Roulette "Incident" at the Rio that never happened.

Yeppers!!

As expected, The Rio finally issued a press release admitting that no such incident of seven hits of the number 19 in a row took place. They claim the table was open and no repairman was working on the display panel. The Rio glossed over any obligation on its part that if it provides a display it must provide an accurate one. The Rio press release made no mention of deferred maintenance in this or in other areas and made no mention of the credit crunch CET (nee Harrah's) is facing. I'm sure that such deferred maintenance will continue until either "The Miracle" takes place or "The Default" takes place. In the meantime the company will keep their press releases and their whitewash handy. Undoubtedly they will also keep trying to sell off assets even at rock bottom prices in an attempt to stave off a credit crisis.

The poker player who snapped the photo of the display and tweeted it into internet immortality was simply wandering by and noticed it, he was not actually playing and he should have known that any such sequence would have resulted in a frenzied table.

Extremely rare events can still be unacceptable.

The Quantas A330 uncommanded dive from FL37 which propelled dozens of unrestrained passengers into the overhead three years ago has been blamed on a computer glitch and the various press releases are stressing the overall reliability of the sensors and the fly by wire computer system. Yes, its a rare event but the computer code that allowed action based on inputs from one of three sensors without doing any data validation first was an unacceptable programming error. The malfunctioning sensor's data should be compared with immediately prior values and with values from other sensors before being acted upon. And a prior warning to the cockpit of sensor disagreement should be sent. Its truly rare, but it happens and a computer that automatically accepts and acts upon such errors is not safe. All those millions of miles flown and thousands of hours of operation without incident were lucky, but not safe.

The algorithm compared two angle-of-attack indicators. If the readings were close, it averaged the two values and declared that the value to be passed on to further flight control systems. If the two values differed greatly the algorithm did not enter into any evaluation mode to determine reliability or data quality it simply discarded each angle of attack indicator's data and defaulted to the angle of attack determination that had been made 1.2 seconds previously. Unfortunately one of the two angle of attack indicators starting releasing spurious values every 1.2 seconds and the algorithm defaulted to erroneous values that made the flight control computer think the plane was at a very high angle of attack and that a stall would take place if the nose were not lowered immediately.

The pilots were not able to counteract the error since the fly by wire system over rides any pilot input that is unsafe.