You hit the key point, speed, but IRBMs also fire on a depressed trajectory and not a ballistic one. They can't travel as far but they get to the target much faster. So yes, you can see them but your response time is in the 10-15 minute range and there's not a heck of alot you can do in that time frame except hope you can intercept them.

But yes, the "hypersonic" bit is more PR than actual threat. Most missiles are "hypersonic", it's the short flight time from the trajectory that's the critical factor. Supposedly why there was a tacit understanding not to place SSBM's right off each other's coast during the cold war, it would destabilize things too much.

The same reason Russia launched an Oreshnik hypersonic missile at the Ukraine. To make a statement. "Here's what we could do..." The target of the missle strike was an old rocket booster facility and the actual damage was minimal (there's speculation that the MIRV's had dummy weights for balance in lieu of actual traditional explosive warheads because that was the easiest replacement for nuclear payloads). So, an arguably worthless target hit with the wrong (and expensive) weapon for the target.

Why? "If you push us far enough, the next one won't have dummy warheads."

Why the cables? "If we decide we not insane enough to use nukes, we could still damage your economies." Yes, there is a lot of interconnectivity but if enough of it is damaged it could cause serious problems for Europe and also make responding to and recovering from a more kinetic attack more difficult.

The case does not blame logitech for the accident. It cites the Logitech controller as an example of:

1. OceanGate using technology not intended or tested for this kind of use.
2. OceanGate's design having no redundancy or fail safe. It was a bluetooth controller with no wired backup and it was the only way you could control the sub.

There are many more details in the suit that layout the case that the owner flagrantly disregarded established engineering knowledge and practice as well as advice and feedback from those knowledgeable in the field because he was smarter than everyone else and "big submarine" was trying to silence him.

They frequently deprecate, remove and (possibly worse) change functionality from release to release. There are sometimes even changes from build to build.

Just review the "Removed Features & Options" in the JDK 21 Release Notes. https://www.oracle.com/java/te....

The release notes don't always capture everything. We had a production issue a few months ago because of a change that wasn't clearly (*) documented, had to dig into the JDK code to figure out what was going on. (*) The change was documented in the commit history but didn't make the release notes.

So... yeah... they do change.

2023 Successful launches:
Russia 19, China 66, US 109

2022 ...
Russia 22, China 64, US 84

2021...
Russia 24, China 53, US 48

2020...
Russia 17, China 35, US 40

2019...
Russia 25, China 32, US 27

The US had more successful launches last year than Russia has had in the past 5. China is also steadily doing better and better. Your right, it wasn't the sanctions. Roscosmos was a shadow of it's former self well before those.

I worked for a large fintech company doing cashless payments. This is a naive understanding of things. One of the biggest drivers for merchants was the speed of the transaction. You see a lot of talk about wanting to be able to track cash flows, spending habits, etc (*). That's not the biggest driver (for merchants), it's the speed of the transaction and not having to physically handle cash (for efficiency and security reasons). An ideal cashless payment takes less than 5 secs total. Cash takes 10+ sec, more like 20-30 sec or even more. For merchants that do a large volume (convenience, grocery, public transit, etc) having everyone use cash is untenable, they would have to double or triple the number of PoS (Point of Sale) machines.

Which brings up the second, and real issue, any chain store relies heavily on their PoS system which went down and is where the real issue lies. It does the pricing, the book keeping and the inventory (and usually forecasting). Nothing price tags anymore, so that 30 sec to pay with cash just became 1-2+ minutes while the clerk figures out the prices, tallies them up, figures out the tax (which in some places is variable depending on what you're buying). Manually booking flights for a major airline? You can just forget about it.

You think want a cash based society, but you really, really don't. It's going to cost you, personally, a significant amount of time and it also going to drive up the prices you pay. You want the system to be as redundant and fail safe as possible. Cashless apps/cards should be able to work independently for a time, PoS systems should be able to work even if the back office goes down, etc.

There is nothing special about cash. We have it because we reached a point where barter was inefficient. No one wants to be lugging 5 pounds of flour and vegetables around hoping the merchant will accept them. We've reach a point where cash is no longer an efficient way of supporting 8 billion (and growing) people.

(*) Merchants do want to track cash flows and spending habits, especially larger merchants... but they generally keep this data for themselves. Payments systems are generally not getting as much of this data as many people think. It was a struggle just to get data that was required for regulatory compliance (can't apply rebates/etc to alcohol & tobacco).

The distribution of data centers in Japan is really unbalanced. ~65% in Tokyo metro, ~25% in Osaka and then a handful of small ones scattered along the islands. It's a problem for DR planning because Tokyo and Osaka are close enough (~400km) it's conceivable (but unlikely) they could both be taken off line. Making matters worse, providers often don't have feature/service parity between Tokyo and Osaka. A number of AWS services aren't available in Osaka for example (but they have been improving that).

With the growing push for data residency, it's a problem. The Japanese government should really consider "You want a DC in Tokyo, you build an equivalent one in Sapporo first." No incentives, tax breaks, etc. Just build it. Right now there is just one. The population of northern japan is pretty low but it's cooler, the land is much cheaper and the Shinkansen runs all the way up to Hokkaido so right of way for a lot of fiber is not a problem.

You must be new to the internet and media in general. "The beating of a conservative" is a daily activity, sometimes with showings on the hour. Careers and lives are ruined every day, on both sides. In some cases it's deserved, in many cases it's excessive and on rare occasions it's completely wrong.

The real tragedy here is that they didnâ(TM)t get R.C. Bray to do Skippy from the Expeditionary Force series. Even more fitting since most ML answers are a âoesolid schmaybeâ. (That you know of.) For those who havenâ(TM)t heard the audiobooks, Skippy is an absent minded, snarky AI who is an asshole. RC Bray patterned him off of Fraiser and he is awesome.

You're conflating two different things. IP theft (which they do with abandon) and this current wave which is trying to block them from legitimately buying and building things. ASML lithography is banned. Nvidia is nerfed. This latest effort is going after license and royalty free RISC-V which was meant to be used by anyone. Why aren't they going after LoongsonISA (which is just a fork/clone of MIPS... which they did get a license for) or AMD's custom EPYC for mainland use?

Most people here know it won't accomplish anything, but the politicians will sell it to a lot of people who think it will. It's political theater.

This is awesome. They're obviously going to examine the situation and come to the conclusion that:

1. Tariffs and sanctions just encourage Chinese R&D and innovation and are, at best, a very temporary "fix"
2. We need build a quality, meritocracy based educational system and puts all students on an equal footing (financially, socially, etc). Many of our advancements were based on the back of the GI Bill which gave 8 million veterans access to education they may not have otherwise had).
3. Incentive long term corporate R&D
4. Fix investing so it is strongly weighted towards a companies long term prospects, not their quarterly results.

Just kidding. Both parties are going to continue to pretend that we're awesome, everyone else is backwards and if we keep our toys to ourselves no one else can possibly play.

Threads here and elsewhere have generally had the theme of "This is management's fault and management won't be held responsible!". Someone in management is finally paying the price (let's put aside whether it was enough for the moment) and people are saying "This won't fix anything!"

Pick one.

The truth is that this thing seems to be rotten from the head down and fixing it is going to take a _lot_ of changes. In management and on the floor. That takes time. You need do a full analysis and determine what's really going on. Then you need to figure out how to fix it without making things worse. If you go through and just rip out everyone and make sweeping changes you are going to introduce more risk, not lessen it.

This is one small step. If we start seeing regular news like this out of Boeing, we're changing this... we're introducing this... I'll start believing that they might actually be taking the problem seriously.

All of the information you listed is already public. They're not selling your data, they're selling a convenient way to access it. Some of it they're not even crawling your systems to pick up, because it's public. They're not compromising your security and I doubt they're putting any meaningful load on your system.

I work on a public, C2B system with about 50 million MAU. The security firms scanning our servers are a drop in the bucket compared to script kiddies and legitimate users trying to do ill advised things (automation, but badly). The worst offenders are the marketing/analytics bots. We found one that has a browser plugin that was feeding the users data back to their servers which were trying to scrap the pages near real-time. Enough users had it installed that it was triggering security alerts because the bot was trying to access pages that required authentication. After repeated attempts to ask them to knock if off we just blocked them. We even debated showing an alert to affected users but decided that would just as easily backfire with people accusing us of invading their privacy (How did you know I had this plugin!?)

Long story short, the internet is a wretched hive of scum and villainy. If you expose anything to the internet, you have to put up with this crap.

While I'm no fan of Microsoft, you can't win in a situation like this.

1. Don't upgrade dependencies and half the people will yell at you for not supporting new features, security, etc.
2. Upgrade dependencies and the other half will yell at you for breaking their sh*t.

18.04 LTS expired last year. You're running outdated software that you got for free and can upgrade for free. Stop b*tching at Microsoft. Yes, they have ESM (Extended Security Maintenance) but you have to pay for it and it's just security patches.

Your running on Linux. You have a two options here:

1. Upgrade Ubuntu.
2. Build glibc 2.28... it's not... that... hard.

Living next to any high power EM source may cause cancer.

Fixed that for you.

The only reputable study I'm aware of that has indicated that AM broadcast stations might cause cancer was a Korean in the early 2000's. But they only saw correlation within 2km of high power (100kw) stations and they weren't confident enough to establish causation. If it does cause cancer, you're going to see the same correlation with FM stations and TV stations. It's all radio/EM, the marketing names we use for it don't matter.