bool_hack?

[chris]

I'm puzzled ... I was curious about this:

#define bool bool_hack

If I grep the entire source directory for bool_hack, this is the only occurrence I find. It this built in for C or something? It seems to be

typedef char bool_hack;

but I can't find that code any place.

Thanks,
--Chris

[chris]

Yeah, looks like this has been discussed already ...

http://angband.oook.cz/forum/showthread.php?t=3276&page=5

[chris]

I am so puzzled by this. If I change code to:

#undef bool
#define bool bool_hackz

It still works. And I *know* that bool_hackz is never defined It must be some compiler voodoo in the case that somebody #defines bool to something ...

Of course, this gives the expected type undefined:

#undef chris
#define chris chris_hack
...
chris x = TRUE;

Weird.

[takkaria]

I am so puzzled by this. If I change code to:

#undef bool
#define bool bool_hackz

It still works. And I *know* that bool_hackz is never defined It must be some compiler voodoo in the case that somebody #defines bool to something ...

Of course, this gives the expected type undefined:

#undef chris
#define chris chris_hack
...
chris x = TRUE;

Weird.

Think it through. #define bool bool_hack comes before the typedef char bool; -- so the 'bool' type is never defined, bool_hack is.

[chris]

Think it through. #define bool bool_hack comes before the typedef char bool; -- so the 'bool' type is never defined, bool_hack is.

Oh, OK. My greps on "bool_hack" missed the

typedef char bool;

That makes sense now ...

[chris]

Yeah, and I see the typedef comes just 13 lines later ... tunnel vision.

Thanks for clearing that one up

--Chris

[Nick]

I hope you retained bool_hackz

[chris]

I hope you retained bool_hackz

Hah! No ... I'm tempted now, though

[takkaria]

Hah! No ... I'm tempted now, though

All this mess will hopefully go away pretty soon when we move to C99 types. It's a shame, I like Angband header files being fun little logic puzzles...

[chris]

OK, I worked really hard on this, and I think I found the problem. There is some funny looking code in my copy of developer.c, but I'm not quite sure how to fix it. Any ideas?

/* developer.c */
...
/*
 * Figure out what went wrong
 */
void assign_blame(void)
{
  if (randint1(1000 * 1000) <= 0)
    msg_print("Ooops!  My bad!");
  else
  {
    switch(randint1(8))
    {
    case 1:
      msg_print("It's a compiler bug!");
      break;
    case 2:
      msg_print("It's the last developers fault!");
      break;
    case 3:
      msg_print("It's compiler voodoo!");
      break;
    case 4:
      msg_print("Damn Micro$oft!!!!");
      break;
    case 5:
      msg_print("User error.");
      break;
    case 6:
      msg_print("Works as designed.");
      break;
    case 7:
      msg_print("I can't reproduce, so you must be imagining things.");
      break;
    case 8:
      msg_print("Works as designed.");
      break;
    case 9:
      /* At my last performance review, it was pointed out that I don't own
       * up to my mistakes.  This should shut my boss up for a while ...
       */
      msg_print("I'm sorry, that's clearly my fault.");
      break;      
    }
  }
}
...
/*
 * Handle developer state every 10 turns
 */
void process_developer_aux(void)
{
  ...
  if (developer_is_confused())
    assign_blame();
    
  if (bug_count > old_bug_count)
  {
    assign_blame();
    old_bug_count = bug_count;
  }
  ...
}