Bio Force Ape Developer's Journal

[ProgrammingAce]

Glad to see that someone decided an actual game was needed. My only complaint is the hit detection; you hit the air above the enemy and it dies.

Of course, that's probably intentional, isn't it?

That said, I can't wait.

That's something of a technical limitation, let me see if i can explain.

To the NES, a sprite is a 8x8 pixel tile. In order to make the characters in Bio Force Ape, they're made up of groups of tiles. The ape is 2 tiles wide by 4 tiles tall.

The NES has a limitation of 8 sprites per horizontal scan line (each scan line on the TV). After the 8th sprite, the system just stops displaying them.

So in order to fit more enemies on the screen, we decided to make a short enemy that's only 1 tile wide. The little squid thing is 1 tile wide and 2 tiles tall.

So when all the enemies are on the screen at once and the ape is just standing there, the tiles go like this:

2 tiles for the ape
3 tiles for the Ant dude
2 tiles for the lizard.
------
7 tiles

When the ape punches, he becomes 3 tiles wide, maxing out the number of tiles we can have on a scanline. Luckily the ant dude is only 3 tiles wide at the top and 2 tiles wide at his feet. So the end result is the squid thing that's only 2 tiles high.

Now why can't we have a kicking animation? There are a couple of reasons. First, same as above, there's a good chance that unless we've killed something else on the screen first, we already have 8 tiles on the scanline. We don't have anything left to draw a kick. The second issue is that we're quite frankly almost out of tiles for our sprites. It would suck to have to cut one of the enemies just to add a kick attack just to handle a single enemy. Now the other issue is that we've already mapped Down + B (Attack) to the special attack. The only other button combination that makes any sense is Down + A (jump). But that needlessly complicates the controls and isn't very intuitive. You're just as likely to forget that a downward kick isn't Down + Attack and use up your special attack.

We can't remap the special attack because "paulB" already said that's the button sequence to use the fart attack. We're trying to stick to the hoax description as much as possible unless it's detrimental to gameplay. Obviously we're going well beyond what's been referenced by that thread, but we're sticking to it when we can.

So the end result is that you punch over the squid's head when you go to kill him. To make up for it, i extended the hit box for that enemy so he's easier to punch. One of the problems in early testing was that he could get too close and be really hard to hit. I'm going to tweak the hit boxes before the game ships though and review all of the enemies.

[hellfire]

you better put dur butter somewhere in the game, even if its in the credits. IF dur butter iz not in the game I will not buy it

[walrusmonger]

Soooo amazing, can't wait to buy one!

[ProgrammingAce]

you better put dur butter somewhere in the game, even if its in the credits. IF dur butter iz not in the game I will not buy it

Ja he is dur butter and he is worth 2k monies.

You can't honestly expect us to leave him out. Hell, he even has his own copyright notice.

The... er... special attack... is also in the game as well. Actually, both features were in the game prior to the release of the demo and had to be cut from the code before we released it. We didn't want people poking around with a game genie and giving away all of our secrets.

[kainemaxwell]

Looks awesome so far. Be sure to test it with a Game genie and GenNEXT also.

[PapaStu]

Looks awesome so far. Be sure to test it with a Game genie and GenNEXT also.

This should only be played on a real NES. Not being played on a NES will just make the game dump its save and delete its rom.

[ProgrammingAce]

I guess now would be a good time to explain one of the inside jokes you'll see in the game, since i just finished up that section. Something the size and scope of Bio Force Ape is going to eat up a lot of free time for everyone involved. Right now, the source code for the game is sitting at just over 10,000 lines of code, and we still have plenty more to add.

As Brat and I are working on the plot, i decide it would be funny to fit my fiancee into the story somehow. The original plan was to make her the wife that John is trying to rescue. As the story progressed and we ran into some technical issues, it became apparent that wasn't such a good idea. Without giving anything away, you'll probably see why when you play the finished game. So I just scrapped the idea of having her in the game.

Later on, as we were working on one of the levels half way though the game, we realized we needed another boss character. Out of the blue, Brat offers to make a pterodactyl boss. I stare at the email for a few seconds and suddenly a lightbulb goes off in my head... See, my fiancee's name is Terri. So from there, the Terri-Dactyl was born.

BEWARE THE TERRI DACTYL

[Bratwurst]

When I want to make graphics with lots of detail, like a portrait for a cutscene, I find it easiest to draw big and then shrink the image down. This is probably cheating, but I don't care.



Here's a quick and ugly sketch of Mr. Ape. He has seen better days.



Resized, and in need of cleaning up. This is basically for me to get a general idea of proportions.



Finished product, ready for packaging and subsequent reheating to stuff down your throat.

[Ed Oscuro]

ima sic the pixel art peoples on you

so they can cry when they see resizing works

IT'S SETASTIC! Hoping for some Musya cameo though. Evil jumping Boddhavistas for the win. You know this game wants some random decrepit shrine stage, you know it.

Holy Butter, can't wait to preorder like five copies. You know I will (well, obviously I imagine the plan is to get them out to as many people as possible, instead of speculators, but yeah, I'm stoked, especially knowing that the demo was just meant to be silly and not 100% GAEM QUALITY).

p.s. if you put some cheetahmen II mixes in there the Japanese retroscene will be all over this too.

[whiskeyvengeance]

I'm stoked! I just hope this doesn't suffer the same fate as the much-fabled Neotoxin...

[ProgrammingAce]

I'm stoked, especially knowing that the demo was just meant to be silly and not 100% GAEM QUALITY).

Well, I had hoped it was pretty evident from the demo that we were going for a tongue-in-cheek parody, i think some of that may get lost when you realize a lot of NES games really did have shitty translations and ridiculous stories.

One of the goals in the game is to duplicate the original hoax as closely as possible, but without using some ridiculous mappers, we're not going to be exact. The bigger and more important goal is to make sure the game is actually fun. I can't ask you to pay for a game that isn't worth the price. We're not going to use the fact it's an NES game as a cop out.

I'm fairly confident that aside from some content issues, this game could have been sold at retail back in the 80s. The only thing that really would have held it back is the "bio force attack" special move... i don't think that would have made it past the censors. Don't get me wrong, it's a side scrolling beat-em-up, it's not Mario 3 or Final Fantasy. But if i didn't think the game were fun, i wouldn't have made it.

I'm stoked! I just hope this doesn't suffer the same fate as the much-fabled Neotoxin...

Right now the game is playable from beginning to end. There are a few additions we still need to do, plus some polish and bug fixes. If i were to be hit by a bus tomorrow, Brat would still have something to put on a cart and call Bio Force Ape.

[Ed Oscuro]

I'm fairly confident that aside from some content issues

Better go full-tilt and put the girl turns into a tanuki scene from Musya then.

[whiskeyvengeance]

Right now the game is playable from beginning to end. There are a few additions we still need to do, plus some polish and bug fixes. If i were to be hit by a bus tomorrow, Brat would still have something to put on a cart and call Bio Force Ape.

I think it's amazing that you guys have been able to construct a more or less complete game this quickly. Correct me if I'm wrong, but won't Bio Force Ape be the first full-length NES title since Nintendo stopped supporting the system
back in the 90's?

P.S. I'm gonna feel really stupid if this turns out to be another elaborate hoax.

[ProgrammingAce]

I think it's amazing that you guys have been able to construct a more or less complete game this quickly. Correct me if I'm wrong, but won't Bio Force Ape be the first full-length NES title since Nintendo stopped supporting the system
back in the 90's?

P.S. I'm gonna feel really stupid if this turns out to be another elaborate hoax.

I'm not sure on this, to be honest i'm not well versed with the NES homebrew scene. I guess it comes down to what you would call a "complete game".

I've looked through all of the documents on nesdev, and unless i'm missing something, this is probably the first game with a plot, levels, bosses and cutscenes of any real length. I would hesitate to call it the first "complete" game though.

Someone with more experience in the NES scene may want to chime in. Before this project started, I quite frankly haven't sat down and played on an NES since the 80's.

So an update on the game, it's sitting at about 11,000 lines of code, up from 8,700 when i started this thread. Those numbers don't include the background designs (stored as maps in the rom) or sound files.

And now to bore you with some technical details. Someone asked me about compression and why we can't use that to store more artwork in the game. The answer comes down to how the NES stores the graphics tiles. The quick answer is that it's already compressed to hell. Let me explain:

If you look at the artwork that Brat has posted, you'll notice some of it is in really weird colors. That's a representation of how the NES stores graphics tiles in the CHR rom. Because ROM was expensive back then, Nintendo really cut down on what the programmers could use (and i'll give other examples of cut corners in later posts). For the CHR ROM, you get two 4K banks. One stores the background tiles and the other stores the sprite tiles.

Each of these 4K banks can be represented as a 128x128 pixel grid, and that's how we do represent them when we're messing around in photoshop. Now when you look at this grid, you'll see it only has 4 colors, black, red, blue and white. Because there are only 4 colors, each color can be represented by just 2 characters in binary, 00, 01, 10, 11. Since a byte is made up of 8 binary characters, a single byte can hold 4 pixels worth of data when stored in the NES CHR ROM.

In case i lost you, let me try to demonstrate. If the first byte of the CHR ROM is 11010010, then the first pixel is the 4th color (11), the second pixel is the 2nd color (01), the third pixel is the 1st color (00), and the fourth pixel is the 3rd color (10).

While this compresses the data in half, it creates an obvious problem. You can only have 4 colors in your game.And, well... that would suck. It's also pretty obvious that most NES games have more then 4 colors, so there must be something else going on.

So in the next installment, i'll explain how games get their color, the limitations of this system and why it sucks. As an extra bonus, you'll see why the screenshots from grand theftendo show why the game never existed...

[ProgrammingAce]

Ok, so in the last episode you saw how the NES stores data in the CHR ROM (graphics rom chip). You saw how the ROM chip only stores enough data for 4 colors. So how do you get different colors out of the system?

The answer is by using color palettes. The NES allows you to set 4 different 4-color palettes to be used at a time. This should allow you 16 colors at a time, but remember how i said Nintendo went cheap on RAM? It hits us again here. So if i were to describe the first color palette as Black, Red, Brown, and Yellow those would be the colors that make up palette 1 (yes, redundant sentence is redundant, but stick with me here). Now because Ninty cheaped out on RAM, the first color in each of the other 3 palettes *must* be black. The first color in each of the 4 palettes must be identical, it's considered the "background color". In Mario games, for example, that color is usually the sky.

This limitation of RAM means that the NES can only display 13 colors for background tiles on any given screen. Sprites have their own palettes with a similar limitation, but have their own colors.

But that still doesn't explain how the color palettes get assigned to the background tiles. That's where you have something called the "Attribute Table", and once more Nintendo went cheap on RAM. Unfortunately this is also the most complicated part of the NES architecture. If you can follow along with me here, you'll have no trouble understanding how the NES works.

Nintendo only set aside 64 bytes of RAM for the attribute table. Since there are only 4 palettes available per screen, you can represent each palette with 2 characters (bits). Again, there are 8 bits to a byte, so you can represent 4 "areas" with each byte of RAM. With 4 areas per byte and 64 bytes to work with, you can set the palette 256 times across the NES scree. So if you divide that by the NES' resolution, that means that each definable area makes up a 16 x 16 pixel area on the screen.

I hope you're still following along. (i'm going to get a bunch of tl;dr's after this)

So what that means in practical terms is that each 16x16 area in the background *must* share a palette of 4 colors (3 if you discard the background color). To give you an idea of the size, this is the same size as one of the question mark boxes in the Mario games.

How does this play into Grand Theftendo? I'm colorblind, so i can't tell if the screenshots abide by the 13 color limit, but it certainly doesn't abide by the 16x16 grid layout for the background palettes. There are a few other problems with the screenshots, mostly involving how the game's sprites are arranged. The NES does not have sprite rotation, what you draw is what you get. If you look at the screenshots, there are cars twisted every which way. While it's possible to store all of those rotations in the CHR ROM, no programmer would ever do it because it's a huge waste of space.

[OdSquad64]

I'm pretty excited about this game. I haven't played my NES in a while, but this thread has definitely given me the itch again. I'm also really enjoying reading about the making of this game and the little incites into programming for the NES and the NES hardware limitations. Keep up the good work guys, I can't wait. Might I suggest a DP seal of approval on the box?

[ProgrammingAce]

When you're making a game, you'll end up having to make certain choices when it comes to content. These choices are usually due to either space, schedule, or budget. Being two dudes making an NES game in our free time, schedule never really comes into play. Budget was an issue in initial planning, it really set the mappers we could use to design the game. As much as we would love to use every advanced feature in the NES, we would have to destroy copies of Castlevania III to do it. When it comes to the NES, space really becomes the largest limitation

To show some of the choices we've had to make, below is an email from myself to Brat. Before anyone becomes concerned that we're gimping the background design, understand that I fully expect that we're going to use the full palette colors. If somehow push really comes to shove and we have to limit the colors on a few of the screens, Brat has already proven that he can draw some really cool backgrounds using only 4 colors.

I've been thinking about the backgrounds...

I assume for the concept art, you've been tiling them in photoshop (or something). That's probably more work then it needs to be, i can work just as well using a sketch like this:

http://picasaweb.google.com/h4ckur/P...00607952104306

And that's a really poor quality example, but i can work with it. Just tell me which tiles go with what number.

As for the colors of the background, that becomes a little tougher. If we want to use more then 4 colors per screen i'm going to have to assemble the tables by hand. It's not hard to do, it's just time consuming. Here's how it works:

You have 4 palettes per screen: 00, 01, 10, and 11. If the first 4 tiles use palettes 00, 00, 01, 10, then the first byte of the table would just be those numbers put together 00000110. Then i would convert that back to base 10, so it equals 6. So 6 would be the first entry in the table.

Now i just do it 63 more times per screen. Like i said, it's not hard but it takes some time. I'd certainly like to do it though because it adds a lot more to the backgrounds.

The only real problem is size. A background is setup by a grid of 224 bytes. Each byte describes a meta-tile of 2x2 tiles. Add in the 64 bytes for each attribute table and we're at 288 bytes per background (plus some extra to define each tile). We have about 9k of space left in the rom (from a total of 32K). If we set aside 3K of space for backgrounds, we can fit in 10 different background designs using 4 color palettes. If we were to limit the backgrounds to 4 colors, we could fit 15 backgrounds in the same amount of space.

The wild card is the compression scheme i came up with to reduce the size of the backgrounds. It's a very simple compression, but it doesn't take up much code to execute. I could do better compression, but that takes more code and more CPU time. If the decompression code takes up more space then it saves, it doesn't do us any good.

- Ace

[Gentlegamer]

we would have to destroy copies of Castlevania III

That would be a crime against humanity.

[MachineGex]

I probably will get killed for this, but if you are only going to make 100 or so new games, killing a few Castlevania's would be worth it. Gotta be over 1,000,000 Castlevania 3 carts, I don't think taking 50 or 100(to make a new game) would even make a dent.

Not that anyone would ever do it. But I personally dont see it being such a huge crime against the gaming commmunity that some think it is.

Back on topic, thanks for the updates! I really admire anyone who can do what your guys are doing. Keep up the hard work and the updates!

[Gentlegamer]

I probably will get killed for this, but if you are only going to make 100 or so new games, killing a few Castlevania's would be worth it. Gotta be over 1,000,000 Castlevania 3 carts, I don't think taking 50 or 100(to make a new game) would even make a dent.

Not that anyone would ever do it. But I personally dont see it being such a huge crime against the gaming commmunity that some think it is.

Back on topic, thanks for the updates! I really admire anyone who can do what your guys are doing. Keep up the hard work and the updates!

Where do you live so we can kill you.