@lupi

History doesn't repeat, but it sure does love to rhyme. Right now, we're watching the tech industry eat itself all over again, as the boom times screech to a halt. So if the last few years felt like the dotcom bubble, you might not be surprised it's where this story starts.

In the mid-to-late 1990s, as the dotcom boom was in full swing, the US Department of Defense was seeking to replace the aging Titan IV, Atlas II, Delta II, and so forth, vehicles that had been pressed into continued service after the shuttle failed to meet the DoD's expectations for operational readiness and capability. They put out a contract to all the major rocket builders for what they called Evolved Expendable Launch Vehicles, and when you're a defense contractor, when the government says jump, you say "how high/how much are you paying us." Plenty of contractors got to work, but only two contenders were selected.

*  Lockheed Martin provided their proposed Atlas V (and maybe the Atlas III? can't recall), which benefitted from a collaboration with Russian aerospace industry (arranged by the US State Department!), providing the RD-180 engine to power them. They're not important here, as much.
*  McDonnell Douglas (later Boeing, after the merger) provided their proposed Delta IV. It shared a name with Delta rockets past, but little else. It would take the cryogenic upper stage from the failed, interim Delta III (my beloved) and mate it to a fully cryogenic core stage. And, importantly, it's the focus of our history lesson here.

As you can imagine, with the tech industry on what seemed like an endless climb towards the heavens, endless investment money pouring in, and so many tech companies promising revolutionary new services for and facilitated by the burgeoning web, a lot of folks were really getting into satellites. Motorola famously took a big gamble on Iridium, one that Really Didn't Pay Off and they nearly scuttled the whole thing (there's a good book about it i've never read, Eccentric Orbits). But to Boeing and Lockheed, the EELV contract wasn't just an opportunity to position themselves for government money, it would afford them the chance to build new designs to suit this upcoming commercial satellite demand.

Boeing would build an entirely new facility in Decatur, Alabama to produce the Delta IV rocket, a titan of rocket manufacturing, capable of producing 40 rockets a year, a flight rate that even Shuttle hadn't been able to realize. They saw number going up, and they wanted to ride that train, they designed Delta IV to be a commercial powerhouse, intended for a high flight rate to keep that factory constantly ticking along.

And then that commercial market completely failed to materialize. The bottom fell out under the dotcom bubble, it burst, and suddenly Delta's greatest strength (its capacity for a high flight rate) was its greatest weakness. With all the fixed costs of their overbuilt infrastructure, the business case didn't close.
Delta IV would forever be stunted by its own potential, horrendously unprofitable but mandated to exist by the DoD for "assured access to space," a policy instituted after Challenger to make sure that there were multiple avenues for them to get things into space.
This would have ramifications of its own (ask me about Vulcan, and ULA's Decatur Problem sometime), but they don't matter here insofar as they're a demonstration of a big mistake, Boeing's hubris, confidence, eyes larger than stomach, stomach larger than portion size, et cetera.


How does Boeing's failure define SpaceX's success?

In the late 2000s, after SpaceX got Falcon 1 to succeed by the skin of their teeth, NASA awarded them and some other companies (Rocketplane Kistler, Sierra Nevada, and later Orbital Sciences) contracts under the Commercial Orbital Transportation Systems contract, partly funding the development of vehicles and systems that would be able to bring supplies to and from the Space Station as the Shuttle retired. For SpaceX, this would be their impetus to develop a new vehicle, as EELV had been for Boeing and Lockheed.

That vehicle would become the Falcon 9, and the Dragon spacecraft that rode it. And in this moment, I cannot express how much Falcon 9 is "What if we didn't fuck up Delta IV." They saw what had happened when Boeing overextended, and endeavoured to design a rocket to be as lean, cost-effective, and potent as possible for a specific market sector.

* Both stages of the falcon 9 use the same tooling to produce their propellant tanks.
* Both stages even use the same engine, with the second stage's engine just being vacuum optimized with nozzle extensions and some other minor tweaks.

Falcon 9 trades efficiency for convenience with both of these, as an upper stage fueled by RP-1 (highly refined kerosene) loses a whole lot of efficiency over the hydrolox upper stages used by the EELVs, but it greatly simplifies a whole lot for them to have common propellants.

*  Additionally, they sought manufacturing efficiency by only developing one size of payload fairing, sized to fit the majority of commercial satellites, a move that prevented them from accepting a great many defense contracts.
*  Lastly, SpaceX's Hawthorne facility was modest, they didn't overbuild, they didn't put the cart before the horse, they made sure it was lean. Hell, even when it came to their launchpads, they were lean, claiming equipment and structures from scrapyard and govdeals to build things up. They even got the old Beal Aerospace proving ground for a steal, it's now their McGregor testing facility. (Beal aerospace is another hilarious story from 90s spaceflight, my favorite era)

All of this meant that even before they started strapping legs to it and trying to hit barges in the ocean, they had a potent, cost-effective vehicle primed to meet the ISS resupply contract, eat Roscosmos and Arianespace's lunch by undercutting them in the commercial market, and even bid on the handful of defense contracts it was qualified for to get little a DoD money as a treat. Falcon 9 succeeds because it was designed by a competent team with a clear vision and a demonstrable awareness of past failure to avoid.

They've managed to stretch the design from its stocky, square-bottomed beginnings to a genuinely remarkable powerhouse of a lifter, and a reusable one at that. It's only blown up a couple times, once from an issue with a supplier (which isn't even something they alone have dealt with, rip Taurus II), and once while it was still pushing the envelope, getting stretched and experimenting with performance increases.

And that's even before mentioning the teams that made it happen. They had so many industry experts, they had people from TRW who worked on an engine for the Space Launch Initiative program, its engine technology finding its way into Merlin, for example. They were no slouches, SpaceX attracted all sorts of talent with decades upon decades of industry experience in the early days, and it really shows in what they were able to accomplish back then.


If they had such smart people at the wheel, how the hell did they fuck up this hard?

Simple. When Elon Musk built his team out in the wetlands of south texas, at a site that had originally only been slated to be a ground tracking station, then a potential falcon 9 launchsite at most, he brought almost nobody with him that worked on Falcon 9, or really, rockets or launchpads in general. This decision baffles me, but then, I'm not the kind of "more money than sense" sort that Mr. Twitter is.

Starship as we know it has a long history of bold promises, from the original unveil of the Interplanetary Transport System/Big "Falcon" Rocket (no, it didn't really stand for falcon, of course it didn't, this is elon musk we're talking about) at the 2016 International Astronautical Conference in Guadalajara, through to the spectacular conflagration we saw today.

Frankly, a lot of that doesn't matter, because up until they broke ground in boca chica, the most real the thing had ever been was that time they purchased a shitload of expensive carbon fiber tooling and entered into a lease with the port of long beach to use an old WWII shipyard for manufacturing, and then backed out of the lease and scrapped the tooling. I think they made one or two test tanks out of carbon fiber and blew them up (on purpose)? Doesn't matter.
The vehicle design changed at least annually, with Musk's initial promises and claims that "it won't be a space shuttle, it won't have wings, or tiles," left by the wayside, as they tried designs that were flat bottomed lifting bodies, designs with fins, and ultimately designs with tiles like we see now.

It was only when they started fabricating the thing (after contracting a water tower construction company to manufacture a singular prototype vehicle, called Starhopper by fans) that a real design for it began to emerge, out of an impetus that can only be seen when you actually start to have skin in the game and have to sell a product, not just an idea of one. And now, that design process was taking place with a team of people who had very little experience with launch vehicles, or launch pads, leading to decisions like:

*  "let's catch the landing vehicle out of the air with giant arms on the launch tower for immediate reuse"
*  "we don't need a flame trench, we're not gonna have one on mars" (though that decision may be equally rooted in not wanting to deal with the paperwork required to do any amount of earthwork in the protected wetlands around their launchsite)
 * "we're gonna make the thing use these weird flappy things driven by tesla motors to control itself on reentry, so it can reenter on its belly and land on its tail"
 * "what if the vehicle WAS the fairing and it opened like an alligator"

Baffling technical decisions aside, we've got more to unpack here. Like the question of, well, what is starship... for?

* launching huge batches of starlink satellites
* launching every satellite we can fit in it super cheap because reusability means cheap right
* point to point suborbital passenger flight
* point to point suborbital logistics for the department of defense
* orbital tourism
* lunar lander for the Artemis program
* mars colonial transporter
* honestly, the list goes on.

and so we begin to see a problem. Does anyone else remember a space vehicle design with lofty promises, including the concept of being a Vehicle To Do Anything And Everything, do it cheaply, reusably, and innovatively? A vehicle that would make spaceflight routine, a vehicle that would make it accessible to everyone, a vehicle that would

hi i'm talking about the space shuttle can you tell i'm talking about the space shuttle? I love the space shuttle a whole lot, but love isn't blind, and even the best spacecraft ever designed isn't without its faults.
Even when its worst moments were the result of human factors choosing to operate it outside of safe limits, bad safety culture allowing egregious risks to be tolerated in the name of meeting deadlines, et cetera, even with that defense, it was still a shadow of what it was promised to be, suffocated under the weight of having to be designed to do everything, and not designed to be good at any one thing.

The team working at Boca Chica are either unaware of the mistakes they're making (and repeating) through lack of experience, or in Musk's case actively ignoring the fact that he's driving those mistakes, because SpaceX's past success has made him feel like they're invincible, and "rip to the Shuttle but I'm Different. Better Actually."


What can we learn from this? What conclusions can we draw?

I dunno really. SpaceX now is where Boeing was in 1998, thinking they're invincible and making huge mistakes in the name of ambition, maybe. The market conditions sure line up for that, as the space industry is undergoing another mini-collapse.

We're seeing a bloodbath as small launch companies fail, leading venture capital investors to back out as spaceflight seems like a riskier investment, on top of VC being super gunshy in general as crypto and everything falters, jumping all over the place trying to find the next big pump and dump, like AI or generative art or NFTs or the metaverse, just trying to find a lifeline to keep their precarious little empires of dirt from getting washed away.

Spaceflight is risky business, the only dependable source of revenue comes from governments, due to the absolute fickle state of the commercial market. The actual sturdy customers, telecoms, only need a couple sats every ten years, they're not enough to anchor you.
SpaceX is attempting to induce their own demand by vomiting thousands of starlink satellites into the sky to try and be the first to bring a profitable (rip Iridium) satellite internet constellation to market.
Rocketlab is branching out by offering end to end services designing, building, and flying payloads for academia and small industry on their Electron+Photon platforms, while also attempting to scale up to a medium lift vehicle with Neutron, and also as of this week, selling Electron as a hypersonic research testbed.
We're seeing bloodshed as even companies we thought were sure shots like Branson-backed Virgin Orbit fold as funding dries up. Those who have survived are trying to make every pivot they can to weather the oncoming storm.