2018/10/13

10:28

Took a peek and the monitor for Reader is only used for skip so that is not too bad at all.

10:35

I should have a fallback if the specified system encoding is not valid.

10:36

Maybe not.

12:38

The hashmap has negative divisions used, so that means I need to mask the lower value to remove the sign bit.

15:43

Looks to be that protected fields are unable to be accessed from subclasses so need to check that out.

15:49

Okay so, it is UnmodifiableMap.__SetIterator__ and it is accessing the parent class iterator. So this needs fixing.

16:01

Looks like protected and package private fields and methods could be accessed if they are in the same package. Oh wow, this is something.

16:05

Wow, I learned something new! Or at least something I really forgot about but now I know. This definitely changes how I create fields and methods.

16:07

Well, this is quite the revelation of something I learned long ago but completely forgot about. But it works and actually kind of makes things better I guess.

19:02

I paired with a friend on SquirrelJME and I fried her brain with exception handling, heehee.

19:20

So for integer overflow parsing, not all values reach overflow so that is something: RV int:31 int:35.

19:23

I need better reporting.

20:27

So now that I have exception handling, I can pretty much run any program now apart from anything needing the few instructions not implemented. Although there is the main library to consider. I suppose what I should do now is make it easier to run tests by implementing the stuff needed for suites and tasks and otherwise. Although I do still kind of need threads and such. But bulk running of tests might just be the next thing to do. Although I can really branch out to many other things now as needed. I could though just move onto double SpringCoat, although I would really need file handling for that to be doing anything useful. I could try to run Squirrel Quarrel, but that would be far too much work and would take some time. So I guess the next goal is implementing SpringCoat on top of SpringCoat. It would be far slower and will likely just use tons of memory, but if it became self hosting it could probably run everything. If I wanted to use the host VM libraries to provide suites and such like as a forwarding scheme, then that could work. Although I would need a way to list the contents of a JAR in order to more effectively handle it. Although actually, no I do not need to do that because the current code just uses resource as stream for classes and such. SpringCoat on SpringCoat like that would only work if the original classes were still available, which they might not be. But it is worth a try to be honest I would say.

23:06

Okay so I really think the next thing to do would be to implement suites and tasks, this way I can launch things. I really need to be able to do bulk tests and also support things like performing launching from a launcher and such.