| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
| |
Discuss a second objection that is based on a misunderstanding
of the meaning of the third error value.
|
| |
|
|
| |
Spotted by P. Olcott.
|
| |
|
|
|
|
|
|
|
|
|
| |
Carefully separated "procedure" and "function" terminology, to avoid confusion.
Removed claims that the space of functions can be searched to find an
equivalent; this is not tractable due to Rice's Theorem, and not germane to the
proof.
Formalized in more detail what the third "error" value of the halt3 function
means.
|
| |
|
|
|
| |
The error_decider returns strictly a boolean value.
(Reported by P. Olcott.)
|
| | |
|
| | |
|
| |
|
|
|
| |
This option allows one to just generate a standalone
program for testing a decider.
|
| |
|
|
|
|
|
| |
The deciders could cheat by defining macros that affect
the behavior of later code. So we put them after the main
function and in particular, the halting_decider is placed
last in the program so its macros are of no consequence.
|
| | |
|
| |
|
|
|
|
|
| |
In cases where the decider is wrong, TRUE and FALSE are
not both viable answers for a specific test case, but
rather for the class of test cases which share the
same error_decider.
|
| |
|
|
|
|
|
|
| |
Generate an additional program which wraps just the decider
in a simple driver main which lets the program be run
independently, with both its inputs specified through
the command line. The result of the decider is reported
to standard output.
|
| |
|
|
|
| |
- Provide prototypes for the deciders, so they are diagnosed if incorrect.
- Add some compiler language and diagnostic options.
|
| | |
|
| |
|
