adventures in game design

[adekto]

lets begin the chapter of games design and its development.
so i have had a brainstorm on many games for the gamebuino, some are new but some are refitted old designs here are a few

ghost mansion

this game has an actual plot and is one of my older concepts, i'm thinking of making this game first, the main focus in this game is to get all the artifacts that are spread around the mansion.

hedge maze

this is a little more ambitious, this actually started as a dungeon crawler but i soon realized that it would be over complicated and not wanting to have a manual bigger than the entire console i think i found a compromise in the hedge maze, i see it more of a tech demo since u can mostly walk around a maze and find the exit, although the idea of multiplayer would be cool here.

ashens game colection

well its ashens so what did u expect? main idea here is to have versions of popstation games, these are more as a joke

i hope u like my idea’s and i like to take any suggestions

[Myndale]

You do some pretty impressive work with limited pixels!

Personally I'd like the see and/or work on something like the hedge maze idea but a dungeon crawler instead i.e. something along lines of NetHack. Takes me back to the days of SSI and their early D&D games like Eye of the Beholder etc...great games at the time, and perfectly within Gamebuino's capabilities gameplay-wise.

[adekto]

thank you very much and i like the idea of a dungeon crawler to, just not sure of the limitations on the hardware, say for example enemy variety it needs a lot of code and image data, allso different walls or walls with doors or items on the floor in the distance can increase game size very rapidly.

heres an image of the first idea that ended up to be hedge maze


by all means go ahead and make a roguelike although i like to see much content, and i'm not sure how u can do a good inventory/battle system unless its very text based

[Myndale]

I did some very rough back-of-the-envelope calculations and I think it's doable if you restrict the engine to displaying no more than 4 "blocks" deep and design your levels accordingly. You could probably fit 8 different walls types in 4kb of PROGMEM, 8 different full-screen monsters in another 4kb and then anywhere up to 32 or so objects in another 4kb (potions, keys etc). Add another 4kb for extra stuff and you'd need a 16kb buffer in your game which you'd then reload from SD card for each level. Tricky, but doable.

That's what I love about Gamebuino, it's easy enough for beginners to start playing with immediately but restrictive enough to provide a decent challenge to those willing to push the hardware to its limit.

[adekto]

I was wondering if theres a way to optimize the images
Like breaking up bigger images (like the walls) since i feel theres allot of blank data in for example the ashens image?

Or what about prosedural textures say for example the wall is a black plane used as a mask to then run a texture wich is like some alorithm

Also theres that option of games opening games from sd with some way to carry over data ( shared saves) and have different levels tiles and enemies

[Myndale]

There are plenty of compression algorithms that would be useful in an engine like this but the smaller the image gets the worse they perform. Also the better performing the compression algorithm is the larger (and slower) the decompression algorithm has to be, which starts eating into code memory and eventually negates the benefits. I've done a lot of work professionally with compression algorithms over the years and in my experience images at this scale tend to be best suited to run-length-encoding like what the old PCX format used:

http://en.wikipedia.org/wiki/Run-length_encoding

At low resolutions a simple RLE scheme will typically get you somewhere around a 50% saving, which is still a very good result when every little byte counts.

[adekto]

i see this compression look simple to understand
hoping this will be build in the GB libs

also i have heard the idea of "video" on the lcd wich sound a bit weirder and i asume the compression to that one is a bit more involved since wel u got to fill 20 or atleast 12 frames a second for smooth transistions and not run out of memory

[adekto]

testing how that tileset works and how much tiles are needed
also im looking into 3rd person perspective if u can see the viewport is one tile behind the player curently i put a crappy visual indication what oriantation u are in


any ideas or sugestions are much appreciated

[rodot]

No no no I'm not crazy when I talk about video!
The frames will be stored in the SD card, so there is virtually no memory limitation. Someone already did that and managed to make it run at 12 FPS : https://www.youtube.com/watch?v=FJBNF5x1_kA
As I really improved the screen perf compared to the library from adafruit, I guess we could even access higher FPS.
I already did a program in Processing to convert any video file in a 84*48 monochrome video using dithering. Now I have to save it to the SD card and play it back.

[Skyrunner65]

hedge maze

this is a little more ambitious, this actually started as a dungeon crawler but i soon realized that it would be over complicated and not wanting to have a manual bigger than the entire console i think i found a compromise in the hedge maze, i see it more of a tech demo since u can mostly walk around a maze and find the exit, although the idea of multiplayer would be cool here.

This would be fantastic in multiplayer, either as a race to get to the middle, to kill each other, or to find all the pieces to defeat a monster that is chasing you.