I think space is one of those places where you can appeal to unknown unknowns being a major risk factor and everyone agrees. But I personally think that the better option is to just do all of this on the moon. Making a human-habitable base on the moon, to me, still makes the most sense for preparing for habitation on other worlds.

Avoiding the need to wait 26 months between launch windows certainly reduces the degree of difficulty. Going from "oh we wish we had X that we didn't think of" to having it is literally almost 3 years on Mars.

It's not to avoid the difficulty, it's to make iteration safer and easier. Good luck seeing what failed if the craft that failed is flying somewhere between earth and mars

For one, leaving the safety of Earth's magnetosphere. We haven't really ventured far enough for radiation to not be tempered by Earth's magnetic field, from what I understand. It would shorten the duration of radiation and help study the effects on astronauts.

So, go to moon. No real need to go to another planet just to irradiate some astronauts and in case shit hits the fan they have chance of surviving

Precisely the goal of Gateway, of which Artemis/SLS are pieces.

Interplanetary space is significantly different environment, compared to low earth orbit, for spacecraft and crews. The only knowledge we have of humans beyond the Van Allen Belts is from Apollo, for relatively short durations.

> Engines running for long time ?

Not with chemical rockets. Perhaps you meant "engines sitting idle for years in interplanetary space, and then restarting reliably".

Really depends on the mission profile. If you are refilling in space for a "short" burn, it could still be longest in human history.