Basically congress has set a requirement to build a 2025 operating system using Perl with 100% uptime and spread development across 20 companies run by idiots. NASA doesn’t have any extra budget to spend with that horrible mess.
Starship has been much cheaper to develop and launch because of SpaceX's significantly less conservative approach to R&D, but it doesn't work yet. We should maintain SLS until SpaceX either gets Starship fully developed or it falls apart.
Starship is risky and if we want Space to be a "critical" capability of America, then we shouldn't be putting all our eggs in the SpaceX basket, even if it is significantly more expensive.
How cheap do you think buying Starship services will be when they are LITERALLY the only option for anyone anywhere?
What!?
Can you expand on this? That's horrific. I hope that's just an analogy and not the truth.
This book has many illustrations of alternate designs considered before the Space Shuttle was finalized
https://www.amazon.com/Shuttle-History-Developing-National-T...
as well as the results of late 1980s-early 1990s studies that considered the possibility of putting together Space Shuttle parts in different ways. Say you could have a bigger external tank, four solid rocket boosters instead of two, and a big-ass orbiter which has five engines instead of three. (That one probably is too much of a boondoggle because they'd have to spend a few billion dollars making a bigger launchpad!)
You might think that, with the research done to develop the parts, it would be cost effective to build something new from Space Shuttle parts but in those studies it was always outrageously expensive to build anything out of Space Shuttle parts because... Space Shuttle parts are stupendously expensive.
In an alternate universe something like the SLS (a non-reusable big-ass booster with lots of stages) could be a lot cheaper than the SLS if it was cost-optimized, but cost-optimization itself isn't cheap!
By contrast to that, the Starship model isn't proven either. Reusable Starship to LEO could be a good business, but orbital refueling is still unsolved and even if it is perfected, current plans are to launch 12-20 Starships to land and return maybe 3 tons from the moon. Think of what a huge thing you could build in LEO if you had 20 Starships!
One way to refuel in space is to carry the cargo fuel in a container that can be detached and reattached to another craft.
The craven avoidance of in-space refueling begs the question of what kind of future is NASA imagining for humanity in space, if something as simple as transferring a fluid from one tank to another is too difficult.
Growing up in the 1980s reading the "science fact" columns in Analog magazine I got the idea that we got sold out when we decided to do Apollo the way we did and that instead of a stunt we could have had a much more capable program that would have led to a more durable presence on the moon.
Today though looking at large numbers of plans to get to the moon it's easy to come to the same conclusion von Braun did (from looking at large number of plans), that the Apollo architecture was much more feasible than any of the others. It wasn't just "get it done on Kennedy's timeline" but a matter of being able to do it at all.
It's not so hard to transfer fluids with gravity, still unprecedented to do it without. I studied theoretical physics so I didn't get too into it but I ran into a few situations in labs handling cryogenic fluids where a liquid wasn't a liquid when it got to the other end of a pipe!
---
From a mission planning viewpoint also I'm not too impressed with Starship scenarios that involve refueling. That is, I've heard it takes anything from 12-20 Starships to fully top off a Starship to go to the Moon or Mars.
You could imagine one launchpad could launch 1500 tons of cargo to Mars a year (1 launch a day staging fully fueled Starships in a holding orbit to waiting for Earth and Mars to align every 2.5 year or so) but if you can launch 20x the cargo to a destination (as opposed to fuel) you could build a baby O'Neill colony in LEO and have one hell of a space tourism destination. It's not so much a shortcoming of Starship as much as it is the reality that Heinlein was wrong and LEO is much closer to the Earth than other destinations.
(For instance in the 1980s the US fielded an antisatellite weapon that could be launched from an F-15 and kill LEO satellites with just a 2.5 km/s delta-V. Try to shoot down a GPS satellite, however, and that's an entirely different problems)