M110 and M29 rework

[AnHardt]

Add 'ok' output for M29 actually closing the file.
M29 now with ADVANCED_OK, if defined.
Handle M110 when received without linenumber/checksum.

Tested with Octoprint 1.2.9, RH V1.6.1, Pronterface 03-Feb-2015.

Proposal for a fix of #3055

[AnHardt]

   M110       -> ok    -> next line unchanged
   M110 Nx    -> ok    -> next line = x+1
Ny M110    *c -> ok    -> next line = y+1
Ny M110 Nx *c -> ok    -> next line = x+1
My M110    *w -> error -> next line unchanged // checksum mismatch, Last Line:
My M110 Nx *w -> error -> next line unchanged // checksum mismatch, Last Line:
My M110       -> error -> next line unchanged // No Checksum with line number, Last Line:
My M110 Nx    -> error -> next line unchanged // No Checksum with line number, Last Line:
   M110    *c -> error -> next line unchanged // No Line Number with checksum, Last Line:
   M110 Nx *c -> error -> next line unchanged // No Line Number with checksum, Last Line:

[thinkyhead]

In our discussion on #3055 we both generally concluded that complaints about M29 are probably now outmoded. You can double-check, but I'm pretty sure that M29 is already fine and already sends "ok" because it goes through the normal command processor in process_next_command, it is recognized by the switch(), and it does fall through to the end where process_next_command calls ok_to_send().

[Blue-Marlin]

@thinkyhead

if (M110) {
  char* n2pos = strchr(command + 4, 'N');
  if (n2pos) {
    npos = n2pos;
    *npos = '\0'; // don't process this 'N' again!
  }
}

would ruin the checksum.

[thinkyhead]

@Blue-Marlin Yes, I figured that out a while ago.