mirror of
https://github.com/elliotnunn/supermario.git
synced 2024-07-02 15:29:33 +00:00
35 lines
1.6 KiB
Markdown
35 lines
1.6 KiB
Markdown
I tried to get System 7.1 (Cube-E) running on top of a NewWorld ROM in
|
|
QEMU. It was really fun. Some observations follow...
|
|
|
|
The NewWorld ROM is recognised as a SuperMario ROM ($077D).
|
|
|
|
The Linked Patches are easier to work with than they look. Just add an
|
|
object file to LinkedPatchObjs that takes advantage of the
|
|
LinkedPatchMacros (InstallProc, PatchProc, leaResident...) and the
|
|
loader will make sure it becomes part of the System. Let there be no
|
|
doubt, the Linked Patches are an integral part of System 7. They should
|
|
be RE'd.
|
|
|
|
Relatively little of the on-disk System deals with hardware directly.
|
|
The ROM kinda-sorta acts as a hardware abstraction layer. Only it has
|
|
heaps of bugs and missing features that need to be worked around. For
|
|
example, I expected hardware interrupts to ruin everything but this
|
|
didn't happen.
|
|
|
|
MacsBug works! It's a bit nicer than spamming the NanoKernel log.
|
|
MacsBug did eventually start giving me nonsense explanations for
|
|
software exceptions. I thought this could be worked around by disabling
|
|
the DR emulator, but this seemingly had no effect.
|
|
|
|
Developing the System *and* the ROM at the same time must have been a
|
|
huge relief for the NewWorld devs. Being unable to edit the ROM directly
|
|
(because my build system wasn't set up) was very frustrating.
|
|
|
|
Things that I couldn't figure out: Balloon Help wouldn't install despite
|
|
much massaging, and this tripped up the Process Mgr. The global MenuList
|
|
was, at a later stage, getting corrupted and I couldn't figure that out.
|
|
Skipping over it, eventually the Memory Manager and Trap Dispatcher
|
|
would die.
|
|
|
|
It was great fun! Definitely worth revisiting later.
|