Bankswitching, And Could It Have Saved Coleco?

[ColecoFan1981]

If the developers Coleco had hired to do all of our favorite games -- such as Donkey Kong and Smurf Rescue -- would have learned about how bankswitching works, would it have saved Coleco from the Video Game Crash of 1983?

 

A single 8K EPROM, through bankswitching, would yield 16K.

 

In theory, since the largest a cartridge went out with was 32K, then, in theory, through bankswitching of each chip, this would have been doubled to 64K. Had any CV cart been made with bankswitching, this would have at least delayed the release of the ill-fated Adam computer system.

 

~Ben

Edited May 15, 2023 by ColecoFan1981

[PkK]

I'm not sure I understand what you are trying to say about the technical aspects of bankswitching. But it won't get you 16K from an 8K chip. You could map 16K into the space used by an 8K chip, but you'd still need a physical 16K of (E)PROM.

Unless you've used up all the 32K of program space available without bankswitching, bankswitching provides no benefit.

 

As we see from historical games, which typically don't even use the full 32K, the size of the EPROM wasn't the bottleneck for most game development (the price of (E)PROM and the development toolchain were the relevant limits). And I really don't see how this could have "saved" Coleco: to me, the size of the cartridge memory seems quite unrelated to the causes of the Video game crash.

[ColecoFan1981]

2 hours ago, PkK said:

I'm not sure I understand what you are trying to say about the technical aspects of bankswitching. But it won't get you 16K from an 8K chip. You could map 16K into the space used by an 8K chip, but you'd still need a physical 16K of (E)PROM.

Unless you've used up all the 32K of program space available without bankswitching, bankswitching provides no benefit.

 

As we see from historical games, which typically don't even use the full 32K, the size of the EPROM wasn't the bottleneck for most game development (the price of (E)PROM and the development toolchain were the relevant limits). And I really don't see how this could have "saved" Coleco: to me, the size of the cartridge memory seems quite unrelated to the causes of the Video game crash.

Donkey Kong, the pack-in game, could have benefited from bankswitching (whether 24K or 16K) to allow for some missing elements from the arcade original (i.e. fireballs on board 1; springs on board 3), if indeed the entire 24K/16K was used up.

 

~Ben

Edited May 15, 2023 by ColecoFan1981

[+5-11under]

First of all, what Pkk said.

Second, I'm guessing they would have known about bankswitching, from Atari, who had been using it since at least 1981.

[Captain Cozmos]

The only time I needed to use Bank Switching is when I went from Great Western to Chase

[+evg2000]

https://wiki.gamehacking.org/Bank_Switching

[Pixelboy]

Coleco made their decision to implement games bigger than 32K via the ADAM, through the Digital Data Pack medium and the ADAM's 24K of RAM. This made sense at the time, because back in the early eighties, ROM chips were expensive, and Coleco evidently preferred to stick with ROM sizes below 32K for cartridges, as most of their cartridge games were 16K, and a few were 20K/24K. Only a handful were 32K and they came at the tail end of the ColecoVision's short commercial life.

 

Had the ColecoVision lasted a few more years (and the ADAM not been Coleco's ill-fated champion product) bigger ROM chips would have become more affordable and then bankswitching in carts would have been a viable solution. Ironically, what would have happened next is what happened decades later via the ColecoVision homebrew community: Lots of MSX ports. Coleco had already begun this trend commercially with Antarctic Adventure and Monkey Academy, which were clearly MSX ports, and this trend would have likely continued and expanded with new bankswitching solutions for ColecoVision cartridges. Adding more RAM via the ColecoVision's expansion port would have been unavoidable at that point, so Coleco would have been forced to release the Super Game Module.

 

But would this really have allowed the ColecoVision to survive the Video Game Crash of 84? I seriously doubt it. Retailers at the time just lost faith in the video game industry as a whole, and just wanted to clear all the home console games from their shelves. Better cartridge technology would not have changed that, especially with consumers moving over to home computers anyway. Pirating games on floppies was all the rage, back then. 

 

Edited May 15, 2023 by Pixelboy

[zzip]

9 hours ago, ColecoFan1981 said:

If the developers Coleco had hired to do all of our favorite games -- such as Donkey Kong and Smurf Rescue -- would have learned about how bankswitching works, would it have saved Coleco from the Video Game Crash of 1983?

 

I doubt it.   The industry grew too fast in the early 80s and a correction was in the cards.    I don't think there was a technical fix that could have stopped the carnage.

[mickster]

only thing that would have saved coleco was two things.

 

1)  Somehow made the ADAM computer reliable.

2) Somehow make the ADAM cheaper to compete with Commodore.

 

Even  then, that would have bought them about 2-3 years tops.

[MrDave]

The coleco has 4 rom slots of 8k for a total of 32k  each slot could have a 64k eprom for a total of 256k but the ram would stay the same and limit the games.. Nintendo carts work easy with bank switching as the sprites are mapped into rom and can be cycled to create neat effects in games like batman and battletoads etc.. Nintendo reversed the crash by forcing games to be approved.. so more expensive bank switched games might of been great but the cheap games would still bring things down.

[Tursi]

It all came down to cost. We sometimes forget how expensive memory (RAM and ROM both) was back then.

 

[+nanochess]

Spy Hunter was 32 kilobytes, released in 1984. And Intel just introduced the EPROM 27256 in 1983!

 

Bank-switching was applied to Atari because it only could access 4K of ROM, and 8K memories appeared around 1980 (also probably the cause that the IBM PC BIOS was 8K). An Intel manual for 1981 shows only memory capacities upto 8K.

 

Bank-switching couldn't save Colecovision at the time, because it was designed to support 32K of ROM directly without bank-switching. The problem was that memory was pretty expensive!

 

I mean a game like Mecha-8 requires 128K of ROM, that is FOUR 32K EPROM memories. At the prices of the eighties, it would have been prohibitive.

 

Coleco upped the RAM memory as it was going down in costs, but still tried to go with tape as it was the most inexpensive, but disk was entering fast to the market.  It was a perfect storm, all looked like perfectly good decisions, but things changed too fast.

 

[Captain Cozmos]

31 minutes ago, nanochess said:

Spy Hunter was 32 kilobytes, released in 1984. And Intel just introduced the EPROM 27256 in 1983!

 

Bank-switching was applied to Atari because it only could access 4K of ROM, and 8K memories appeared around 1980 (also probably the cause that the IBM PC BIOS was 8K). An Intel manual for 1981 shows only memory capacities upto 8K.

 

Bank-switching couldn't save Colecovision at the time, because it was designed to support 32K of ROM directly without bank-switching. The problem was that memory was pretty expensive!

 

I mean a game like Mecha-8 requires 128K of ROM, that is FOUR 32K EPROM memories. At the prices of the eighties, it would have been prohibitive.

 

Coleco upped the RAM memory as it was going down in costs, but still tried to go with tape as it was the most inexpensive, but disk was entering fast to the market.  It was a perfect storm, all looked like perfectly good decisions, but things changed too fast.

 

I am dabbling on something now with an MSX game that on face value should work on a Colecovision bank switching or not.
The issue I found is that you can add more to the roms all day long.  Switch the upper bank, new levels ect... but the 1k memory limitation was the real problem.

The game I decompiled to see if it port-able I have working on an ADAM but not an unexpanded Colecovision because the playfield is huge and uses 2k of upper memory to map it.
I should change that to pre-working.  I have it loaded, graphics but I have to remap some memory because it overlaps EOS.  It also has to have sound changes.

 

Point is, in my opinion, bank switching roms may give you extra content but not necessarily better game play.