Apollo 11 glitches: the alarming 1202, the 4 sec fastforward and the soft landing bug

It's forty years on from the moon landing, the culmination of the work of 400,000 people. I can remember as a kid seeing the hazy picture on TV (One of the few bits of TV I saw as a kid, as I grew up on the Pacific island of Nauru but must have been in Australia on holiday). While everything appeared relatively fine to us, there were some issues with software that varied from potentially critical to positively funny. Their computer (click the link for a simulation) resembled a calculator, with numbers being entered as nouns and verbs, and number output being looked up in a booklet for explanatory descriptions. During the landing, the alarm sounded and "1202" displayed then the computer rebooted. Neither Armstrong or Aldrin knew what this meant. While the alarm repeated alternating with a "1201", so did the command from Mission Control, "It's Okay!" The same error had been encountered in trials but had stopped by itself. Luckily, after several reboots, the alarms also stopped. This was later found out to be caused by an incorrect manual instruction to leaving an upgraded radar device (needed for the later departure), which had overloaded the computer once the landing started. The manual was updated and the bug fixed in later missions.

Next the astronauts looked at their surroundings on the moon's surface. (Even before the computer reboots started, Armstrong had said to Aldrin, "I think we’re gonna be a little long", as he realized they seemed to be out of position in their approach to the moon) They noticed they were passing landmarks 4 seconds earlier than they expected. When the lander slowed and straightened preparing for the descent, the expected landmark of an empty crater was a rocky plain instead. That was when Armstrong turned off the autoplot and took over the controls (and doubled his heart rate!) looking for a safe landing spot, while Aldrin called out their speed and height. Soon he added another number "90 seconds", until they ran out of fuel. While Armstrong was prepared to risk an out-of-fuel landing, they touched down with an estimated 16 seconds of fuel left. (Later analysis showed unexpected gravitational variations had confused the autopilot, which was corrected for later missions) While the manual landing was perfect, it went against the expected plan. I can remember as a kid thinking "one small step for man, one giant leap for mankind" was a strange quote when he seemed to jump down onto the moon not step off. The expected rough autopilot landing was meant to jolt the ladder free, but Armstrong was too smooth! The ladder was still partly up and he had to jump!

There was an independent software test team who had received recognition awards for their work, and they were asked to run some new tests soon after the landing. While we still test in a similar fashion on some projects(!), the development team were extremely limited in what they could fix. While software teams today can experience the same sorts of glitches, the hardware is unrecognizable. The computer consisted of 36000 strings of wire with iron disks. Closer to chainmail than the computers of today, a major bug found in the lead up to the launch had taken weeks to correct! Comparing it to today, it is amazing that the computer they used had significantly less capacity than a electronic greeting card!

As well as the landing glitches, there had also been data and communications outages from Houston to the ship. According to the landing procedures, any of these events should have triggered an abort, but confidence in the combined team abilities kept them going. Even with the biggest team in the world, and the most comprehensive planning, gut feel and seat-of-the-pants got the job done!

{NOTE: this post was updated as I discovered more first hand accounts of what happened)