SquirrelJME/SquirrelJME

# 2019/06/28

## 09:00

Okay so the value being read should be the `seeker` value which is not valid
at all when RatufaCoat runs.

## 09:10

Okay so load pool is exactly the same as SummerCoat, same values and such.

## 09:25

Okay so, I think I saw something in the JAR minimizer. The BootRAM writes
7 fields while the non-boot JAR writes 8 fields. So this definitley might be
why things are messing up. Need to see what is needed. Okay, only 7 fields
are needed. Hopefully this does something?

## 09:36

I need a way to handle out of memory errors. But fixing the non-boot JAR error
made RatufaCoat behave a bit different.

## 10:11

Wait a minute, perhaps in SummerCoat the names of the JARs are placed in the
actual JARs, because it is just a ROM pointer.

## 10:13

Okay just checked and that is not the case. So not sure!

## 14:43

Okay so `__initRamLinks()` is being called with `Invoke 01dc2e54 (`
`1114112/00110000, 16777216/01000000)`. So this looks right.

## 15:17

Okay so the first init ram links read ended up being zero.

## 15:24

BootRAM is being initialized but it seems no values are being written to
anywhere.

## 15:33

Okay so it seems the initial value is not being written to BootRAM at the
given address?

## 15:59

I am stupid, I was not storing the read value in a register.

## 16:42

Okay so the boot library is not valid in either VM, so the value must be
written wrong.

## 16:49

No, just the wrong offset in the ROM was being read for the main class.

## 16:55

Actually the original value is correct.

## 15:56

I just have to offset the value correctly I believe.

## 19:30

So the RetroArch engine does not need to be updated as much, I am going to
put in ROM stuff that is built-in like an initial JAR program.

## 19:52

Okay so I do not want to copyrom anymore, but it just gives me invalid
argument on init. So wonder what causes this.