Implements Communication Error Bit, resolves #20

[daniel-j-h]

For #20.

@kent-williams @dcyoung can it happen that the sync bit and the error bit is set? (or really: multiple bits, not specific to those two) Also any insights to error codes for other bits? Or are they unused on the firmware level, too, at the moment?

[dcyoung]

You can definitely have a sync bit AND error bits set on a single DATA_BLOCK.

See the last post #27 for a better description of the error codes. There is at least one error code value we may have to avoid producing, so instead of relying on individual error bits to separately indicate different issues, we may want to just consider bits e0-6 as a single 7bit integer error code, of which we avoid using 1 particular error code value. This will also let us simplify the checking to a single error code:

    // Consider the DATA_BLOCK contents...
    // check that the error bits (e0-6) are all 0 (ie: DATA_BLOCK contains NO ERRORS)
    if ( !(responses[received].sync_error >> 1) ) {

      // Check if DATA_BLOCK is a sync reading (ie: first reading of a new 360deg scan)
      // & if we've already been gathering DATA_BLOCKs
      if ( responses[received].sync_error == 1 && received > 0) {
        // Note the end of the scan
        last = received;
        //FIXME: add check that "last" does not exceed SWEEP_MAX_SAMPLES
        
        // BREAK to return the scan (NOT including this DATA_BLOCK)
        break;
      }
    }
    else {
      // some error
      //FIXME: Handle each type of error code and shutdown gracefully if need be. 
      switch(responses[received].sync_error >> 1 ) {
          case 1: // communication error
          case ....
       }

[daniel-j-h]

Thanks - I somehow forgot about this. Changed it to an error check by shifting out the sync bit.