What's absolutely brilliant about the development of the original Game Boy is that at the time, Gunpei Yokoi's team was facing significant incredulity. "Why play games when you're on the bus, or on the toilet? It's going to be so uncomfortable. Play at home on you couch in front of the TV with family and friends. Play at the arcade for cutting edge experience. Nobody wants to change batteries and see just a few shades of grey".
Their vision was to take "withered technology" and package it in a simple to use device, with simple, short games.
That team single handedly started mobile gaming. Much respect to everyone on it.
> Gunpei Yokoi's team was facing significant incredulity. Why play games when you're on the bus, or on the toilet?
> That team single handedly started mobile gaming.
No. This is revisionist history.
Us kids in the 70s already had portable LED games by Mattel that we played on the bus and toilet. I had at least 6 of these, maybe 8. Here’s the earliest one, from 1976:
“Mattel stated that the game immediately sold out.”
None used cartridges but they were so cheap and small (esp. the LCD games), it didn’t matter. You could own a bunch without breaking the bank.
Such games are now called “handhelds” to distinguish them from the GameBoy generation of portables, but we did not call them handhelds back then as I recall.
My point is a successful market already existed in the 70s and early 80s. Nintendo didn’t do that. They evolved what already existed by shrinking cartridge-based consoles into a portable form factor. Both were already successful, established markets before.
It's not "revisionist history", just a different take than yours.
I detailed in another answer that the Game and Watch and other past handheld didn't offer a console-like experience or one that you'd directly compare to a modern phone game.
I don’t know anything about Game and Watch. The mattel LED and LCD games I’m talking about were mobile predecessors with many of the same genres of today.
There’s a reason it was the DMG. Dot Matrix Game(?).
It was the first really successful portable system with interchangeable games and a dot matrix screen as opposed to custom LCDs per game like G&W.
The Milton Bradley MicroVision came first around 1980. Nintendo referenced it as a precursor. But the screen was 16x16 pixels so ‘graphics’ is kind of misnomer.
I think I remember seeing something once where the screen was included with the game, because it was a static LCD sort of like a Game and Watch game.
The GB was in the right place, at the right time, from the right company, with the right hardware and software, at the right price point. Instant success.
For anyone old enough for this to make sense but who hasn’t seen a Game and Watch game: they worked just like the cheap Tiger Electronics LCD games you could get.
Just like a simple digital watch, everything that could be on the screen was designed upfront and then the individual bits could just be shown or hidden.
Nintendo eventually made three collections of the Game and Watch games for the GB. Each collection was a ton of fun with each game available in classic (faithful) or updated (graphics replaced with sprites, same gameplay) modes.
There's a fighting game (I think a port of Street Fighter II) in this style and it's just wild how they make the static screen elements work with that type of game.
Whoops, I see Game Boy Color was added in there, but he forgot to revise the part about the cartridge size being only 1MB. Some GBC games reach 4MB in size.
edit: There's also an 8MB game about Trains, with some short videos and hi-color images in there.
And to be fair, I don't think there was anything stopping a DMG game from reaching 4MB. The pak has the same aperture size of basically 32k either way, and the bank switching for either 4MB or 1MB is on the cart itself.
Right. I don’t think Nintendo had a mapper to do it at the time. I doubt third parties did.
But if you wanted to let a DMG address 2 gigs of memory 32k at a time you could do it.
It’s a little hard to imagine a game needing > 1MB if it’s just B&W. But I suppose you could put a ton of text in or do some kind of demo-scene style ‘video’ playback. But no one would have for cost reasons back then.
None of the standard mappers used in Game Boy/Game Boy Color games provide 32K switchable banks. They all provide a fixed 16K bank at the beginning, and a switchable 16K bank afterwards. If there are 8 bits of bank selection, the maximum cartridge size becomes 4MB, but some mappers allow more than 8 bits to select a bank.
One of the common mappers even enforces that you may not pick bank 0 for the switchable bank, picking 1 instead.
Mappers found in bootleg multicarts do allow you to change the base address and size of the game, which will move the fixed bank.
That would be levels and levels of indirection because 32K can be loaded at once and to address 2G in 32K chunks you need over 32K just for the address of the block. And then you need a part of the code loaded that actually does that swapping. You would need at least 3 levels for every memory address. Each 32K chunk would have its own address that would be loaded from two 32K chunks, that’s how giant its address is. And then that block would be loaded in.
I think that it would be literally impossible to have random access memory of 2G.
I'm not sure what you mean. 4gb is 31-bits (<4 bytes) of address space. If each page had 32K of address space (15 bits), that would require 64K different pages, which just requires the remaining 16 bits of the 31 bits to identify it. So the block address is just 2 bytes long, not 32K bytes.
For these paged systems, you typically swap in the the page you're interested (iirc, gameboy did it by writing the page number to particular memory addresses) and then work with it for a while, and often execute code directly contained in it, which will eventually lead to loading the next page you need when you need it.
A single fullscreen image for game boy is almost 6KB. If you were making something pretty you could have several images per level and go over a megabyte without much effort.
You can’t individually address every pixel, you have to use tiles. So unless you’re using a mapper chip that could swap them out on a per scan line basis you have much stricter limits.
Yes you have to use tiles, but I'm not sure what you mean by "much stricter limits". The game boy can load 384 tiles at once, which is enough to cover the entire screen without duplicates.
You can't freely mix the first 128 with the last 128 but that's easy to deal with.
The Game Boy is by far my most favourite console of all time. I’m currently writing a game for it in ASM (a pinball-dungeon crawler). The hardware is simple to understand, and the restrictions require you to be creative.
There is a whole book dedicated to the game boy by Florent Gorges but sadly I don’t think its translated into english. I got them in Spanish. Unique interviews of the creators.
There is a variation on the new-pcb and components approach called the "Ultra Boy Colour" which doesn't require any parts from an OEM Game Boy Color.
It can use a clone CPU that was manufactured (until recently?) for the "GBBoy Colour" (and could also be pulled from the sort of failed SGB-like clone called the "Extension Converter for GB").
> I redesigned the Gameboy Color from the ground up
> based on the premise that it wouldn't require any
> components to be harvested from a real GBC, so it
> would be compatible with both the original GBC and
> clone hardware taken from a GB Boy Colour- since
> the KF2007 clone CPU from the GBBC is practically
> a drop-in replacement for the CGB CPU found in a
> real GBC- no software compatibility issues at all.
If you don't mind your GameBoy being as fast as a supercomputer when it came out, then I think you could get an ARM SBC running a GameBoy emulator on Linux rendering into an OLED screen for less than $150 and a bit of your sanity.
If you're looking for hardware fidelity, I don't think it'd be in the "democratised" price range if you want to build a single unit that fits in a handheld form factor.
I think you’d at best be buying parts taken from old Game Boys and putting one back together. I think I’ve seen an FPGA Game Boy implementation somewhere, which may be the closest you’d get.
It’s a simple architecture that works directly with the hardware, without an operating system. That’s a good gateway to embedded programming, and it’s fun.
Their vision was to take "withered technology" and package it in a simple to use device, with simple, short games.
That team single handedly started mobile gaming. Much respect to everyone on it.