By this argument, Leonardo would have learned a lot more about flight from trying to build a rocket to the Moon than from trying to build a glider or studying bird flight.

That is, attempting something that is impossible given your current level of engineering is not likely to produce any useful results.

> Leonardo would have learned a lot more about flight from trying to build a rocket to the Moon than from trying to build a glider or studying bird flight.

Ridiculous, given that flight is entirely about swimming through the air, while space rocketry is mostly about how to get through and out of the atmosphere ASAP.

(Not to mention, rocketry as a field was already making better progress than flight back then.)

> That is, attempting something that is impossible given your current level of engineering is not likely to produce any useful results.

This is how we've always been learning, though. Even Leonardo and other early pioneers of flight constantly attempted things impossible at their level. It's those attempts that led to progress - the theory and methods to solve such problems on paper came about only in the century.

Incidentally, it also turned out that studying bird flight was a waste of time - bird flight is too complicated for us to replicate even these days, and we mostly don't bother, because simpler systems yield better results for the kind of needs we have now.

We already know all the basics about agriculture for example. Hardly any of the problems we need to solve to do agriculture in Antarctica are the same as the problems we need to solve to do it in Mars. The soil medium is different, the processing needed to make it useful for agriculture is different, the light levels are different, the atmospheric environment is different.

Ultimately Antarctica is so similar to, say, a mountaintop in the USA, or even a lab in a US city compared to Mars that I dont really see the added value. The unknown problems we will face on Mars are there on Mars, not here.

Will a city on Mars ever be viable, let alone self sufficient? I have my doubts, but there’s only one way to find out.

Now try growing grain indoor with lights instead of just lettuce with no caloric value for multiple years with no outside input. And also balancing crop carbon needs against human breathing needs in a closed capsule without having something like mold or algae or other bacteria from throwing everything off balance because you lack 99.99% of the ecological diversity of anywhere on Earth. Or having access to fossil-fueled derived fertilizers which takes massive amounts of energy to produce without natural gas. It sure as hell doesn't make sense to ship fertilizer from the surface of the Earth to Mars, or water, or really anything other than handfuls of extremely specialized equipment like computer chips or extremely difficult to obtain or rare chemicals or elements. You are going to need sustainability of all your basic resources for years, if not decades, before you ever have a chance of assaying and mining and processing such materials for yourself.

How much enriched nuclear material can we really send up in a rocket at a time in order to fuel the massive energy requirements of such a colony? We sure as hell don't want to send up multiple tons of enriched material at a time in a rocket to potentially fail during launch, and we would need that much energy to build any sort of functioning industry on Mars so it is anything more than a glorified emergency bunker where we send people to die.

Knowing the basics is still far removed from demonstrating self-sufficient capability at civilizational time scales. How many successful, isolated biodome projects are there? My understanding is that every one so far has failed. Theory ain't practice.

Of course, my point is that Antarctica is insufficiently different from e.g. Boston to make any difference to any research we might perform on earth. Meanwhile there will be conditions on Mars we can’t anticipate or fully simulate anywhere on earth. We need to do both.

> That is, attempting something that is impossible given your current level of engineering is not likely to produce any useful results.

Engineering advances fundamentally by trying to do things you haven't done before. If you already know how to do it, you higher a technician and hand him the instructions. (Many software developers for example are closer to trades people/technicians than actual engineers.)