of each page. The redesign has numerous new features:

25 categories instead of 11. Existing categories have been split into subcategories. The "chases and escapes" category is now added, along with "multicategory" (due to a rather high number of these recently), and others. The video game dreams category has 4 subcategories, for real, fictional, combo, and general video game dreams. The travel one has 5 - neighborhood (? don't recall), local, foreign, fictional, and general. The categories are also better and uniquely arranged. The target category has dreams with similar properties to those in the category above or below it. This way, you can find dreams of similar content to your favorite category.
New headers. Each header has the icon for the dream category followed by the dream number itself (such as "[ ] 8: dream title" where the "[ ]" part is the icon representing that category.).
Easily identify noteworthy or special dreams. Before the category icon, various other icons are used to represent noteworthy or special dreams. These include ribbons for the first and second best records within a type (not in a category), the worst record, a book for dreams that are well-recalled and have been posted as a story (there aren't many of these), and, most common, a page for a noteworthy dream (typically a dream that has an unusually high rating somewhere or is special in some other way). One dream uses 6 icons (7 if you include the category icon) - yep, the best of the best which holds a lot of first and second-place records (the bulk of the icons are from records). If I expanded on it, writing a story (it would have to be mind-game-enhanced though since there aren't enough details recalled for it), I could make it use 7 (or 8) icons instead.
The main document is cleaned up. What a mess! The current document is going on 80 KB in file size and it still grows bigger. The new document is barely 1/3 that. The bulk of what causes the large size is both the FAQ, explanations of outdated ratings bars (the old, often misleading 0 to 10 scale I used, getting less accurate the further from neutral it got), and the cluttering of the newest updates. The new version is much cleaner and much more to the point.
Separate pages for the FAQ and recent updates. The FAQ is quite large so it is posted on a separate page. The recent updates list will still be there, but only include the last 5 updates or the last 3 months' worth, whichever is longer. All previous updates, back into April of 2005, will be on a separate page.
List of records. While the FAQ sort of covered this, it really didn't go into much depth. The new design now lists nearly 3 dozen records. Of these records, there's a first place best, second place best, and first place worst record for almost everything. There's also the oldest, longest, most parts (unique segments), recalling dreams, and a few others as well. I'm currently undecided as to whether to have this in the main dream journal page or a separate page (not in the FAQ either, a completely separate one). I'm currently leaning 70% toward having a separate page, the other 30 toward having it in the main dream journal page.

There's probably others that I don't recall, but that sure is quite a change. Afterall, with now over 800 dreams, the dream journal could really use it. So far though, only about 120 of 738 or so have been processed. New dreams added are automatically done for the dream journal so I don't have to bother with any new ones. I would really like to finish this before I reach 900 dreams (and knowing my average recall rate, this is likely to be reached around May of 2010).

#9 Roof construction - worse condition than expected: When it rains, it drips and in places it shouldn't be. I'm referring to a leaking roof. When I was assisting my father with fixing it, I was shocked as to just how bad the roof's condition was, at least for the worst spot. I even took photos and videos. The videos are for archival only - they will not be posted. Only one photo is posted though, of which is below. The others are for archival only.

Using the laws of physics and some logic, I would think that the culprit isn't really too much in the way of rain or age, but rather the small slope of the roof. Normally, I see houses with slopes at 4:1 at the most level, typically 3:1 and 2:1, rarely something as steep as 3:2. Ours is just 6:1 or so with the garage being a tiny 10:1. The bulk of the leaking is where the 10:1 slope is. With a slope like that, water's acceleration is only 9.95% that of normal gravity, very little and any small kink can cause water to be held in place. I tried to convince my dad to increase the slope, but he didn't listen so I have a feeling that these problems will come around again fairly soon. While the carport also uses the 10:1 slope and a garage extension uses a 40:1 slope (yes, 2.5% normal gravity), these are meant to house cars and aren't much of a concern.

#10 42 new dreams, 3 in one night for the fifth time: Weird dreams, lots of them. I seem to have a surge in weird dreams lately. I don't know why, but I do. I had a good recall rate during April, especially around April 22. On that day, I recalled 3 dreams in one night, for the fifth time. Before the first case of this happening occurred, I was wondering when a case of 3 dreams in one night was going to occur. I knew it was going to be rare. Well, it wasn't. 831 total dreams with 5 cases of this - that's one case every 166 dreams, in a sense. If at least 1 dream is recalled, I have a 1 in about 80 or 90 chance of this occurring. It turns out, it's not so rare after all, at least no where near as rare as I originally thought. I'm now kind of like this for a case where 4 dreams occur in one night. So far, that hasn't happened. I'm thinking that it may not be until 2015 that it'll happen.

#11 Nelson Carlson Lake, again, and my skills with water: Swimming is one of my top favorite things to do. Trouble is, it's rare. The nearest known lake with a beach is 31 1/2 miles away - Rice Lake. Trouble is, it appears that it's almost all houses. There was a public beach that I recall being there, but at just a measely 60 feet, it wasn't worth the effort so the original, long-standing champ of all known nearby lakes, Nelson Carlson, was visited instead. Its beach is about 400 feet, probably more, although the lake is another 1 2/3 miles further out. It's the lake often mentioned in my blog. Unlike previous cases, I wanted it for my birthday. My birthday was clear back in January though and water is frozen solid at that time for months to come. Knowing how sensitive I am with cold, and how inheritly dangerous it is, swimming at the lake in January is certainly a no go. I had to wait several months for my birthday gift to be given and it finally happened on July 23. It seems like there might be another upcoming case, but I currently don't know anything about when it will be, not even the exact week is known. August, however, is the likely month.

So, you might be wondering what my skills are with water. In short, they're very limited. They aren't too decent, although I've figured out a bunch of it on my own (often through using physics and logic), sometimes by mistake. Here's a rough idea on the skills rated 1 (total unknown) to 5 (master):

Skill	Experience	Brief notes
Floating	Intermediate	My primary survival method for very deep water
Surface swimming	Starting	Figured out on my own, but easily exhausted with 50-foot range
Treading water	Starting	Figured out unexpectedly - 20-second maximum duration
Going underwater	Intermediate	No problem, other than not knowing where I'm going
Swimming underwater	No idea	How is it even possible?
Diving	No idea	Other needed skills are too weak to do this
Doggie paddling	Beginner	Very slow, but a 200-foot maximum range

I pretty much have floating down. If I were to fall in water deeper than I am tall, this would be the only way for me to handle it, leaving me at the mercy of any currents. Surface swimming was figured out by looking at the physics and using some logic. Trouble is, I get exhausted after just 50 feet. I can only do it on my back (head pointed toward the sky). Treading water was figured out unexpectedly upon noticing how, by repeatedly moving my arms outward and inward in a certain way, I was held up. I thought it worked kind of like an airplane's wings, although 40% certainty on this makes it quite unreliable. I can hold my breath for up to 20 seconds underwater, but when I do go under, I always close my eyes. I have no idea what it looks like underwater for real (cameras from TV or looking down when just at the surface are the closest I can get right now). Swimming underwater is one skill I don't have. I don't even see how it's possible. Push your arms back to go forward (blindly for me; Newton's laws of motion (equal and opposite reaction)) then forth (to reset my arms) and you're going nowhere (since you're canceling your forward movement), as I'm seeing it. Diving is a skill limited by other skills. Diving requires deep water as far as I know (no, not cliff diving, unless you consider a 3-inch ledge a cliff) and since my swimming skills are so limited, this skill is out of reach. Doggie paddling is pretty much the same as underwater swimming. Trouble is, I'm 100 times faster on land in a sprint than I am with doggie paddling. On land, I can reach 13 or 14 mph, but I can barely manage an inch per second with doggie paddling and it's apparently for the same reason as underwater swimming. That inch per second may actually be due to currents. The up side is that my range is much more than 50 feet, possibly 200 (untested).

As you can tell, I have some progress, enough to survive in the event I need to, but my skills are very limited and weak. The bulk of it from the lack of doing it and the fact I've never had lessons.

#12 Download problems - Flash, Firefox, or software conflict?: For about 2 years now, since when version 2.0.0.8 was released (roughly - around June or July of 2007, shortly after joining YouTube), I've been having these problems with downloading files. I first thought that some program was blocking the downloads so I tried getting rid of Flash and going back to a version of Firefox that I know worked (2.0.0.0) without problems and everything was blazing fast. I installed Flash with this older version and the speed problems started again so I thought Flash was the culprit. So, I wipe out Flash as entirely as I could. I then installed Firefox 3. Nope, slow. I'm now thinking of going back to Firefox 2.0.0.0 and staying at that version for a long time... or even changing browsers again (certainly not IE - I want something without ActiveX and can run Flash without these problems. Mozilla advertises Firefox 3 as being fast, but it's not to me, even without Flash. It appears to be either some program conflicting with Firefox, Firefox itself (others don't have it and I have essentially a clean install (no add-ons, plug-ins, or anything, ruling these potential culprits out), or something completely different. It's hard enough downloading a 130 MB file, but if I use IE, that same file, even with Firefox running, downloads without a hitch hinting that it's not internet connectivity troubles and more software-related. I don't like IE any more, but I only use it strictly for testing purposes.

#13 JPG quality improvement with an unknown: JPG - its lossy and I rarely use it as a consequence. Every time I use it, I notice some strange bleeding effects (even at 1x zoom on my 144 dpi monitor, I can easily see this at maximum quality, 100). Pure hues seem to become more muted, causing artifacts to be easily noticed. I always thought JPG was like that... until recently. While fiddling around with GIMP's JPG save settings, I noticed that, with quality 100, even with 8x zoom, I can't tell a difference. Differences only started to be visible around quality 97. Those two photos involving the roof repairs above are quality 80 using this new method. The only downside is that the file size is bigger, of which I expect. The unknown is browser support.

What is the discovery? Use the "1x1, 1x1, 1x1" subsampling instead of GIMP's default "2x1, 1x1, 1x1". Also, use "floating point" as the DCT method for another quality boost. I made a few test cases to explain what I mean.

JPG Quality Test - baseline

JPG Quality Test - old method quality 90

JPG Quality Test - old method quality 60

JPG Quality Test - new method quality 90

JPG Quality Test - new method quality 60

These are the test cases. It's essentially the same as the images at the top in the first headline, but adjusted accordingly. I'm using my BMP original (snagged a screenshot, then saved it as BMP) only I scaled it down to half the size (from 1024x768) and saved it as PNG, a lossless format. From this PNG, I saved the image as JPG 4 times. The image below the top one is the results of the old method with quality 90 used. The one below is quality 60. In these two test cases, you can easily see the discoloration present with the emerald and the bluish platform it is on. The colors are muted some, especially the reds. The two images below use the new method at quality 90 above and 60 below. Compared to the original, you practically can't tell a difference between this and the original with quality 90. With 60, you can, but it's much better quality. The image at the bottom is what the settings I used for the fourth image (new method, quality 90). By doing this, I can get far higher quality than I thought possible with JPG.

#14 3D modelling program - a start: Menus, done. User interface, done. Planning, mostly done. Save file format, stumped. DirectX, can't seem to get motivated. For any 3D design to be done, right now, I have to either use the highly limiting MED from Gamestudio that I abandoned in 2006, or use slowly use Excel and plot points, connecting them. I've tried Blender, but it's confusing like crazy and it's slow to use. I like the kind of modeling program where I can use both predefined shapes and create vertices for custom or odd shapes, and is also fast to use. MED qualifies in every regard here, except for two major things. The first is that texturing isn't consistent enough due to the odd way the objects are scaled/moved. The second is that I have no idea what the file format is. The first case is the biggest problem. The second is fairly close behind. There are other minor problems (such as the light source always moving when there are cases I wished it would be static, the undo system is quite glitched, and the program is otherwise quite buggy), but for the most part, it's still quite worthy. I'd just like something better and more effective, particularly where I can view the whole level/area all at once in the same view. 65,535 vertices is far from enough in that case. GG13's scenery could easily use up 200,000 vertices. Gamestudio's WED doesn't show the texture of the object. For now, I'm left with either not having a modeling program/3D level editor or having to build my own (and programming skills are so lacking...).

#15 That stuff down there is big time blue!: Ever had something stuck in your head for months? In my case, it's the exclamatory sentence "That stuff down there is big time blue!". A closely related one is "That stuff down there is o-o blue!". The whole thing is said in a rhythmic way. There's often another that's related, but different: "and it's o-o below" or "and it's o-o down". What's this "o-o"? It's the sound of the "oo" in "boot", only it's said twice (like a sixteenth note, an eighth rest, and another sixteenth note). The one with "o-o blue" at the end has "blue" said as "balloo", like the word "balloon" only without the "n" and the first part said as "ball", with the short A sound. It derives after doing what I often do with Tails, in Sonic - put him in the water and wait for the water to defeat him. Oddly, I haven't played Sonic in quite a long time so I don't know why this is going on... for about 3 months now.

#16 "Creepy crawly" - insects, especially ants: Yet another thing. Whenever I see a beetle, moth, or any insect, especially an ant, I often say "creepy crawly". It basically started during the construction stuff. There's a lot of ant hills around my house. When I'm doing the construction stuff, waiting for dad to saw something, I'm often covering up the ant hills and watching how the ants react and eventually move the pebbles around. Sometimes, I may give an ant to a "super creepy crawly", a spider, as food for the spider, and watching the spider web up the ant and even once watching the spider eat it.

#17 2 years at YouTube - and it went unnoticed: It wasn't until about a week or two after I was at YouTube for 2 years until I found out that I've been with them for that long. I was actually surprised. Gamespot lasted only a month (due to their numerous bugs at the time), but YouTube has gone on for 2 years and still counting. In fact, I don't see any sign of leaving YouTube any time soon despite the fact they have dozens of bugs that I'm aware of and there's now no longer any known way to contact them.... One such bug is the loss of paragraphs in video descriptions - returns are ignored causing every sentence to be bunched up, a problem present for, get this, 16 months. Going on for 3 months is the bug where I can't unmark a comment as spam. If someone flags a comment as spam, I have absolutely no way to mark it as "not spam" and thus their comment will remain marked as spam until YouTube fixes this long-lasting bug. I can mark comments as spam just fine, but not unmark them. Sure there is the "not spam" link, but clicking it has the same behavior as clicking regular text, as if there was no link there at all. There's many others around that I'm aware of, but these bugs are no where near as bad as those encountered on Gamespot. Sometimes I encounter cases where I'm completely unable to upload videos (like when I tried uploading the "max score in 8 minutes" video for Super Mario World for several days. The most annoying thing about YouTube is the fact that they have to take the whole site down to perform maintenence. I never do this kind of thing nor will I ever do it. Can't they just create their code in test files and merely just copy and paste the test files over? Instead of taking the site down for 90 minutes or so (which seems to be the typical case), only a few seconds is needed, if that. Still, it's not quite enough to deter me from YouTube. Once these bugs are fixed, it'll be much better. I sure hope they do, but without a way to report bugs, I'm out of luck....

#18 Any moment now - story copyright registration: On December 18, 2008, I submitted my story for copyright registration. I'm thinking that it'll take 5 to 8 months to complete registration. This means that copyright registration should now be completing at any moment. When it does, my book will be available on Lulu. The only thing left right now is getting the "sales pitch" paragraph on the book's back cover.

I've often been getting requests for a T-shirt that contains my book's front cover and you may be wondering why I haven't done anything. The sole reason for not doing it is copyright registration - I want it to complete first. Once it does, then it will be available shortly after and there's a very good reason. It's advertising but with a negative price tag (that's better than free). By you having the shirt with the image on it walking around on vacation or in a town, you are, in a sense, advertising the book and, because you're buying the shirt where I can profit off of this, the price tag for the advertising is negative, a rare thing. If a set of speakers were available for $50 and you bought them, you'd pay the store $50 (plus tax). If it was -$50 and you bought them, the store would pay you $50. That's what I mean by a negative price tag and this is why it's a rare thing. I've only seen it once - pay $4 for lights but get a $5 rebate, a -$1 price tag (probably more around -$0.60 after roughly 7% tax and standard interest are factored in). Unlike this tiny -$1 case, it's potentially a few hundred in the negative, depending on how the shirt sales go. Yep, even outside video games, I'm looking for weaknesses I can exploit (legally of course).

#19 Focused TV ads - more useful: They play, but a lot of them are irrelevent to me. I'm referring to ads on TV. For a brief idea, to get a background, I've listed various types of ads that I see, their frequency, and how they apply to me.

Ad type	Frequency (5 is high)	How it applies to me
New car		Useless: I can't drive
Car insurance		Useless: I don't have a car
Flood insurance		Useless: I don't own a house
Life insurance		Questioned: don't know if I really need it at my age
Fast food		Doesn't apply: I'm not into fast food anymore
New pizza		Rarely applies: Last useful when Pizza Hut announced its 16-inch pizza
New movie in theaters		Useless: I strongly dislike movie theaters
New movie on DVD		Rarely applies: I'm hardly into movies
New computer		Indirectly applies: I prefer to build my own, for better control and to avoid Vista
New music CD		Doesn't apply: Instrumentals are rare and I have yet to encounter anything worthy
New video game		Rarely applies: Game is for a system I don't have (most likely), too violent (M-rated), or not my type
New episode		Often applies: Depends on the type/content of the show
Weight loss		Doesn't apply: Paper towels to degrease pizzas while avoiding sodas and candy is about as effective
Vonage		Useless: long distance calls are very rare and its already quite cheap
Cell phones		Useless: I have no need for one
Inn/hotel/motel		Doesn't apply: The last time I stayed at one of these was back in 2000 and is mega rare
Travel sites		Doesn't apply: It's so rare and it's been more than 20 years since I've been elsewhere than ND and MN
Call to order a product		Rarely applies: I have yet to find something of worthy enough use/interest
Legal claims		Doesn't apply: So far, these items don't apply to me - the one for Seroquel is the closest thus far
Allergy meds		Useless: I'm not allergic to anything that I'm aware of
Pain relievers		Useless: I practically don't have any pains
Sleep aids		Useless: I don't have problems getting adequate sleep (check the wakeup and sleep times in my daily entries)
Tax relief		Useless: I don't have tax debt (I also don't see how one can get $10,000 in tax debt in the first place)
Pet food		Applies, but my parents get it
Pet treats		Applies, but my parents get it
New credit card		Questioned: current company policies disliked; credit card = debt and wasted money;

As you can tell, almost all TV ads are either useless or don't apply. A few rarely apply, applying once every 5 or so years. This makes me unaware of numerous things that I would otherwise find useful. Take video games. I'm into classics. Now that the new systems are out, most of what I see doesn't apply because I don't have the system. When I do eventually get it, if I do, I'd have almost certainly forgotten about it. Most games I see on TV are online-based games or M-rated ones, neither of which I'd bother playing. Where's Mario? Where's Final Fantasy? Where's RPG's (to generalize)? Where's platformers (to generalize)? I haven't seen any RPG's or platformers advertised since probably 2002, perhaps even earlier. The last time a video game ad was of any use was when Sonic 3D Blast (on the Genesis) was released... more than a decade ago. The next game after it was one that I recall used 24-bit color (the 24-bit color was a key element in the ad). It was 2D and I recall a scene of the character jumping over an edge and landing in a small puddle of water. I have no idea what system this game is toward or any hints of its name. It occurred supposedly in the late 1990's making it toward the PS1 system, Nintendo 64, and whatever other system(s) there were during the genre. I have yet to get this game, whatever it is.

Because of this, I have a great idea that could help resolve some of this. That idea is individual-focused advertising. How it works is fairly simple. Digital TV is a near-requirement. With built-in software for things like DirecTV, digital TV recievers, etc., there would be various options to control the types of ads shown. For example, some of these may include travel, pets, food, cars, and other such general categories. A user rates them from 1 to 5 depending on how important it is to them. For example, let's say someone is involved with a lot of travel but doesn't have any pets. They may set the value for travel higher so travel-related ads (destinations, airlines, hotels, etc.) appear more frequently, but pet stuff (food, treats, etc.) lower so they appear less frequently (the lowest setting should still show them, for the case of group-watching, potential friend telling, and to prevent abuse (abuse: stopping ads from showing at all)). Depending on the setting, points are given. Points are used as weights to influence the frequencies of such ads appearing. A rating of 1 in a category gives 2 points toward it, 2 for 3, 3 for 5, 4 for 7, and 5 for 10. The totals are then added for all categories (this can vary with the user) with the upper range limit logged for each category. Let's say there are 3 categories (for simplicity in my example). The example user would have cars as having 5 points, pets as 2, and travel as 10, making the end points as 5, 7, then 17. When choosing an ad, a random number generator picks a number from 1 up to the actual grand total. Let's say 8 was chosen. Since 8 is above 7 but less than or equal to 17, a travel-related ad would show. Let's say that 5 was chosen next. It's greater than 0 but less than or equal to 5, so a car-related ad is chosen. Of course, instead of 3 categories, there'd be several, like 12 to 20. Now, how to make sure the ad session's alloted time is set properly, that I don't know. There'd have to be measures to prevent abuse though, because, if someone hacks into the system to otherwise disable all ads, this can cause problems. The values should be adjustable though because change happens. That user who may have once been into travel may suddenly get much more involved with cars instead (putting travel at 3 and cars at 5).

This system provides some advantages to advertisers. Instead of just focusing on the type of user likely to be watching a given show (like news, a movie, Mythbusters, etc.), covering a range of audience otherwise irrelevent to them, they'll be able to directly focus their advertising on a user that could actually use it. For users using satellite or cable TV services, it can also be put into an ad index so their ads can cover practically any station allowing users who are into such advertised products to more likely be aware of them and because the user is more into such an item, more likely to get sales as a result.

When the time comes, it may be possible to have more specifics within each category. I'm into older video games much more than newer video games, so if a store has such things available and posts an ad for it, I'd be aware of it and possibly check it out. Of course, the fine details need some adjusting. This is only an idea I thought of since early May of 2009.

#20 TV ads - ban those of higher content ratings: I'm watching an educational show rated PG then an ad session comes up as usual. Trouble is, the ad is for content rated MA. Shouldn't the content ratings block such ads from even displaying or showing at that time on TV? Those Viagra ads are more for a TV-14 type show due to the nature of the product. Ads for new cars and cell phones are essentially G-rated and would always be allowed. One thing that bothers me is when an ad comes up for boxing where actual fighting is present and it comes up while watching even G-rated shows. Most of what I watch is G and PG, rarely 14, but never MA (in increasing order for those not familiar with the US TV rating system (there's Y and Y7, but those are for kids shows)). I would think that TV ratings should affect ads too. Dating service ones are more G than PG.

#21 Car fuel efficiency - a new approach: It often targets newer cars, not gas-guzzling existing ones. This is refering to the government's plans for improving fuel efficiency. I have an idea myself. Instead of targeting newer cars, target the older cars. Have it so that, for example, all cars must get at least 18 miles per gallon by 2015. The 2015 comes from having enough time (6 years) needed so that one can save some cash to get their cars upgraded as needed. Trucks, by their hauling nature should get 15 miles per gallon instead at the minimum. I know of some 10's and even 8's still around. By getting this up over the next 6 years, it can offset quite a bit. I don't have statistics at hand, but let's say that 30% of cars in current operation get under 18 mpg. Of these, the median is probably about 12 mpg. If these were upgraded as needed, the effect would be about the equivalent of making new cars use about half the fuel (not 2/3 due to the fact there are probably many more older cars than newer ones). Cars that don't meet this by the deadline should pay 50% more in gas until the specs are met. For $4 a gallon gas (whenever that comes around again), it can be quite an expensive fill-up.

#22 Social Security Numbers - redesign the system: I've heard a lot on the news about some using the social security number of babies and those that are dead. Well, as I'm seeing it, the odds for thieves are quite a bit in their favor. For the basics, there are 9 digits formatted as "XXX-XX-XXXX". Mathematically, a thief has a one in a billion chance of guessing your number at random. This may seem comfortably rare, knowing how long it takes to win the Powerball lottery (you're 4 times as likely to win the jackpot), but considering the fact that the US has about 320 million civilians, the odds are very decent for a thief to randomly guess one, nearly 1 in 3, about that of rolling a 1 or a 2 on a standard 6-sided die. Given this, I think the system needs a redesign to include another 4 digits (making the current 1 in 3 more like 1 in 30,000, about the odds of getting 4 of a kind in Poker on the first draw and a 1 in 10 trillion chance for getting your number. I would think the format should be similiar to the existing one, only with the extra numbers likely at the end or something, such as "XXX-XX-XXXX-XXXX".

#23 Laws involving money - scale with inflation: It hasn't been updated since 1940. That hasn't been updated since 1921. This hasn't been updated since 1973. Take the FDIC case. It was supposedly set at $100,000 during 1930-something (I don't know the exact year - check Wikipedia if you want it), but, even in 2005, it was still $100,000. My suggestion involves having these and other such amounts of money scale with inflation. $100,000 back then may be $250,000 now. It may be $300,000, if inflation checks out, by 2023 (assuming a constant rate of 1.3% per year (a number I recall); I had to solve an equation to figure that out.). By scaling with inflation, it'll make it easier to keep such laws up to date, typically updating once every year, 5 years, or 10 years.

#24 Comparison of mini pizzas: These are the deep dish single pizzas that I refer to. They have one big advantage and one downside. The big advantage is that I can eat whenever I want, without having to worry about not having a plate or something. Making up for the plate is a paper towel, to help with the degreasing. The one downside is that these things tend to be rather greasy so I need to focus more on the degreasing which costs more time. Each of the brands has their own pros and cons. I've summarized this in this convenient table (the more stars, the better - see the table footnotes for more in-depth details):

Brand*	Opening***	Degreasing†	Cheese‡	Sauce††	Crust‡†
Tombstone (cheese)
Tony's (cheese)
Red Baron (4-cheese)
Red Baron (cheese)
Great Value** (cheese)
Digiorno for one (cheese)

Table footnotes:
* Some brands are left out as they are either store-owned or none of my local stores carry them. Thus, comparisons are impossible to make.
** This is a store-brand pizza from Wal-mart.
*** This is how easy or difficult the package containing the individual pizzas is to open. The easier it is to open, the higher the rating.
† Degreasing is where I use paper towels to soak up the fat (see my "fat removal trick" in the tips and tricks index for details on this), something I've been doing for years and I know it works. The more degreasing that is needed, the longer it takes to eat. The more degreasing a pizza takes per unit of mass, the lower this rating (bigger pizzas will naturally require more degreasing).
‡ This is how good the cheese part of it tastes to me. The better the taste, the higher the rating (note that this is relative to the one furthest from neutral (on either side) one against neutral rather than direct compatibility).
†† This is how good the sauce part of it tastes to me. The better the taste, the higher the rating (see the note for the footnote above).
‡† This is how good the crust part of it tastes to me. The better the taste, the higher the rating (see the note for the footnote 2 above).

As you can tell, the Digiorno Pizza For One is the best of the best. Too bad it's hard to find and the price is the highest per unit of mass. These pizzas are 70% bigger (about 9.3 ounces compared to 5.5) but tend to have double the price per pizza (making it, essentially, about 20% pricier per unit of mass). However, for the quality, I'd easily accept another 20% more in price with these.

Following behind, since the Digiorno ones are hard to find in quantity (often 0, but whenever they are present, it's just 4 instead of a more needed 15 to 30 every 2 weeks), are the Tombstone ones. Its primary downside is the package itself - it's a pain to open, often taking a whole minute just fighting it, going on 3 minutes for the extreme end (4 is the record at the moment). Outside this, it's an excellent pizza and easy for me to buy in quantity. If even this isn't available, then I focus on Red Baron's 4-cheese variety which is very close behind (I typically get 8 of these for every 9 Tombstones). The only 2 downsides is that it's a bit tricky to buy in quantity and they tend to be a bit pricier per unit of mass.

If even this isn't available, then I just go further down the list with the other Red Baron variety then Tony's. Tony's takes a long time to degrease so I tend to avoid them, using them as a last resort. However, their price is quite low and they're fairly easy to buy in quantity. If all else fails, then the store brand, Great Value, is the true last resort. It's crust is horrible because it's 90% crust (with a tough and low-quality taste), and essentially no toppings. The toppings part (sauce and cheese) are the only good things about this particular brand. I start getting into wanting Macaroni or Hamburger Helper by the time I reach Tony's but these require a lot of dishes and quick-to-spoil milk.

The "perfect mini pizza" would be having one as easy to open as Tony's, have low fat (this influences the degreasing - 6 grams for every 130-gram serving is good like the large Digiorno pizzas have), have the cheese of Great Value in a way like Digiorno's (by quantity and how its spread out), the sauce of Tombstone, and the crust of Digiorno. For this combo, I likely wouldn't mind a $4.50 price tag for a 2-pack of 6-ounce (170.1-gram) pizzas at all. Higher compatibility means a greater likelihood toward buying the more expensive brands (with limits of course), but it's of little use if there's enough in quantity.

Daily Events

May 1: Number-drawing function fixed... mostly. The transition from normal to exponential is flawed. I made a sample level to get a sense on the looks and got a video of it. I fixed a bug, enhanced the icons (brighter lighting on objects, made hearts consistent, added details on how the level creator will work (so I can make levels more effectively), and adjusted scenery details, mostly adding new headers and adjusting target level size. Woke up at 9:27 PM and went to bed at 3:47 PM.
May 2: Pools for water, mud, and lava are done. I found a new technique to improve quality through noisemaps - stretch the important to double size. Wavy water for the dedicated water worlds is done, for the part the character goes in - B+. I started the character at the end of the day. The base forn and some easy ones are done. I started run, which has 10 frames (5 unique). Lighting is worked out - 3 tones used. Before all this, adjustments to the plans were made. The main one involves scoring - time bonus and one other. From this change, other adjustments had to be made. The time given for each difficulty setting is worked out - 3 notches lower (originally 4) meant a doubling of the allotted time. Woke up at 12:10 AM and went to bed at 3:38 PM.
May 3: Character states, done! Leviburst was a bit tricky. Lighting is about 1/4 done. Walk/run came out better than expected. Bounce is the best/most realistic I've done it. Finish the lighting, finalize the character, fix the text draw function, figure out colliection detection, then try out simple game mechanics is next. House work around 9:00 AM slew me down some. Woke up at 1:05 AM and went to bed at 4:58 PM.
May 4: A travel dream was recalled. The character is done - 48 states. The text draw function is redone though one bug still remains - lots of extra space at the end with word wrap = 2 odd spaces on the first line. This class 1 bug is not problematic for what I need though. Time experiment started... 14 1/4 hours before I went to bed. I hope I can pass 33 hours again. 7:19:40 AM is when I turned it on and 7:19:57 AM is when it started playing. I use the latter of the 2 times as the 0-second point. Winter Land at 55% true speed is used. I also finally got Cinnamon Altoids, a rare thing now. I got the video on making Hamburger Helper without the hamburger. I'm still at Winter Land - wow! I've had it nagging in my mind - name the character in Platform Masters Jeremy. Woke up at 3:38 AM and went to bed at 9:34 PM.
May 5: I was getting a bunch of 31's, now I got 32 hours, 24 minutes, and 26 seconds as the duration. The text draw function is supposedly fixed - thanks to thinking of a new idea - log the end of each line during stage 3. Stage 4 is the source of many problems, along with word wrap. The low density clouds are done, except testing them. In the middle, I got a video showing the recent progress. Near the end of the day, I got an SNES off Ebay. Super Metroid is the main game I got - the other was for my sister. I should have it by Thursday of next week. Woke up at 6:10 AM and went to bed at 12:39 AM.
May 6: Slept 7 1/2 hours - upset stomach.... I set up for high-precision movement, required by collision detection. It's not too difficult, especially in the later worlds, to get 500 mph where falling through the land is possible. I also need it for optimal scoring in the time bonus since each millisecond matters. At the end of the day, I tried getting malloc going again - I got it working for a pointer in a struct, but, like months before, not outside. I thought YouTube might help so I got a video explaining the issue using my test functions. It was good on take 1. I'd like this resolved before I make the level editor. There's still a bunch of other graphics to add as well from the 14 other worlds, some partially done. Woke up at 8:45 AM and went to bed at 10:29 PM.
May 7: Did I just waste the whole day fighting malloc? 5 hours of converting and malloc is now supposedly working - it's in a struct. The problem now is AlphaBlend - it doesn't modify the contents of the buffers. Apparently, BitBlt isn't working either. I had this same problem from 2 conversions ago. Once, the window class wasn't being found, even though I didn't modify this content or even touch it. I even tried directly setting the values - no change and I don't see the 4x1 area at all! No drawing, no progress. If not resolved on Saturday, I'm going back to direct arrays and sticking with it to the end. I'm probably going to ditch the light edition. Today's systems could easily take 64 MB. I forgot to charge the MP3 player battery as I was going to bed. 3 nose bleeds occurred - evil strikes again! Woke up at 9:48 AM and went to bed at 12:01 AM.
May 8: Malloc abandoned - without changing the drawing stuff, all works again.... When was Radeon 7000 (for 64 MB memory) released in a budget system? 2006? It was released in 2000. 64 MB is hardly a dent in today's systems. TSO is the same. I first reverted back, taking 4 hours (yep, about 15 hours down the drain fighting malloc). I adjusted for the other 3 sizes (2x1, 8x1, and 1x1 - 1x8 deleted). I began to add moving platform functionality but was stumped on how to do it. A shower and grocery shopping followed. I, gradually, worked out the system, but needs optimization. I set up, ate, but was too tired to resume. I changed to Village Zone then OWTLI (the latter when I went to bed, after dealing with an overflowing toilet). Woke up at 10:52 AM and went to bed at 12:36 AM.
May 9: Slept 10 1/2 hours.... An odd/interesting dream was recalled. Moving platforms done (dynamic type - the character is required for the other 2). I need this for the level creator. I also worked out the fine details on the game's enemies - 16 so far. I've met both of my goals for enemies. The character's actions, the menus, and on-screen data are likely next - ditch most of the debug panel. Woke up at 12:40 PM and went to bed at 3:32 AM.
May 10: Ma's day and no dream.... Daily check, readjust enemy sats, eat, fight parents' bed/mattress, decide on icons for screenshot data, make giant font, eat, make the score icon, then form the checklist. I got 2/3 of the entries done, but have no idea on the progress. I also got a video of "Goofy Fuzz" climbing up stuff - yet to be processed. Woke up at 1:03 PM and went to bed at 4:53 AM.
May 12: Checklist done - tempted to play JF2 for height - lasted 40 minutes. JF1 followed for 20 minutes. Dominating the day was Spyro 1, completed 100%. 2 songs and a video are wanted. Payment still pending - it had better clear when I wake up - if not, then sister may not get her present. I tried to play Skullmonkeys but couldn't find the disk... after searching for an hour. Woke up at 3:28 PM and went to bed at 9:36 AM.
May 13: Time and status are done. Lives and current level are yet to be done. Subheaders for "status" still remain as well. I got a video of the out of bounds glitch in Spyro 1 (and got the music from that level - both processed as well). Nothing else worthy happened. Woke up at 6:38 PM and went to bed at 12:07 PM.
May 14: I attempted "current level", but it wasn't coming out right. I then went at Super Mario Sunshine and got the video for the unlimited coins glitch - 8 takes were needed. Take 5 was borderline and partially processed. Being C+ for the grade, it was then rejected. Take 8 was spectacular, except for one mistake in the longer route case. I got "lives" done but had second thoughts on the "current level" case - leave it out and it'd be an oddball. If I can finish that, I can make a simulation screenshot. Woke up at 8:21 PM and went to bed at 12:58 PM.
May 15: I've finished the icons but redid some - for consistency reasons mainly, but others due to a design change. I also created 2 simulation screenshots - one for the onscreen data for overall looks during actual game play and the other for the end of level results. I also got the SNES and sister's present - I played it some (not my type - fighting) but only to make sure it and the system worked. I didn't get Super Metroid. I got a video of the 2 simulation screenshots as well - it may be the last for a while.... Woke up at 11:44 PM and went to bed at 4:27 PM.
May 16: Daily check, work on game while waiting for good stuff to come on on TV, Skullmonkeys whole game, work on game - ocean world, nearly finishing an island, visit sister for mowing her lawn - mower wouldn't start (due to bad spark plug and/or coil), eat at Denny's - 50 minutes just for the food, I finished the island and figured out how to do the waves. HWW also finally changed to a new forum, PHPBB. Woke up at 1:48 AM and went to bed at 9:55 PM.
May 17: Ocean world is pretty much done - the waves need lighting effects, and the reef remains. Work on game's scenery started the day. Leave for sister's to mow her lawn - mostly done... until mower wouldn't start (ours too!) Sort stuff for garage sale (give gifts to sister beforehand), resume game while others went to Applebee's and continued to day's end. Moving/installing new couch was before sorting. I should have Super Metroid tomorrow. Woke up at 9:53 AM and went to bed at 11:36 PM.
May 18: Slept just 7 hours.... Super Metroid didn't arrive. I checked out the door several times (like 6 to 8 times) but didn't see it. I started the 8th world - I've mostly finished the design of the object at a scaling of 8 (object is classified)... after about 4 redos with the design and intended scaling, finalization of the base design, and animating it. Intermittent was Super Mario Sunshine and trying the Yoshi-Gelato-Beach bug - I see somewhat on what is going on. Looking for a way to pass time, so Super Metroid came was a self-challenge - fire 10 fruits at a Pianta on the beach near the market before they disappeared or flashed. Woke up at 7:46 AM and went to bed at 10:30 PM.
May 19: The object with a scaling of 8 in the 8th world is done, minus animation. The objects at a scaling of 2 and 4 are also done minus the lower half. Super Metroid has yet to arrive - suspicion++. Medium font redone - much better (it was way too thick and seemed like it was bold. Woke up at 9:25 AM and went to bed at 1:18 AM.
May 20: Still no Super Metroid - seller contacted.... The objects at a scaling of 2 and 4 in the 8th world are all done, one of which is a bit funny. I also like the world's official name too. I now just need the more distant stuff. Codewise, it appears that only vsync is the only thing missing. I seem to have an option without the use of DirectX to get full screen, thanks to someone on YouTube. Woke up at 11:42 AM and went to bed at 2:45 AM.
May 21: World 8 done! I'm thinking of doing the overworld next. As to Super Metroid, it wasn't shipped but I also end up getting a free game! The delay is resulting in extra progress with Platform Masters.... 3 worlds down, 12 to go! I also, near the end of the day, had the runs. Between Mario World and Final Fantasy Mystic Quest, I let the seller decide on which I'll get so I won't know what free game I'll get. I should get them by Tuesday or Wednesday. Woke up at 12:45 PM and went to bed at 2:37 AM.
May 22: Super Mario Sunshine dominated the day, almost 15 hours. Due to an unexpected barbeque, grocery shopping won't be available until Sunday and for best results, I need to add a lot of hours now.... I thought those block levels were hard - blooper surfing and red coins in hotel are much harder than those. The blooper surfing one is much, much harder, the hardest of all. No progress was made with Platform Masters though.... Woke up at 1:29 PM and went to bed at 8:26 AM.
May 23: Super Mario Sunshine again. I looked at Super Metroid reviews and images. The same goes for Final Fantasy Mystic Quest. I'm very familiar with Super Mario World. I tried looking for other blue coins in my main game but couldn't find any. I then started a low gravity game - it was much easier than I thought.... I was going to get strategy videos for Super Mario Sunshine but was drowsy most of the day.... Woke up at 6:21 PM and went to bed at 5:38 AM.
May 25: Finally a dream! Super Mario Sunshine again, for strategies segments. I then found a way for 999+ coins in Pinna Park - I had to process that. I got Delfino Plaza's music and processed it. Eat, shower, grocery shopping, daily check, Super Mario Sunshine for strategies video (idea somewhat rejected), then realized the potential for 999 coins. I processed the video as I started my site's update. Can I finish before those games arrive? Woke up at 3:54 PM and went to bed at 10:07 AM.
May 26: Headlines half done from what I have. My evil nose decided it needs to strike - 3 nose bleeds - 100 minutes wasted right there, including one just as I went to bed. Working on the headlines while repeatedly checking YouTube on and off made up the bulk of the day. Games should arrive tomorrow.... Woke up at 8:53 PM and went to bed at 10:18 AM.
May 27: They finally came... and dominated 18 hours of the day. Super Metroid was first... until I got stumped. I have 399 energy at the moment. I'm pretty much where I left off last time... 13 or 14 years ago when I rented it. Bomb-climbing is very useful. Final Fantasy Mystic Quest followed. I'm already at level 35, on the ship part. I have all weapons, spells, and all but one item. Am I already at the end!? At roughly 5:00 PM, I got another one of those prank calls, mentioning of an alien invasion while pretending to be one of my imaginary friends (and pronounced incorrectly too). Woke up at 9:47 AM and went to bed at 6:02 PM.
May 28: Awake for 24 1/2 hours!? Super Metroid done with a little help. I was stuck in the same spot. I was last time I rented the game. The location was that of the first visit to Norfair. I didn't know about the Varia Suit - I had to use GameFAQs to get unstuck. After messing around in Norfair, I got stuck again and out came the guide. Once unstuck, getting the grapple, I figured the rest out myself to the end. The main boss for Maridia was the hardest of all - 6 game overs because of it. Woke up at 5:10 AM and went to bed at 4:13 AM.
May 30: Slept 15 total hours - 16 1/2 went by due to waking up and going to sleep repeatedly. Every 2 or 3 hours past the first 6, I'd wake up for a bit. I played Super Metroid in full again, taking just under 5 hours. I got lost in Maridia while searching for the boss costing almost an hour. Beyond that, not much else happened. I thought of reaching an excessively high level (like 40 or 50) in FFMQ. I was getting drowsy by level 8 and stopped at 10 - all in Level Forest without defeating the boss. Woke up at 8:46 PM and went to bed at 11:17 AM.
May 31: Slept 12 hours... I think I'm back to normal. Recording a full game video of Super Metroid (3.4 hours for a completion) was first after eating. Processing it took up the rest of the day. The vocal track was, oddly, faster than it should've been causing odd misalignments. This took up another 2 hours. I intend on segmenting it and starting to upload it tomorrow. 26 segments are the expected minimum but could be as much as 40, depending on how I segment it. Woke up at 1:20 AM and went to bed at 5:20 PM.

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
	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
	Extremely minor event / Very bad day

Footnotes:
None