4 ms·

The problem with this argument, when talking about space, is that precedent is often so extremely conservative and over-engineered that there's likely to be a h

The problem with this argument, when talking about space, is that precedent is often so extremely conservative and over-engineered that there's likely to be a huge middle ground between "good enough" and what NASA did in the 60s. NASA built a launch pad with a flame diverter and water quenching system - and it worked! But now they have it, they're unlikely to test launches without it. Space X clearly think that there's a middle ground - something which is simpler, quicker and cheaper than the status quo, but which is good enough. Clearly they failed - and arguably they were irresponsible not to do more tests. Given the environmental repercussions of failing, they probably should have started with something closer to precedent and worked iteratively to simplify it. But when you're talking about this sort of very low volume system (how many launch pads, total, are there in the world?), precedent is only worth so much. (It's also the reason so little progress was made in re-usable launchers until SpaceX came along). Hardware is not software - I completely agree. But you can only make progress by testing hardware. Assumptions are rarely correct.

davedx 3y ago
Yup, apparently the pad at 39A was overdesigned too (for larger rockets than were launched from it). Engineering is all about finding the tolerances. It's a difficult line to walk.
mft_ 3y ago
Broadly speaking, any individual or organisation working in a regulated industry is directly incentivised to either maintain the current status quo of conservatism, or increase it further. Maintaining existing conservative approaches brings a huge amount of 'CYA' with little direct downside. Trying to even slightly roll back from a conservative position places all of the risk on the individual or individuals responsible, with a huge downside versus a small incremental upside that few will likely appreciate. It's mostly a thankless endeavour I see this regularly in a different industry (with FDA, EMA) and it's what SpaceX are dealing with here. While I'm not arguing that their decision to launch with the existing stage 0 was correct, their decision to do something differently has introduced that much larger downside - further delays and potential censure from the FAA, and opprobrium from commentators and internet forums alike.

davedx 3y ago

Yup, apparently the pad at 39A was overdesigned too (for larger rockets than were launched from it). Engineering is all about finding the tolerances. It's a difficult line to walk.

mft_ 3y ago

Broadly speaking, any individual or organisation working in a regulated industry is directly incentivised to either maintain the current status quo of conservatism, or increase it further. Maintaining existing conservative approaches brings a huge amount of 'CYA' with little direct downside. Trying to even slightly roll back from a conservative position places all of the risk on the individual or individuals responsible, with a huge downside versus a small incremental upside that few will likely appreciate. It's mostly a thankless endeavour I see this regularly in a different industry (with FDA, EMA) and it's what SpaceX are dealing with here. While I'm not arguing that their decision to launch with the existing stage 0 was correct, their decision to do something differently has introduced that much larger downside - further delays and potential censure from the FAA, and opprobrium from commentators and internet forums alike.