This should give every coder a moment of pause.<p>Consider. Here we have a $2.5 billion interplanetary mission that is easily the most ambitious space exploration mission since the Apollo project. And it's using a completely new and untested landing system. Only bits and pieces of the landing system have been physically tested, and not even remotely all together. Because it turns out it is rather non-trivial to simulate an atmospheric reentry from an interplanetary trajectory, in the Martian atmosphere, under Martian gravity. In the end it came down to relying on our prior understanding of the Martian environment and simulations and measurements of the landing system and trusting the software to do its job. Because the round-trip delay for a human to be able to effect some part of the landing is nearly half an hour, and the entirety of the landing would take about a quarter that much time. Indeed, by the time we on Earth would receive a signal indicating that the atmospheric entry had begun in reality the entire landing would have already occurred several minutes in the past.<p>If you ever thought that a lot was riding on your code, this will certainly put that in perspective.
It's hard to get a sense of scale, one minute it's wow so high up look way down at the craters then suddenly oh! That's not a crater that's the ground.<p>Beautiful I didn't know they had such high quality video I figured like it seems on all the others missions data is kept to a minimum and we'd only see black and white 640x480.
The final drop is quite messy, there´s a lot of dust.<p>I wonder how happy NASA is with the sky-crane solution - would they do it again as the new standard for mars rovers or is it in the "nice try" category?