I take pride in making my websites accessible for everyone regardless of the monetary incentive to do so, and to make inaccessible websites simply because you don't think there's money in it is very shameful.
Not saying that hamburger menus can or cannot be accessible; just saying that, as an engineer, you really have an obligation to make something that is accessible. This is a basic, non-negotiable requirement for all software.
The cheapest and most popular rocket today in terms of cost per kilogram sent to low-earth-orbit is the SpaceX Falcon 9, which is estimated to cost about $2700/kg. If the Starship meets its goals, it will cost $10/kg to LEO. It has a payload capacity and infrastructure which can economically deliver massive payloads to any solid surface in the solar system. It could send three full-sized bulldozers to Mars in a single trip. This is the largest scale space project ever conceived and portents a major shake-up to the space industry and a big step towards establishing a permanent human presence off Earth.
>The cheapest and most popular rocket today in terms of cost per kilogram sent to low-earth-orbit is the SpaceX Falcon 9, which is estimated to cost about $2700/kg. If the Starship meets its goals, it will cost $10/kg to LEO.
I've read that the cost will be $2700/kg even if Starship is completely expended; that is, no reuse. But far, far, far larger payloads, of course.
It is. The Falcon 9, of which there have been hundreds of launches already, costs $60 Million per launch. And somehow Musk thinks he can make a rocket that is 10 times bigger 30 times cheaper? Wishful thinking. Best case imho is that they take the launch cost down to about 200M-ish over the next 15 years. That's a cost of $1000/kg and sounds about right to me.
How much of a setback would disaster be? Would SpaceX keep going, or would failure leave room for competitors to take over? (If there are competitors? I don't even know tbqh.)
SpaceX has followed a tech startup-like model of iterative development AKA "move fast, break things", in which many rockets go boom and (as saberdancer said) every boom is studied to reduce the odds of the next one go boom. Basically an edit-compile-debug loop writ large. This is why SpaceX's official countdown schedule states "excitement guaranteed" at T-0:00.
>Would SpaceX keep going, or would failure leave room for competitors to take over? (If there are competitors? I don't even know tbqh.)
As saberdancer said, there is no serious competitor. Rocket Lab has achieved ~~reusability~~ recovery of small rockets and has a good launch record, but a) has not yet figured out how to get them to land other than on the ocean and b) can't launch them weekly/daily like SpaceX. Bezos's Blue Origin actually landed rockets before SpaceX, but only suborbitally. ULA (US) and Arianespace (Europe), the two large commercial-payload providers pre-SpaceX, aren't at the point of test launches of any reusable rockets.
The above is all before Starship, which if it works will give SpaceX a) a fully reusable launch vehicle (existing SpaceX rockets' upper stages aren't reusable), and b) the ability to launch gigantic payloads. The combination of the two will further massively cut launch cost. I've read that even if Starship is fully expended, it will still be the same $2700/kg that existing SpaceX rockets cost, with that aforementioned gigantic payload capacity.
It would range from unfortunate to a non-issue. The entire flight is not expected to be successful, the purpose is data gathering. The worst case scenario is that it fails on the pad and damages the launch infrastructure, which would be a huge setback to the project. If it fails anywhere else it's probably fine.
I haven't followed this specific launch in much detail, but part of why SpaceX has been so successful is they expect a certain amount of failure and don't pin the failure of the program on the failure of any one test. This allows them to iterate faster than other programs that might have less test launches and thus rely more on the success of the launches they do perform. I've heard for example that this rocket is already out-of-date and the next planned rockets already have a bunch of improvements.
Not a big setback at all. They would have to analyze reasons for failure and make changes. There are no serious "reusability" competitors, especially at Starship size.
SpaceX is trying to iterate quickly on this rocket so they have several already assembled. Next two launches will also be fully expendable.
Disaster is complete demolition of the launch site, setting their schedule back many months and costing them a rumored hundreds of millions to rebuild. They've run enough tests to support reasonable confidence that won't happen. The directionality towards mass manufacturing the rockets, instrumenting to observability scale data collection to feed engineering course corrections, and resilience to setbacks is promising.
The hundreds of millions figure doesn’t sound implausible either way necessarily, but I wonder if it includes the R&D costs? I.e. building number 2 would be incrementally much less expensive than building number 1.
I wouldn't be surprised R&D numbers were included in the rumored figures, but considering their lessons learned today from the extensive concrete spalling incurred by the launch, there is even more R&D to perform with the next build-out. Especially when we take into account some observers are claiming SpaceX is installing a deluge system [1] for the next build.
I'd like to hear from some rocket nerds why launch pads don't sit on top of a giant, deep water body, as the deluge systems as I vaguely understand them seem primarily to manage the shock and secondarily the heat, and concrete is a difficult material to manage over a long period of frequent use as a shock absorber.
I am a kernel hacker, and I have worked with io_uring, and I can safely judge that it is very good -- but the main issue is that it represents a totally different approach to I/O (and syscalls generally, which are just I/O in other words), which is going to take the ecosystem a while to reform around. Note that the sample code in the OP's article is much more complex than the traditional approach. It's also very Linux-specific, so any software which takes advantage of it will be less portable or will have to write multiple I/O backends. It's also nontrivial to understand and use effectively, so adding good io_uring support to a project is an effort.
My hope is that Linux ports of kqueue or libdispatch can use io_uring where possible and we would automatically get the benefit, but I'm not familiar enough with the guts of these projects to know how timely or even feasible that is.
Fair supposition, but lighthouse controls for latency and throughput and the same tests have been run from various parts of the world by various parties with indistinguishable results. Feel free to run them on your own network, it's pretty straightforward:
Many of the other comments are missing the point. It's ultimately because there are a lot of desperate people in SF. The bay area has the highest homelessness incidence in the country. Mental health services are poor to non-existent. The city is full of people with no options, no healthcare, and no support system. We could criminalize homelessness even more, and brutalize the desperate with more police interventions, but that's a lazy and unsympathetic explanation for why this is happening. Society has failed the poor and SF is exhibit A.
This. In europe we treat sick people, in the US they leave them on the streets, and then task the police with beating them regularly.
I live in europe, in a nice house. We don't have any guns in our house, and we don't fear random weirdos trying to break into our house, because our society takes care of and treats sick people. Because we are not scaring and harrassing them all over the place, they aren't running rampant around areas of our city.
I don't care about having health insurance, because I already have it as a citizen, paid for with my taxes. If I'm fired from my job, I still have health insurance.
Surely resources for people living in poverty are no longer necessary for someone once those resources have the desired effect and they stop living in poverty?
Looks to me that the logical alternatives to means-tested aid are either no resources for anybody, which seems undesirable, or resources for everybody, no matter how well-off they are already, which seems wasteful. What do you suggest instead?
Are all those homeless originally from SF, or do they migrate there?
If it's the latter then I assume that social services are just overwhelmed?
I'm pretty sure that if I was made homeless tomorrow, and knew I had no prospect of pulling myself out of that situation, I'd migrate to somewhere with a climate like that of SF.
The overwhelmingly vast majority migrate. Some of them might have had a sublet apartment or lived at a friend's couch for a while before becoming homeless.
So the statistics say "most homeless lived in SF prior to becoming homeless", but if you actually check local school records you know they didn't grow up here.
In addition to the year-round mild climate, SF's local laws (and enforcement or lack thereof, e.g. legalizing theft of less than $1000 or intentionally spreading HIV) attracts desperate people, as well as people who simply enjoy the lifestyle, from all over the US, by rewarding them for doing what they do.
Basic "supply and demand" also applies to crime, playground needles and sidewalk turds: if you lower the "price" (punishment) at constant demand (people who want to steal or shoot up in public), the supply increases.
You’re putting a lot of stock into SF’s failures, but keep in mind that other cities ship their homeless to SF, and as a result, SF gets a disproportionate share of homeless.
Not saying that hamburger menus can or cannot be accessible; just saying that, as an engineer, you really have an obligation to make something that is accessible. This is a basic, non-negotiable requirement for all software.