Sat Jul 18 17:24:59 CEST 2009
rearranging throw and abort
Quite a marathon. The internal interfaces are getting more brittle,
which helps a lot for manual refactoring like this.
Let's look at all the demo files to see if they work.
Looks like there are quite some bugs left though.
These seem to be all C-level bugs, so let's first install some proper
gdb startup routines.. The idea is that the global abort will run a
Problem: it's only known whether there are any exception handlers
after the longjump. Todo: move this to before.
The advantage of the current form is that it is relatively easy to
JIT into a subroutine threaded Forth.