Re: Too hard, too frequent, too unreliable
Yeah, the SRE-educated burst into laughter at that point. They might have been SOLD five nines, but clearly only two were delivered.
2994 publicly visible posts • joined 23 Jan 2015
I'm seriously having a problem understanding that this is real. In order for this attack to succeed, we need someone who
1) Uses bitcoin.
2) Uses Tor.
3) Uses a bitcoin mixer service.
4) Does NOT use https to access the bitcoin mixer service via Tor.
Isn't that like just one guy? What am I missing?
Also, if you're already using Tor, why on earth would you suddenly care about speed over blind trust in the exit node for that final hop?
I've not been tracking it closely, but China-India relations have been complex at best for several decades. In particular, China's oft-stated desire to end "unipolar world" is shown to be a farce where India is concerned. They want to be the sole dominate power. Which is pretty well true of any power, really.
The idea of any power being truly independent is a farce if it depends on another power for anything necessary to defend itself in war. You may believe that such concerns are atavistic, but China's relationships with its neighbors belies that idea.
The recent escalation of hostilities on the China-India border is what seems most relevant to me, from a timing standpoint.
What does the IP version have to do with it? You have a map from services to ip address & port number. How do we gain by making our services to hunt for this information?
I'm really trying to understand this. DNS was created to allow coordination in the absence of organizational discipline. It's what allows the intranets to connect to make the Internet. Why is it the correct solution when everything is in-house?
You really want to compare the US and China on human rights. Okay... How many are at Gitmo? How many are in Xinjang? How does one end up in Gitmo? How does one avoid the abuses in Xinjang? Does the US assign your wife a husband if you go to Gitmo? Do we perform live donor organ transplants?
I do not consider the US record on human rights to be that great--according to our ever-escalating standards. But the concept of individual rights does not even exist in China.
If someone says that you are their enemy, believe them.
The Chinese have been pushing an aggressive anti-American rhetoric for decades. I was alarmed by what they were saying in the late 80s. Given that the Chinese lack the concept of human rights as we have it in the West, their party congresses have often stated that it is the duty of every citizen to aid in the "struggle". Every citizen.
But supply line vulnerability is near-total. The size of the blobs in firmware make it intractable to figure out everything that is happening at a low level
None of the above suggests that the US government is not aggressively engaging in its own espionage. It has been since president Wilson left office, at least. As it must. Which is completely irrelevant to the decisions it should make regarding the behaviors of the Chinese government.
Yeah--because English-Irish relations have been SO wonderful since forever. And English-Scott. And English-Frog<bs><bs>ench.
Give it a rest. People get along better with people who are like them. The idea that we should make the effort to get along with people NOT like us is a fine one. You should try it.
The NSA has two missions. They were kinda skipping on the second one for a couple of decades, but the congressional wrap on the knuckles has adjusted things a bit.
Their second mission is to protect "our" communications and data. With the "our" being Americans', not just the government's. This is why, for instance, that the NSA is on record as opposing back door encryption.
I'm pretty sure that's the point of the original comment.
NVIDIA wants to sell product into the Linux environment. Great! We'ld love to have you join in! What? You don't want to join in?
Anyone selling to the Linux market is making use of the labors of the kernel devs, the distribution maintainers, and everyone who is going to the effort of maintaining a Linux box. There is a price to be paid for the use of these efforts. Pay it, or stay away.
Seriously, would it not have been simpler and faster to contact them and say, "We've got a thundering heard. We need you to block 99% of the traffic with this user agent.
And let the fix change the user agent stream. When the herd thins enough, you can drop the block percentage.
Of course now, with SRE, the rollout is paused by the system at the 15% level because of the increased network traffic driving up system load.
I would be shocked if there is ANY evidence cite which rules out Oort cloud activity in favor of an asteroid. "At the same time" in this context almost certainly means "+/= 10-20MY". Whether it was a single event, or an orbital alignment that caused numerous events over this period has got to be entirely speculative.
If you don't get the reference, pull your head out & read a bit about the history of Western Civilization and the fitful and uncertain development of government that is, in some meaningful fashion, tied to consent of the governed.
What is being called for in this article is the full-on abrogation of the right to compete in the market of ideas by those _currently_ unacceptable to the author.
Don't assume you won't be next.
I think you may have misunderstood. It appears to me that he is saying that a dual stack network is more of a pain to support than a v6 with 4-to-6 translation at the edge(s). Therefore, _once_ a network decides to go v6, he expects it to go v6 instead of dual stack.
I could be wildly wrong, but I expect most SMBs to stay v4 for a long time. I can see no reason for hw to drop v4--so what is the need to change over? The sw tools are all in place to support v4 with 6to4 at the edge, and they too are not going to go away.
When setup of fresh v6 networks is as simple as v4 networks, expect new networks to start being v6. Especially as the new kids come in without v4 experience.
The event horizon of a black hole is the surface described by points at which the escape velocity equals the speed of light.
Lots and LOTS of people have trouble understanding what this implies. A photon created at the event horizon and aimed directly away from the center would (in a perfectly symmetric world & absent other interactions) never return. But if it were angled 179 degrees from the center, it would behave like a cannonball.
Whats more, a photon formed a short distance inside the event horizon, and angled 179.9 degrees away from center, might only deflect to being 179 degrees from center as it passes the event horizon.
In the frame of reference of an object falling into a black hole, there current escape velocity is a mere curiosity. What matters is things like the difference in the gravitational pull experienced by one part of the object relative to another. As mentioned, for supermassive black holes, this is negligible at the event horizon.
Sentiments like this justifying sloppy behavior on the part of the programmer are a major source of tire fire our industry has become.
I can hardly write five lines of code without some sort of bug. But guess what? I write 0 lines of production code without tests. What's more, I have the training to know when my tests are inadequate.
When I was at Google, the solution to memory leaks was often to grow the alloc. We had some small service that was running on 80G allocs (biggest available). I threw the flag, found the leak, fixed it. That service then ran on 10G or less.
Learn to find your own mistakes. Learn that the people who point out your mistakes are your best friend. Teach these lessons to others. And stop whining that your tools should stop you from playing the fool.
1) It is NOT the same code. There are lots of ways to guarantee that a pointer is not null. Many of them don't involve a conditional branch.
2) Calling code "faster" because it is executed sooner is just weird. You get a (very) slight reduction in code size. Execution speed improvement is limited to improved cache utilization.
The convention against the use of chemical weapons works because it is relatively easy to trace the physical components. Nuke treaties can (in theory) work because ICBMs are large enough to be traceable.
Bits? Not so much.
The real justification for this stuff is that if we don't continually improve our capabilities, the Chinese will get so far ahead of us that we will be wiped out in a minute. Okay if you want to live in Hong Kong these days, I guess. But for me? No.
I REALLY wish I had a good answer for this.
You're making a HUGE assumption here. When the mob gets big enough, there is no private property. Even in the US.
Did you know that "personal need" is a defense against the crime of theft? Apply that to a vaccine during a pandemic, and you've got really strong ground. Of course, that's stupid, because you've just defunded the research for the next bat virus Pooh Bear's minions decide to grace us with.
But we're talking politics here.
o'tta get better....
So Safe Harbor was replaced by Privacy Shield.
Assuming that the three letters now in play suffice, I expect the replacement for Privacy Shield to be entitled "Half Privacy".
This is a charade played out by the EU government to placate its citizens while permitting the merry game of monetization of privacy to proceed full speed ahead. We're going to see a "really, truly better, I really, really mean it" fig leaf about three days before the deadline. Which will take two-three years to be ruled invalid.
I'm about as from from an LD as you can get, but unless they've been making some huge to-do about how competent they are with managing IT contracts, that really, REALLY doesn't matter. Especially since almost certainly everything but final oversight was handled by careerists.
I've had far, far too many interviews over the last 24 years. The majority have involved whiteboarding code.
If folks aren't great at thinking on their feet, then asking them to solve ANY problem on the spot is going to be an issue.
If the interviews are being nasty, then, sure--that's stress. But by far, for me, that's come from non-whiteboard situations.
If you're attacking people for syntax errors in whiteboarding, you're an ***.
If you're jumping on ANY mistake as soon as it hits the board, you're an ***.
These days, CoderPad and it's ilk have become more favored. That's acceptable 1-on-1, but you need some process to have it work for more.
And if you can BS your way through a technical interview, that's on the interviewer. I've had candidates try that. I've been polite.