since blocking/throttling/etc is purely anti-competitive
Be careful with that line of thought.
"Anti-competitive" has very specific legal meaning, usually requiring that the behavior benefit the actor in some way. However, exclusive-access deals are not often seen as anti-competitive, since any other company could have made a better bid for access.
As an example, let's say that an ISP launches their own video-streaming service, and it's the only one that gets full bandwidth, while Netflix and Hulu and throttled. That's an easy case for anti-competitive behavior.
However, rather than launching their own service, they can open up bidding to be the "exclusive streaming video provider" for their network, and allow Netflix and Hulu (and everyone else, including the tiny little startup with no budget) to bid for that exclusive contract. In the end, the ISP still makes millions of dollars for throttling video, and the consumers still have very little option to move to other ISPs. Even discounting municipal monopolies (which are themselves just exclusive contracts), every other ISP is free to enact the same cash-grab policies.
Do note that in that context, it's easy to argue that since the startup and the major players are competing for the same contract, it's actually pro-competition! It's not discriminating against small companies; it's providing them an opportunity to win a business partnership for a growth opportunity! Of course, the little startup has no real chance to match the bid of the established companies, but to those who like this plan, that's just an indication that the small streaming service should grow more first, perhaps by competing for exclusivity with a smaller ISP for which the big players won't pay as much.
This scenario is a direct parallel to how a brick-and-mortar business grows, by getting local contracts near the company's physical home, that probably won't be noticed by the industry's major players, and probably isn't worth sending a sales rep out for. On the Internet, though, everything is global. The big players can afford to bid on every single opportunity, so a small stream service will likely never find a open niche for growth.
It seems to me that the best option would be to provide more options.
As a creator, let me pick who's covering the processing fees. If I'm aiming for wide appeal, it might be beneficial to pay my own fees, just to keep the apparent prices low. If I have a small appeal to some highly-contributing individuals, a few cents in processing fees won't change our relationship that much.
As a patron, let me pick when and how I'll be charged. Let me pick a schedule, and just show me the fees associated with it. If I'm picking something that causes costs to rise, I'm fine paying for it, but give me the option up front. I might be fine shifting my payment schedule a few days if it cuts down the fees, but let me make that choice. To compromise with the creator, a middle-ground option might be that the creator picks a plan for which he's willing to pay the minimized fees, and the cost of any deviations from that schedule would be paid by the patron choosing to be different.
As a business, Patreon can do a few other things to mitigate the apparently-high risks involved. Rather than passing chargebacks directly through to the creators, Patreon could hold a one-month buffer of donations, collecting the individual charges (and chargebacks), then releasing them to the creators 30-60 days after the charge deadline. Essentially, Patreon itself takes the role of a clearinghouse, maintaining account balances and payouts for each creator. Having floating balances would also open the door to allow creators to support other creators directly from their accounts (preferably with reduced fees), promoting collaboration under the Patreon banner.
Now, implementing these choices would be a significant development and logistics effort, but certainly possible.
You could also look at Troy Hunt's FAQ and blog, where he specifically states that there is no record of searches on the site (beyond server crash logs and non-scraping analytics), but that would require actually trusting a well-respected infosec expert.
I've used that on isolated systems before, in days long gone by. Now, many login systems recognize that the username and password partially match, and it rejects the pair.
I'm a bit dry at the moment... I know I had a can of Instant Froth around here somewhere...
Much better... Now then...
It's not a question of being "anti-science" or not, but more the question of "why?"
In the 1960s, landing on the moon was a huge accomplishment. We conducted important science, established permanent lunar installations of ongoing significance, and it paved the way for our current space-based experiments on board the ISS. Even today, there is a long list (that I've seen before, but can't find at the moment) of experiments that we want to put on the moon.
However, one crucially-important thing has changed between 1969 and today: robotics. We can send a robot to the moon and call it disposable, rather than have to also send fuel for the return trip home, supplies to sustain life, and a pressurized vessel to contain it all while the astronauts are up there. There's a reason the Apollo program required the largest, heaviest, and most-powerful rocket ever flown: Putting mass into space is exponentially expensive. Each Apollo mission cost (on average) about four times as much as the whole Mars Science Laboratory program.
By sending robots to the moon (and Mars, and elsewhere), we can continue to reap the scientific benefits without literally burning American tax dollars and risking American astronaut lives. Once there, the robots can last for much longer than a human, running experiments until they fall apart... and then just a bit more. Frankly, robots are superior explorers to humans in just about every way except for three.
First, robots aren't as adaptable as humans, though they are getting better. Space-bound rovers are designed with adaptability in mind, and the engineers controlling them from Earth are brilliant at remote repair and alternative uses, but a rover won't likely be able to recover from an accidental roll down a hill, even if the damage is minimal.
Second, robots are still limited in their capability. We can't just drop down a new camera and say "here, use this." There has been some work into making reconfigurable robots that could upgrade themselves, but ultimately it's still just cheaper and easier to send a new set of wheels with the new camera.
Finally, robots just don't make good humans. Humans are fragile and sensitive, and we get so upset when one is damaged and is... decommissioned. If the goal is for humans to leave Earth and look towards colonizing other planets, we still have a lot of questions to answer about how to keep those people safe and healthy. That's why we have the ISS. There are a lot of ongoing experiments running on board the ISS, and that's satisfying our current science needs (and exhausting what little budget we have).
In summary, that's why we are where we are today. We use the relatively-nearby ISS to run human-based experiments, and send expendable robots to further places, maximizing the scientific knowledge gain while minimizing the expense of rockets, engineering, and lives. As much fun as it would be to fling more humans at that floating gray target for the sake of patriotic glory, it really doesn't contribute much to mankind's future. We've already taken the giant leap that was beneficial in 1969, by starting extraterrestrial exploration. The next one will be a permanent colony, but we're not quite ready for that yet, regardless of which president wants it.
You have the required unsubstantiated claim and pretty decent bait, but overall it's not very catchy, mostly because it's almost completely detached from the subject of the parent post. It would have been more effective to first steer the conversation towards your bait, such as with a tie-in line like "The legacy airline software often has major bugs that have been left in because they're too hard to find and fix. I have to wonder if..."
You also cast your net too wide, by targeting three languages with wide dissimilarities. Just "Rust or Go" would have been more effective as a compiled choice, or "Node.js or Python" would target the interpreted languages, but combining the two without addressing the differences weakens your overall presentation.
Yeah, that's kind of my thought, too. I could understand a section of the White House being open for unsecured electronics, and a "official business" section being completely secured.
The White House already takes precautions with personal wireless devices, including by requiring officials to leave phones in cubbies outside of meeting rooms where sensitive or classified information is discussed.
"No electronics" is the standard practice for any classified space. I'd be extremely concerned if unsecured phones were allowed in classified meetings, but I can't really say I'd be surprised, considering this administration.
Alternatively, WikiLeaks could have consulted a few trusted security researchers to get any insight from the code, and released that insight with limited snippets of code. While that would likely aid attackers in making a similar infrastructure, they'd have to invent their own boilerplate, likely allowing the different reimplementations to be identifiable. The insight from the experts would also contribute more to coherent and realistic discussions on the actual capabilities of the tool, rather than encouraging more "the CIA is hacking everyone!" panic.
Even if the toolset had been sold to one "highest bidder", that would only be one other attackerto identify. The shared infrastructure would be a little confusing for researchers at first, but continued attacks would show distinct operation patterns as a signal rising above the noise. Yes, that does actually strike me as being more secure than opening the tools up to everyone at once, since it's now so much easier to hide any given attack in the higher amount of noise.
Zero-days and malware are just a part of the operation.
Any attack also requires an infrastructure to send the phishing emails, host fake login pages, make bogus links look trustworthy, and mask the origin of attacks. Often, setting up that infrastructure is the most time-consuming and expensive part of an attack, so it's often reused for several attacks. That is one of the most reliable mechanisms for identifying the source of an attack, by identifying the infrastructure networks used, and associating groups of attacks together, then connecting specific attacks with specific political actions.
Now, thanks to WikiLeaks, any attacker can start to build their own infrastructure from source, that looks just like the CIA. This in turn opens the door to more successful untraceable attacks and false-flag operations. By raising the banner of "journalism", WikiLeaks has yet again contributed to more damaging attacks and escalating conflicts.
Once upon a time, the term "journalist" carried a social expectation of trying to present the truth without harm. Dumping unfiltered source code doesn't offer any new insight except to a few good researchers, but it does enable significant harm and neuters those same researchers' usual techniques.
Because it's a $1 part, but it takes 3 hours of my time to prep, execute, and clean up the project. The benefit is that I get an old cassette deck back. However, since the vast majority of media I would use with that deck is already available to me on digital media, that isn't much of a benefit. Even once the machine is repaired, it's only going to work until the next piece fails, all of which already have 20 years of time on them since they were last known to meet quality standards. I could do a full rebuild, cleaning, and inspection, but that's also now a full day of effort, if not more.
Then there's the consideration for what else I could do in that time. I could play some games, read a book, go watch some YouTube videos, or a number of other things that I personally would find much more enjoyable than tearing apart a dusty 90's tape deck. That might be someone else's favorite hobby, but it's not really mine.
The decision is a lot more complex than simply saying it's a $1 part.
I agree in principle, but that statement isn't strictly accurate. It depends on the service and its implementation.
There is always someone who can handle your information when it's stored in a cloud service. If it's encrypted properly, they may not actually have access to view or modify it. A closer analogy would be a bank safe deposit box. While the bank has keys to the vault and is responsible for the vast majority of its upkeep, you're the only one with the key to look inside the box.
In the case of Twitter, I doubt there's much that's encrypted, if anything. Other services like Apple's iCloud have made a significant effort to encrypt the majority of their stored data such that it's only accessible by the client, but still manageable by the provider. Of course, the only real way to determine what's happening with your data is to read the service policies and terms of service.
I'd rephrase that a bit: Blockchains are appropriate where the effort required to corrupt the blockchain is significantly greater than the value to be gained by defrauding the other involved parties.
That's very similar to a basic axiom of game theory: Players will be honest when the perceived cost of cheating is higher than the perceived value to be gained by cheating.
In a private two-party system, a blockchain doesn't add anything, because either side can trivially rebuild the entire chain to fit their narrative. Publishing the chain brings in a trusted third party (a public record). In that case, the cost to cheat includes the cost to change the public record, which varies by implementation. In countries with endemic corruption, the cheating becomes simpler, because the trusted record-holder (say, a copy of the blockchain in a bank vault or published to a newspaper) can be changed for a certain fee (like bribing the bank officers or newspaper archivist). Even in a widely-published chain like BitCoin, the whole network can be adjusted by a quorum of nodes (though at great expense), and that in turn actually raises the value gained by cheating, because then it appears even more unreasonable to accuse someone (and all of the required accomplices) of the deception. When truth is determined by a simple majority, conspiracies become more appealing.
In comparing technologies, it is important to compare them equally. Attacks on blockchains are different from those in more traditional business models, so it is tempting to simply say "blockchain has none of these normal risks" and assume it is perfect. It makes a more reasonable comparison to ask "what is the cost to compromise this?" and compare the results of the analysis. That also allows a grounded discussion about whether the project in question actually warrants the increased complexity of having a blockchain, or if a more traditional system is still secure enough for the project's value.
I'll quote from my completely-made-up super-secret unredacted investigation document:
The Dallas witness said that an ambitious group of Soviet university students had discussed an assassination and met every Thursday in "Kampainstart Tavern", a bar in Moscow. Through CIA Agent Skip Towne, an operation was begun in Moscow (see referenced document #3.14) utilizing the American-sympathetic professor Dr. Doktor, resulting in the cooperation of Igor Tratorov, a student known to frequent the bar in question. The most potent truth serum available to Dr. Doktor (which was slightly less potent than the vodka available) was a standard dose of truthinol. Under the serum's (and vodka's) influence, Tratorov revealed that the discussion was actually a plan to assassinate the senior chief janitor at the Kremlin. Tratorov was presented the option to defect to the United States, but chose to remain in the USSR to complete his studies, while continuing to assist the CIA as an operative under Agent Towne.
This would add no useful information to the story of Kennedy's assassination, but it would endanger the lives of everyone named, as well as trigger a few dozen investigations into colleagues, friends, and family, some of which might still be assisting the United States through old connections. The named Agent Towne might no longer be playing the spy game, but if he started introducing another American in Moscow before leaving, that person would certainly fall under suspicion now. Poor Igor may have regretted his actions and gone on to support the Soviet (and later Russian) government faithfully, but now he'd face punishment for a decades-old capital crime.
Even if all of the people were safe, there is still intelligence value in knowing the methods used. In this example, "truthinol" is established to be similar in effect to alcohol. Combining that with another released or stolen document that says something along the lines of "factanol is shown to be ten times as effective as the old truthinol" means the strength of the modern truth serum can be estimated with reasonable accuracy.
That's how intelligence and counterintelligence works. It's a long game, played over generations and using corrupted people as pawns. Every agency knows what information it's looking for, and it constantly looks for any tangential information that might get it closer to its goal. With modern technology, public releases can be processed in a few minutes to find any useful pieces to an adversary's puzzle. Redaction of those releases is the best way to counter that capability.
Having worked in Perl (and many other languages) for about 15 years now, I'm curious how many of those polled actually use Perl regularly.
Whenever I have to introduce someone to my Perl scripts, their first reaction is usually the typical horror, which fades in a few days after they start using it. Yes, there are comments. Yes, there is decent design. No, the regular expressions are not worse than any other implementation. No, the "clever" one-liner you copied off of a PerlMonks golf challenge will not pass review.
Sure, there are a few weird warts on the language ("bless" being the most obvious example), but it's no worse than any other, and significantly better than some of today's much more popular languages. Mostly, I find that Perl just has a bad reputation because it allows you to write ugly code, just like C allows you to corrupt data and Java allows you to consume obscene amounts of memory. The language choice does not excuse being a bad programmer.
The obligation is only due to contract law. To avoid even that, the cards usually say things like "no cash value" or "may be deemed invalid at any time", and usually carry expiration terms. The shop could also just close up and go out of business.
The only real reason gift cards maintain value is that there would be a PR nightmare if they were suddenly invalidated. That risk is a liability to the company, countering the financial asset the company gained by selling them originally. It is removed by fulfilling the value of the card, which is itself a liability to the company. All of the real on-paper assets and liabilities cancel out, leaving just an intangible asset: Forced customer loyalty.
If I hold a gift card for a restaurant, for example, I can go there and essentially eat for no additional charge. However, I'm likely to bring a friend, or order just a bit more than the card covers, or even just remember the good meal and come back later. Gift cards (and most other kinds of scrip, for that matter) serve as an easy and convenient way to get customers in the door, and hope that they'll spend more in the long run.
If we assume that there is a pocket of antimatter a few trillion light years away, there are more questions raised:
Why is our matter-favoring pocket so big?
Why don't we see smaller antimatter pockets?
Why don't we see evidence in the cosmic microwave background?
And still, why is there any bias in some parts of the universe over others?
There are lots of examples of region discrepancies discovered throughout history. Land meets ocean, atmosphere meets fades into vacuum, and the solar system meets interstellar space. There are even fairly clear boundaries between galaxies, but all of those regions are explainable within our laws of physics, using testable models. We can make an experimental coast by testing the material properties of rock. We can compute the strength of gravity, and thus determine the size of celestial regions in which a given body will dominate.
We do not have any theory (at least, that corresponds with observations) describing how such pockets could exist, even outside our observable universe. Thus far, all such theories require the assumption that the laws of physics simply do not apply outside of a completely arbitrary boundary within which we happen to exist.
The problem with that theory is that it assumes we're in a "special" region of the universe, and some other area is different.So far, that hasn't shown to be the case.
To use your example, let's say that the first distribution was wildly uneven, with about 75% of the antimatter in one half of the newly-developing space. For simplicity's sake, we'll say that our observable universe is perfectly equal to the matter-dominant side of that split. Now, we should be able to observe every particle, and find that it's matter-to-antimatter ratio is 3:1. That's fine. We should also be able to look at old (distant) regions, and see back to when the universe was still undergoing those distributions, and we should see the results of other uneven distributions. We should see some antimatter-heavy regions and some matter-heavy regions, though we'd still expect to see that general 3:1 ratio.
We don't see that, though. Instead, we've seen no sign of any antimatter-heavy regions anywhere in observable space, regardless of age. This would imply that if such an uneven distribution happened, it happened only before any of our observable universe formed, and all expansion afterward has been perfectly homogenous matter. That's where the probability becomes very unlikely. It's not unlikely to have randomly-bad distributions. It's unlikely that there would be no further sign of such events, if they were prevalent enough to cause our whole observable universe to be so biased.
This is a common practice already, but now with added buzzword-compliance.
For decades, organizations have issued scrips of various kinds. From gift certificates and coupons to the ubiquitous gift cards exchanged today, there's always some new way to get customers to invest in your product before they buy it. This guy now has his own scrip currency, with the gimmick of being a "cryptocurrency" so people can generate their own, essentially paying him in their time and recognition of his brand instead of an actual recognized currency.
Your car will work great in this neighborhood, but once you leave the community onto the decade-old main road, you lose the self-driving ability. Then that road gets upgraded (at great expense to the taxpayers, with the latest sensor design), but the new vendor doesn't speak the same protocol or make the same decisions as the old one, so the boundary between the two systems becomes an uncertain zone, with some weird issues reported by drivers, as the two systems disagree on the best course for the vehicle.
Eventually, even with full compatibility, you're still describing upgrading the entire road system with beacons and sensors, and building a distributed computing system. Sure, you can break it into parts, but the whole system must eventually exist.
In comparison, a swarm is even more compartmentalized, but never needs the system-to-system integration. Each vehicle is its own self, and the nationwide coordination is just engineered emergent behavior. In such a swarm, even systemic corruption is limited to putting one vehicle at risk.
Yes, there is a human factor involved, but that's because we are not "building a[n] autonomous system from scratch". We are building an autonomous system that has to exist in an environment populated by humans who are already occupied with their own lives, processes, and priorities.
The "other methods" have been tried, on both limited and large scales (as far as such experiments are concerned, at least), and we've learned many valuable lessons from them. For example, we have elevators, which are centrally-controlled fully-automated one-dimensional travel in a very tightly-controlled environment. They work almost perfectly, but still need a few key features to ensure safety when that tightly-controlled environment malfunctions. As another example, there are several places with fully-automated train systems, with human operators only present to hit emergency stop buttons if something unusual happens. Those systems still have had a few major failures when sensors failed invisibly or the humans didn't notice something amiss.
I'm sorry, but as an ex-roboticist, I have to disagree. The centralized approach is actually expected to be significantly more difficult.
Driving isn't the problem. There have been partially-automated train systems for decades, and some fully-automated ones more recently. We can easily make a system, centralized or not, to get travelers to their destination. The far more difficult aspect is dealing with the unpredictable interference. Children run into the streets. Animals think train tracks are comfortable beds. Storms knock down trees and flood routes. Every winter, potholes turn small cracks into large hazards.
To detect those problems, we have two approaches. The centralized approach is to have a vast array of sensors constantly monitoring every foot of roadway. That's a lot of sensors, so even being cheap inductive loops still puts the total cost in the billions. Unfortunately, "cheap" and "secure" are often mutually-exclusive. If the sensors can be hacked at scale, reality starts permitting the movie plots involving forcing an armored car or emergency responders to take the criminals' chosen route. Of course, with that many sensors, you also need a massive infrastructure project (and budget) to handle the input. That centralized coordinating computer has to be a supercomputer, even with modern processing, just to properly handle the ever-changing status of the roads. That's not even including the routing and coordination aspect, which would also need to scale as people are traveling. Coordinating a few million vehicles in a hurricane evacuation is no small feat.
Fortunately, the other approach makes a lot of those scaling issues disappear. By having a swarm of autonomous vehicles, the total sensing domain is limited to what's in the vehicle's immediate area, and to a lesser extent what will be in its future route. Rather than monitoring the whole road space, each vehicle can monitor just the road it's interested in. Detected hazards can be communicated to other vehicles, but that's merely advice given out of courtesy. Each vehicle looks out for itself, and as such the available processing capability naturally scales with the processing capability that is required. As technology improves, the new technology is deployed with new vehicles, remaining compatible with old vehicles operating in the same shared roadway, including those with no autonomous function. They are treated as minor hazards, just like any other object in sight.
What's your point? That gently-used devices last longer than heavily-abused ones?
My acquired-for-free salvaged iPhone 4S also lasted for about 4 years before I replaced it, and still functions decently for the rare occasions that I need a second device.
On the other hand, I've broken a Craftsman wrench on its first day of use.
The key here, as noted in the summary, is that Craftsman is expecting to replace tools at any time, and they factored that replacement into the initial price of the tool. As a company, Sears (now Stanley Black & Decker, apparently) has chosen to be contractually responsible for replacing the tool if or when it breaks. Other companies and other brands have made other choices, but they're all willing participants. Apple is arguing that if they were held to the terms of a carrier contract, they would be bound to a contract in which they have no voice, since the contract is between the end user and the carrier.
Slashdot Mirror
since blocking/throttling/etc is purely anti-competitive
Be careful with that line of thought.
"Anti-competitive" has very specific legal meaning, usually requiring that the behavior benefit the actor in some way. However, exclusive-access deals are not often seen as anti-competitive, since any other company could have made a better bid for access.
As an example, let's say that an ISP launches their own video-streaming service, and it's the only one that gets full bandwidth, while Netflix and Hulu and throttled. That's an easy case for anti-competitive behavior.
However, rather than launching their own service, they can open up bidding to be the "exclusive streaming video provider" for their network, and allow Netflix and Hulu (and everyone else, including the tiny little startup with no budget) to bid for that exclusive contract. In the end, the ISP still makes millions of dollars for throttling video, and the consumers still have very little option to move to other ISPs. Even discounting municipal monopolies (which are themselves just exclusive contracts), every other ISP is free to enact the same cash-grab policies.
Do note that in that context, it's easy to argue that since the startup and the major players are competing for the same contract, it's actually pro-competition! It's not discriminating against small companies; it's providing them an opportunity to win a business partnership for a growth opportunity! Of course, the little startup has no real chance to match the bid of the established companies, but to those who like this plan, that's just an indication that the small streaming service should grow more first, perhaps by competing for exclusivity with a smaller ISP for which the big players won't pay as much.
This scenario is a direct parallel to how a brick-and-mortar business grows, by getting local contracts near the company's physical home, that probably won't be noticed by the industry's major players, and probably isn't worth sending a sales rep out for. On the Internet, though, everything is global. The big players can afford to bid on every single opportunity, so a small stream service will likely never find a open niche for growth.
It seems to me that the best option would be to provide more options.
As a creator, let me pick who's covering the processing fees. If I'm aiming for wide appeal, it might be beneficial to pay my own fees, just to keep the apparent prices low. If I have a small appeal to some highly-contributing individuals, a few cents in processing fees won't change our relationship that much.
As a patron, let me pick when and how I'll be charged. Let me pick a schedule, and just show me the fees associated with it. If I'm picking something that causes costs to rise, I'm fine paying for it, but give me the option up front. I might be fine shifting my payment schedule a few days if it cuts down the fees, but let me make that choice. To compromise with the creator, a middle-ground option might be that the creator picks a plan for which he's willing to pay the minimized fees, and the cost of any deviations from that schedule would be paid by the patron choosing to be different.
As a business, Patreon can do a few other things to mitigate the apparently-high risks involved. Rather than passing chargebacks directly through to the creators, Patreon could hold a one-month buffer of donations, collecting the individual charges (and chargebacks), then releasing them to the creators 30-60 days after the charge deadline. Essentially, Patreon itself takes the role of a clearinghouse, maintaining account balances and payouts for each creator. Having floating balances would also open the door to allow creators to support other creators directly from their accounts (preferably with reduced fees), promoting collaboration under the Patreon banner.
Now, implementing these choices would be a significant development and logistics effort, but certainly possible.
It does usernames, but no wildcards.
When Google was founded, its motto was "don't be evil". Uber's motto seems to have been "be very evil".
It'll certainly be interesting to see how Uber's practices change as it matures... if it survives.
You could also look at Troy Hunt's FAQ and blog, where he specifically states that there is no record of searches on the site (beyond server crash logs and non-scraping analytics), but that would require actually trusting a well-respected infosec expert.
I've used that on isolated systems before, in days long gone by. Now, many login systems recognize that the username and password partially match, and it rejects the pair.
The best I know of is https://haveibeenpwned.com/. You can search for a single email address, or set up monitoring for your domains.
If this collection has email addresses, I wouldn't be too surprised to find it added to the collection there.
Legally run OS X.
For some folks, that's justification.
I'm a bit dry at the moment... I know I had a can of Instant Froth around here somewhere...
Much better... Now then...
It's not a question of being "anti-science" or not, but more the question of "why?"
In the 1960s, landing on the moon was a huge accomplishment. We conducted important science, established permanent lunar installations of ongoing significance, and it paved the way for our current space-based experiments on board the ISS. Even today, there is a long list (that I've seen before, but can't find at the moment) of experiments that we want to put on the moon.
However, one crucially-important thing has changed between 1969 and today: robotics. We can send a robot to the moon and call it disposable, rather than have to also send fuel for the return trip home, supplies to sustain life, and a pressurized vessel to contain it all while the astronauts are up there. There's a reason the Apollo program required the largest, heaviest, and most-powerful rocket ever flown: Putting mass into space is exponentially expensive. Each Apollo mission cost (on average) about four times as much as the whole Mars Science Laboratory program.
By sending robots to the moon (and Mars, and elsewhere), we can continue to reap the scientific benefits without literally burning American tax dollars and risking American astronaut lives. Once there, the robots can last for much longer than a human, running experiments until they fall apart... and then just a bit more. Frankly, robots are superior explorers to humans in just about every way except for three.
First, robots aren't as adaptable as humans, though they are getting better. Space-bound rovers are designed with adaptability in mind, and the engineers controlling them from Earth are brilliant at remote repair and alternative uses, but a rover won't likely be able to recover from an accidental roll down a hill, even if the damage is minimal.
Second, robots are still limited in their capability. We can't just drop down a new camera and say "here, use this." There has been some work into making reconfigurable robots that could upgrade themselves, but ultimately it's still just cheaper and easier to send a new set of wheels with the new camera.
Finally, robots just don't make good humans. Humans are fragile and sensitive, and we get so upset when one is damaged and is... decommissioned. If the goal is for humans to leave Earth and look towards colonizing other planets, we still have a lot of questions to answer about how to keep those people safe and healthy. That's why we have the ISS. There are a lot of ongoing experiments running on board the ISS, and that's satisfying our current science needs (and exhausting what little budget we have).
In summary, that's why we are where we are today. We use the relatively-nearby ISS to run human-based experiments, and send expendable robots to further places, maximizing the scientific knowledge gain while minimizing the expense of rockets, engineering, and lives. As much fun as it would be to fling more humans at that floating gray target for the sake of patriotic glory, it really doesn't contribute much to mankind's future. We've already taken the giant leap that was beneficial in 1969, by starting extraterrestrial exploration. The next one will be a permanent colony, but we're not quite ready for that yet, regardless of which president wants it.
I give your troll about a 4/10.
You have the required unsubstantiated claim and pretty decent bait, but overall it's not very catchy, mostly because it's almost completely detached from the subject of the parent post. It would have been more effective to first steer the conversation towards your bait, such as with a tie-in line like "The legacy airline software often has major bugs that have been left in because they're too hard to find and fix. I have to wonder if..."
You also cast your net too wide, by targeting three languages with wide dissimilarities. Just "Rust or Go" would have been more effective as a compiled choice, or "Node.js or Python" would target the interpreted languages, but combining the two without addressing the differences weakens your overall presentation.
Better luck next article.
Yeah, that's kind of my thought, too. I could understand a section of the White House being open for unsecured electronics, and a "official business" section being completely secured.
The White House already takes precautions with personal wireless devices, including by requiring officials to leave phones in cubbies outside of meeting rooms where sensitive or classified information is discussed.
"No electronics" is the standard practice for any classified space. I'd be extremely concerned if unsecured phones were allowed in classified meetings, but I can't really say I'd be surprised, considering this administration.
Alternatively, WikiLeaks could have consulted a few trusted security researchers to get any insight from the code, and released that insight with limited snippets of code. While that would likely aid attackers in making a similar infrastructure, they'd have to invent their own boilerplate, likely allowing the different reimplementations to be identifiable. The insight from the experts would also contribute more to coherent and realistic discussions on the actual capabilities of the tool, rather than encouraging more "the CIA is hacking everyone!" panic.
Even if the toolset had been sold to one "highest bidder", that would only be one other attackerto identify. The shared infrastructure would be a little confusing for researchers at first, but continued attacks would show distinct operation patterns as a signal rising above the noise. Yes, that does actually strike me as being more secure than opening the tools up to everyone at once, since it's now so much easier to hide any given attack in the higher amount of noise.
Zero-days and malware are just a part of the operation.
Any attack also requires an infrastructure to send the phishing emails, host fake login pages, make bogus links look trustworthy, and mask the origin of attacks. Often, setting up that infrastructure is the most time-consuming and expensive part of an attack, so it's often reused for several attacks. That is one of the most reliable mechanisms for identifying the source of an attack, by identifying the infrastructure networks used, and associating groups of attacks together, then connecting specific attacks with specific political actions.
Now, thanks to WikiLeaks, any attacker can start to build their own infrastructure from source, that looks just like the CIA. This in turn opens the door to more successful untraceable attacks and false-flag operations. By raising the banner of "journalism", WikiLeaks has yet again contributed to more damaging attacks and escalating conflicts.
Once upon a time, the term "journalist" carried a social expectation of trying to present the truth without harm. Dumping unfiltered source code doesn't offer any new insight except to a few good researchers, but it does enable significant harm and neuters those same researchers' usual techniques.
I'm unimpressed.
Because it's a $1 part, but it takes 3 hours of my time to prep, execute, and clean up the project. The benefit is that I get an old cassette deck back. However, since the vast majority of media I would use with that deck is already available to me on digital media, that isn't much of a benefit. Even once the machine is repaired, it's only going to work until the next piece fails, all of which already have 20 years of time on them since they were last known to meet quality standards. I could do a full rebuild, cleaning, and inspection, but that's also now a full day of effort, if not more.
Then there's the consideration for what else I could do in that time. I could play some games, read a book, go watch some YouTube videos, or a number of other things that I personally would find much more enjoyable than tearing apart a dusty 90's tape deck. That might be someone else's favorite hobby, but it's not really mine.
The decision is a lot more complex than simply saying it's a $1 part.
I agree in principle, but that statement isn't strictly accurate. It depends on the service and its implementation.
There is always someone who can handle your information when it's stored in a cloud service. If it's encrypted properly, they may not actually have access to view or modify it. A closer analogy would be a bank safe deposit box. While the bank has keys to the vault and is responsible for the vast majority of its upkeep, you're the only one with the key to look inside the box.
In the case of Twitter, I doubt there's much that's encrypted, if anything. Other services like Apple's iCloud have made a significant effort to encrypt the majority of their stored data such that it's only accessible by the client, but still manageable by the provider. Of course, the only real way to determine what's happening with your data is to read the service policies and terms of service.
I'd rephrase that a bit: Blockchains are appropriate where the effort required to corrupt the blockchain is significantly greater than the value to be gained by defrauding the other involved parties.
That's very similar to a basic axiom of game theory: Players will be honest when the perceived cost of cheating is higher than the perceived value to be gained by cheating.
In a private two-party system, a blockchain doesn't add anything, because either side can trivially rebuild the entire chain to fit their narrative. Publishing the chain brings in a trusted third party (a public record). In that case, the cost to cheat includes the cost to change the public record, which varies by implementation. In countries with endemic corruption, the cheating becomes simpler, because the trusted record-holder (say, a copy of the blockchain in a bank vault or published to a newspaper) can be changed for a certain fee (like bribing the bank officers or newspaper archivist). Even in a widely-published chain like BitCoin, the whole network can be adjusted by a quorum of nodes (though at great expense), and that in turn actually raises the value gained by cheating, because then it appears even more unreasonable to accuse someone (and all of the required accomplices) of the deception. When truth is determined by a simple majority, conspiracies become more appealing.
In comparing technologies, it is important to compare them equally. Attacks on blockchains are different from those in more traditional business models, so it is tempting to simply say "blockchain has none of these normal risks" and assume it is perfect. It makes a more reasonable comparison to ask "what is the cost to compromise this?" and compare the results of the analysis. That also allows a grounded discussion about whether the project in question actually warrants the increased complexity of having a blockchain, or if a more traditional system is still secure enough for the project's value.
I'll quote from my completely-made-up super-secret unredacted investigation document:
The Dallas witness said that an ambitious group of Soviet university students had discussed an assassination and met every Thursday in "Kampainstart Tavern", a bar in Moscow. Through CIA Agent Skip Towne, an operation was begun in Moscow (see referenced document #3.14) utilizing the American-sympathetic professor Dr. Doktor, resulting in the cooperation of Igor Tratorov, a student known to frequent the bar in question. The most potent truth serum available to Dr. Doktor (which was slightly less potent than the vodka available) was a standard dose of truthinol. Under the serum's (and vodka's) influence, Tratorov revealed that the discussion was actually a plan to assassinate the senior chief janitor at the Kremlin. Tratorov was presented the option to defect to the United States, but chose to remain in the USSR to complete his studies, while continuing to assist the CIA as an operative under Agent Towne.
This would add no useful information to the story of Kennedy's assassination, but it would endanger the lives of everyone named, as well as trigger a few dozen investigations into colleagues, friends, and family, some of which might still be assisting the United States through old connections. The named Agent Towne might no longer be playing the spy game, but if he started introducing another American in Moscow before leaving, that person would certainly fall under suspicion now. Poor Igor may have regretted his actions and gone on to support the Soviet (and later Russian) government faithfully, but now he'd face punishment for a decades-old capital crime.
Even if all of the people were safe, there is still intelligence value in knowing the methods used. In this example, "truthinol" is established to be similar in effect to alcohol. Combining that with another released or stolen document that says something along the lines of "factanol is shown to be ten times as effective as the old truthinol" means the strength of the modern truth serum can be estimated with reasonable accuracy.
That's how intelligence and counterintelligence works. It's a long game, played over generations and using corrupted people as pawns. Every agency knows what information it's looking for, and it constantly looks for any tangential information that might get it closer to its goal. With modern technology, public releases can be processed in a few minutes to find any useful pieces to an adversary's puzzle. Redaction of those releases is the best way to counter that capability.
Having worked in Perl (and many other languages) for about 15 years now, I'm curious how many of those polled actually use Perl regularly.
Whenever I have to introduce someone to my Perl scripts, their first reaction is usually the typical horror, which fades in a few days after they start using it. Yes, there are comments. Yes, there is decent design. No, the regular expressions are not worse than any other implementation. No, the "clever" one-liner you copied off of a PerlMonks golf challenge will not pass review.
Sure, there are a few weird warts on the language ("bless" being the most obvious example), but it's no worse than any other, and significantly better than some of today's much more popular languages. Mostly, I find that Perl just has a bad reputation because it allows you to write ugly code, just like C allows you to corrupt data and Java allows you to consume obscene amounts of memory. The language choice does not excuse being a bad programmer.
The obligation is only due to contract law. To avoid even that, the cards usually say things like "no cash value" or "may be deemed invalid at any time", and usually carry expiration terms. The shop could also just close up and go out of business.
The only real reason gift cards maintain value is that there would be a PR nightmare if they were suddenly invalidated. That risk is a liability to the company, countering the financial asset the company gained by selling them originally. It is removed by fulfilling the value of the card, which is itself a liability to the company. All of the real on-paper assets and liabilities cancel out, leaving just an intangible asset: Forced customer loyalty.
If I hold a gift card for a restaurant, for example, I can go there and essentially eat for no additional charge. However, I'm likely to bring a friend, or order just a bit more than the card covers, or even just remember the good meal and come back later. Gift cards (and most other kinds of scrip, for that matter) serve as an easy and convenient way to get customers in the door, and hope that they'll spend more in the long run.
That's the point, though...
If we assume that there is a pocket of antimatter a few trillion light years away, there are more questions raised:
Why is our matter-favoring pocket so big?
There are lots of examples of region discrepancies discovered throughout history. Land meets ocean, atmosphere meets fades into vacuum, and the solar system meets interstellar space. There are even fairly clear boundaries between galaxies, but all of those regions are explainable within our laws of physics, using testable models. We can make an experimental coast by testing the material properties of rock. We can compute the strength of gravity, and thus determine the size of celestial regions in which a given body will dominate.
We do not have any theory (at least, that corresponds with observations) describing how such pockets could exist, even outside our observable universe. Thus far, all such theories require the assumption that the laws of physics simply do not apply outside of a completely arbitrary boundary within which we happen to exist.
That's the realm of faith, not science.
The problem with that theory is that it assumes we're in a "special" region of the universe, and some other area is different.So far, that hasn't shown to be the case.
To use your example, let's say that the first distribution was wildly uneven, with about 75% of the antimatter in one half of the newly-developing space. For simplicity's sake, we'll say that our observable universe is perfectly equal to the matter-dominant side of that split. Now, we should be able to observe every particle, and find that it's matter-to-antimatter ratio is 3:1. That's fine. We should also be able to look at old (distant) regions, and see back to when the universe was still undergoing those distributions, and we should see the results of other uneven distributions. We should see some antimatter-heavy regions and some matter-heavy regions, though we'd still expect to see that general 3:1 ratio.
We don't see that, though. Instead, we've seen no sign of any antimatter-heavy regions anywhere in observable space, regardless of age. This would imply that if such an uneven distribution happened, it happened only before any of our observable universe formed, and all expansion afterward has been perfectly homogenous matter. That's where the probability becomes very unlikely. It's not unlikely to have randomly-bad distributions. It's unlikely that there would be no further sign of such events, if they were prevalent enough to cause our whole observable universe to be so biased.
This is a common practice already, but now with added buzzword-compliance.
For decades, organizations have issued scrips of various kinds. From gift certificates and coupons to the ubiquitous gift cards exchanged today, there's always some new way to get customers to invest in your product before they buy it. This guy now has his own scrip currency, with the gimmick of being a "cryptocurrency" so people can generate their own, essentially paying him in their time and recognition of his brand instead of an actual recognized currency.
It's a deployment problem, then.
Your car will work great in this neighborhood, but once you leave the community onto the decade-old main road, you lose the self-driving ability. Then that road gets upgraded (at great expense to the taxpayers, with the latest sensor design), but the new vendor doesn't speak the same protocol or make the same decisions as the old one, so the boundary between the two systems becomes an uncertain zone, with some weird issues reported by drivers, as the two systems disagree on the best course for the vehicle.
Eventually, even with full compatibility, you're still describing upgrading the entire road system with beacons and sensors, and building a distributed computing system. Sure, you can break it into parts, but the whole system must eventually exist.
In comparison, a swarm is even more compartmentalized, but never needs the system-to-system integration. Each vehicle is its own self, and the nationwide coordination is just engineered emergent behavior. In such a swarm, even systemic corruption is limited to putting one vehicle at risk.
Yes, there is a human factor involved, but that's because we are not "building a[n] autonomous system from scratch". We are building an autonomous system that has to exist in an environment populated by humans who are already occupied with their own lives, processes, and priorities.
The "other methods" have been tried, on both limited and large scales (as far as such experiments are concerned, at least), and we've learned many valuable lessons from them. For example, we have elevators, which are centrally-controlled fully-automated one-dimensional travel in a very tightly-controlled environment. They work almost perfectly, but still need a few key features to ensure safety when that tightly-controlled environment malfunctions. As another example, there are several places with fully-automated train systems, with human operators only present to hit emergency stop buttons if something unusual happens. Those systems still have had a few major failures when sensors failed invisibly or the humans didn't notice something amiss.
I'm sorry, but as an ex-roboticist, I have to disagree. The centralized approach is actually expected to be significantly more difficult.
Driving isn't the problem. There have been partially-automated train systems for decades, and some fully-automated ones more recently. We can easily make a system, centralized or not, to get travelers to their destination. The far more difficult aspect is dealing with the unpredictable interference. Children run into the streets. Animals think train tracks are comfortable beds. Storms knock down trees and flood routes. Every winter, potholes turn small cracks into large hazards.
To detect those problems, we have two approaches. The centralized approach is to have a vast array of sensors constantly monitoring every foot of roadway. That's a lot of sensors, so even being cheap inductive loops still puts the total cost in the billions. Unfortunately, "cheap" and "secure" are often mutually-exclusive. If the sensors can be hacked at scale, reality starts permitting the movie plots involving forcing an armored car or emergency responders to take the criminals' chosen route. Of course, with that many sensors, you also need a massive infrastructure project (and budget) to handle the input. That centralized coordinating computer has to be a supercomputer, even with modern processing, just to properly handle the ever-changing status of the roads. That's not even including the routing and coordination aspect, which would also need to scale as people are traveling. Coordinating a few million vehicles in a hurricane evacuation is no small feat.
Fortunately, the other approach makes a lot of those scaling issues disappear. By having a swarm of autonomous vehicles, the total sensing domain is limited to what's in the vehicle's immediate area, and to a lesser extent what will be in its future route. Rather than monitoring the whole road space, each vehicle can monitor just the road it's interested in. Detected hazards can be communicated to other vehicles, but that's merely advice given out of courtesy. Each vehicle looks out for itself, and as such the available processing capability naturally scales with the processing capability that is required. As technology improves, the new technology is deployed with new vehicles, remaining compatible with old vehicles operating in the same shared roadway, including those with no autonomous function. They are treated as minor hazards, just like any other object in sight.
What's your point? That gently-used devices last longer than heavily-abused ones?
My acquired-for-free salvaged iPhone 4S also lasted for about 4 years before I replaced it, and still functions decently for the rare occasions that I need a second device.
On the other hand, I've broken a Craftsman wrench on its first day of use.
The key here, as noted in the summary, is that Craftsman is expecting to replace tools at any time, and they factored that replacement into the initial price of the tool. As a company, Sears (now Stanley Black & Decker, apparently) has chosen to be contractually responsible for replacing the tool if or when it breaks. Other companies and other brands have made other choices, but they're all willing participants. Apple is arguing that if they were held to the terms of a carrier contract, they would be bound to a contract in which they have no voice, since the contract is between the end user and the carrier.