Max cartridge capacity

[MCes]

what about a cart full of RAM that on startup will copy a .PRG file from a hidden ROM to the RAM?

May be a new cheap standard to making cart from own .PRG software/games

https://sleepingelephant.com/ipw-web/bu ... php?t=9950

[kamaleon70]

I think this is what I was looking for. https://www.tfw8b.com/product/commodore ... k-rom-pcb/
Having also additional 3k of ram would be perfect.

[Wilson]

IMO when you start talking about +35K configurations, the "load" time sort of kills the cartridge experience. Not sure if it would be feasible to build a solution does the ROM to RAM copy any faster?

[MCes]

The ROM-RAM copy take less than 1.5-2s,
for example the prototype of "DREAMcard" load activity from startup to starting "Cheesy Trials":
https://youtu.be/tAGjg8AVstk

The VIC 20 used has a special kernal that is able to ignore the cart auto start, in this manner is evidenced the real time needed for the "DREAMcart" software startup without summing it to the natural HW RESET delay.

[Wilson]

Yes, I use a similar approach for a FE3-based project of mine, and, maybe I’m just self-critical, but 2 seconds feels like a long time for a cartridge, which I would expect to need no load time at all.

I’m willing to believe the average end user does not care all that much.

[MCes]

I estimate approximately 1.5s a full system ROM-RAM copy.

Remember that at startup a reset HW is generated, and its length depends on 10uF electrolytic capacitor (1s? 5s?),
then the kernal test the RAM banks and if RAM is expanded it take a lots of seconds....

If Commodore thinks the normal boot delay is acceptable, I might think that even a maximum of 1.5 seconds to boot a cartridge game might be acceptable...

[tokra]

Or just pop up a simple text-screen saying "Booting..." and maybe flashing the border to acknowledge to the user that the system is starting. For 2 seconds though this seems like overkill.

[Wilson]

then the kernal test the RAM banks and if RAM is expanded it take a lots of seconds....

Cartridges bypass the RAM test though; another reason I feel like the expectation is for them to be super fast booting.

Or just pop up a simple text-screen saying "Booting..." and maybe flashing the border to acknowledge to the user that the system is starting. For 2 seconds though this seems like overkill.

I made a splash screen with a simple progress bar for my program (it does some code generation too and takes more like 8 seconds). I don't dislike the "Booting..." idea; I think some sign of life from the machine is always good.

Anyway, if there's no easy way to speed up the copy in hardware- so be it. If someone really needs to save the copy time, they can probably modify their program to run from ROM. I'm not a hardware guy, so I didn't know if that was easily doable.