And that's to just use their services. To actually transfer via Bitcoin incurs Bitcoin's transaction fee, which is variable and depends on how long you're willing to wait. This fee occurs every time you transfer anything via Bitcoin and in independent of the exchange or whatever.
The fee goes straight to the miners as a reward for locking that transaction into the blockchain. Remember, miners do more than "make" bitcoin - in the early days they created actual bitcoin itself, but these days, with the reduction in production rate, miners make money locking the transaction into the blockchain.
So the more you pay, the more the miners get, so naturally when looking at all the pending transactions, they will order them from pays the most to least and then hash those into the blockchain. Pay among the top and it's confirmed within minutes. Be completely cheap and it'll take weeks.
I'm finding it hard to believe that this technology has not been available for a long time. Of course the system requites installation of the signaling systems. Here in the United States, major freight carriers have chaffed at even installing modern safety systems (The more we learn about Amtrak derailment the stranger it gets). But then again, those are "non-revenue generating" while this certainly has the potential. But also remember that unions have a say in manning. Again here in the US, unions fought tooth and nail when rail companies got rid of the caboose.
At the commuter train/LRT level, there are many systems that are completely automated - no drivers at all. Granted, the switching isn't too complex, but they can be commanded to switch tracks and manage themselves with supervision done at a central control station.
Operating at the full size train level is only a slightly more complex problem, mostly because now the tracks are owned by many people and you really need to get them all to install a common communication an d signalling system. (Right now trains can be monitored remotely).
The problem with Amtrak is also been solved - the technology has long existed and it's nowadays called Positive Control. The train will periodically beep and the driver has to hit a button. If the driver fails to hit the button, the train is brought to a stop automatically. This helps catch distracted and sleepy drivers.
Even speed limit enforcement has been automated away - if the train is coming too fast for the speed limit, the train automatically slows down. The unions generally hate this as it reduces the driver to a monkey. OTOH, the safety record of these systems is quite stellar, and most rail lines only experience it once before the system is rolled out on all the rolling stock.
Though, the Shinkasen drivers of Japan, where they not only have these systems in place but also very rigorous protocols and how they act (they lift their arms up as if to salute, then point at the control they are going to adjust, then adjust it - it looks like part of a military march) don't seem to be monkeying around, but seem to be very professional about it. Even though there is speed limit enforcement, they still can control the speed of the train, and still can get the train to be within a minute of the stated arrival time, even if weather conditions make it so the train has to slow down (it snows in Japan, too, and the trains have to operate in it at reduced speeds).
f you imagine it's a new browser and don't pine for the fjords^W old Firefox UI it isn't bad.
Well, firefox makes you edit text files to adjust the UI, so if you're a greybeard that should be a pleasant blast from the past. These days though, editing text files to adjust the UI seems very antiquated and a step into 30 years ago when we did it only because there was not enough CPU power or memory to graphically edit UIs so it was a cycle of edit, relaunch, test, quit,... I get it', it's more secure when websites can't redo your browser UI on you, but still, having to hunt for third party websites that have the magic incantations to adjust the UI and lack of an editor really put you back.
Or it can do a combination of signal strength and speed and try to figure it out that way. After all an 802.11ac signal is worthless if it's weak but you have a stronger N signal nearby or things like that.
A long long time ago, I had a WiFi card that could connect to an 802.11g AP on the 10th floor from the 3rd through a concrete-and-steel building. This was annoying because I had an 802.11b AP nearby that would because of its proximity and greater signal strength be faster than the 802.11g one would because the distance.
And yes, it actually did work, but wasn't that fast.
Even without providing rodent buffets, cars will be shorting out in a few years due to the wire insulation BIODEGRADING. If these executives and engineers think it's such a great idea, let's see them use this wiring in their homes.
Organic parts can be made to last longer. In fact, a boat manufacturer replaced their shaft seals with new ones that lasted 2 to 3 times longer than the existing rubber ones, in testing. Good change, right? Customers can buy seals that last twice as long as the old ones (thus halving the number of tricky seal changes, too, saving money all around since you wouldn't need to dry dock the boat as often, labor, etc).
Problem was, it was soy based I believe, so critters started getting to them and after the summer was over, there were a lot of sunken boats.
Moral is, the soy stuff can genuinely be better from all objective points - it can last longer in the varied conditions a car goes through without the wiring getting brittle and cracking and causing short circuits. It just had an unexpected side effect that critters love to eat it. Even that can be taken care of - perhaps thicker insulation with the application of a bitterant to basically make the wires not tasty at all. (And since bitter is a flavor associated with poisons, it also appeals to the critter's survival instincts that what they're eating is probably bad for them).
and how exactly does that do anything at all to improve the situation? or are you suggesting Open source hardware would somehow be magically design flaw free?
And fundamental problems are still fundamental problems. The reason practically every processor has the same issues is because the same optimizations we used to make processors faster had the same fundamental design error.
I mean, either someone designed the core branch predictor block and everyone worldwide copied it for every processor, or everyone implemented it differently, yet it has the same Spectre flaw, implying that the flaw is inherent in the way branch predictors work.
The only way you can guarantee the designs are error free are to abandon everything that makes modern processors fast - OOO, speculation, branch prediction, and plenty more, including potentially pipelining (the fundamental technology everyone is trying to speed up by avoiding pipeline stalls). Go back to the old fetch-and-execute cycle and where memory operands are fully decoded and retrieved prior to even considering fetching the next instruction.
Everyone will hate it, because now your 4GHz processor will be as fast as a 500MHz one.
It's ironic how some of the biggest internet companies couldn't figure out how to get work done without having everyone physically present. All of these location-dependent problems would go away if they just allowed their employees to telecommute.
Some problems are solvable by telecommuting, some problems aren't.
Stuff that involves hardware (which hey, Apple does a lot of) really cannot be done by telecommuting until you get full telepresence robots (and even then, some things really require you to be there - engineer's knack, for example). Expensive test equipment may be utilized, or even just general tools like oscilloscopes. And then there's physical hardware that requires transportation, Apple's already secretive enough that new hardware has to be locked up and multiple layered security (so fat chance bringing it home).
There's also a lot of value in bringing the team together - collaboration can be much faster. Perhaps you come up with a cool idea, but then instead of just walking down the hall and presenting it, you have to type it up in an email or worse yet, the company IRC-like chat room (slack, whatever) and everyone misunderstands you so you spend another hour re-describing it. And heaven forbid you need to draw something...
If you can stay back a few nodes, not so bad, a few million bucks is needed (masks are expensive at about $100K/each for the older processes), so perhaps a regular 10 metal chip requires a couple million bucks.
And while most of it is autorouted and autoplaced, you still want to hand edit the designs. Remember the reason we're at 10 metal is because for most general random logic, the limiting factor is wiring. The vast majority of transistors in any design is used in memory - caches and the like. The density falls off rapidly when you get to general logic blocks like ALUs and registers and other random logic parts. And because of the low density, you edit the layouts to sprinkle in lots of new transistors that do nothing. This is because when you're fixing bugs, you're going to need extra logic parts, and having them already in the silicon itself means only changing a few masks instead of an entire mask set.
That's why there are steppings - A0 to A1 to A2 are metal-only changes and A0 meant a full mask set, but A1 and A2 only had marginal changes so perhaps 5 masks changed (so instead of spending $2M per pop, you spent an additional $1M). But go from A2 to B0 means you changed everything again, so a complete redesign. The reason for this is you ran out of spares, or more likely, the fix involves speed-sensitive paths and you just cannot route the spares in a way to keep the speed up.
Anyhow, I still don't see how an open design would avoid something like Meltdown or Spectre, because those vulnerabilities came about because of novel ways of using the weaknesses of things we did to make processors so much faster, and given the entrenchment of the designs, everyone implemented the same issues, so it's really an issue going all the way back.
Puerto Rico doesn't want to become a state because being merely a territory means Puerto Ricans don't have to pay Federal income tax (unless they work within a U.S. state or for the government). As a territory, they get nearly all the benefits of statehood, without having to pay Federal income tax.
It also means that they are under the control of Congress and abide by all the laws that it passes. But even bigger, they have no representation at all. US territories follow US laws and all that, but have no representation in the House or the Senate, so they don't have a say in any laws that get enacted.
That's why they aren't taxed - perhaps the old refrain "no taxation without representation" sound familiar? It also means those territories are mostly forgettable - they're required to follow US laws even though it can be bad for their own interests (because no one in the House or Senate cares, or knows about, those issues).
Seems like there is more damage being done by the Meltdown / Spectre hysteria and attendant patch frenzy than by exploits of the vulnerabilities themselves.
That's because the most serious vulnerabilities affect virtualized systems more, and given the popularity of the cloud, they are extremely affected because there is no control over what software is run.
These attacks may seem theoretical, and for a desktop user, they mostly are. Even server users with dedicated machines are generally safe as they own the entire machine and thus control the software on it. Virtualized hosts, not so much, and this ranges from simple VPS hosting companies, to cloud service companies like Amazon and Google. In this case, accessing not-your-memory can be beneficial, like trying to find the encryption keys used by someone sharing your machine. Cloud providers are especially scared because during high peak retail times, encryption keys may be spun to many instances to handle the load, and crafty attackers may simply spin up instances to extract those keys.
Meanwhile, your checks, credit card transactions, banking accounts, insurance, merchant payments pass through virtualized systems. Everyone here depends on virtualized systems. 100% of the people here depend on virtualized systems, and it isn't just Intel and AMD that have speculative execution bugs, the POWER chips in IBM Z Mainframes and AIX boxes do too...
Even worse, IBM has only been releasing half-patches, the complete fix is due in mid-February. With the speed that Intel and AMD have released fixes, and OS vendors have too, IBM seems slow off the mark.
And I won't really call it a bug, but an optimization people did that no one realized would open up serious security vulnerabilities. It's a cheat because you're not doing all the proper checks when you should, because if you did, you'd slow down and stall the pipeline. Instead, you begin speculative execution, kick off tasks to fetch operands and other metadata (including security and maps) and hope by the time the instruction needs the information, it's there.. It's why we moved from fetch-and-execute cycles to full OOO execution - the speed up is immense. It's also why everyone did it, and why we're all seeing the result now.
Of course, doing it right avoided the issues, but then your processor ends up slower and thus, less competitive.
And I see them with the huge phones in their hands (hands, plural, because they can't hold it in one hand - too big), so that isn't accurate at all. It's interesting to watch them wrangle a 6" phablet in two hands, which obviously was bought because they needed the screen size than practicality.
Anyhow, apparently the real reason is the iPhone X sales in China is disappointing. Apparently the screen is too small. If your phone requires at least two hands to hold it, it's the right size. If you can hold it in one hand, it's too small. If you require a third hand, bingo, you got the best phone out there.
Apparently size over practicality. Even if it requires two hands to hold it so you can make a phone call (using voice assist, because with two hands holding the phone, there are no spare digits that can reach the numbers).
And yes, they flash the phone everywhere. It is a status symbol, and the iPhone X unfortunately is not "bling-y" enough.
Indeed. You can see them recycled in giant piles at the scrapyards. The problem is we recycle things of value. Salvageable parts, metals, even the copper in cabling is good. You know why the recycling market pays more for raw copper rather than for cables per weight? Because of the effort needed to strip out the copper (recycled) from the plastic (not recycled). Cable sheathing is made of many different types of plastics and rubbers. No recycler ever puts the effort in to recycling the sheaths when the copper is worth 100x more to them.
Burning plastic (fastest way to strip a wire) has the lovely effect of breaking it down into small bits. You successfully achieve what the landfill does in minutes rather than years.
Cars are heavily recycled, to the tun of about 80-90% of a car is recycled back to raw materials. The rest of it is dumped. Car manufacturers have been trying to increase the amount that you can recycle, or reduce the amount that stays stable when dumped (i.e, most plastics). A lot of stuff like fiber reinforced materials cannot be recycled - the fibers make it heavily impure and a pain to remove. It's why recycling fiberglass turns it a rough powder that is then injection molded.
Burning plastic isn't actually all that great - it generally releases a lot of nasty chemicals and solvents in the air that are not good for the environment. It's also hard, because a lot of car wiring has to withstand high temperatures, and often are fireproof, which makes burning them in incinerators even nastier. That's why they switched to more organic materials - burning generally isn't as bad (but it's still horrible - as any crematorium will know, burning dead bodies releases nasty compounds even though we're "natural").
Biodegratable ones can be degraded in special environments that rapidly decompose the materials - and can be designed to last basically forever under normal environments, but put them in special composters they will be gone in weeks, and generally release byproducts that can be captured and used (e.g., methane)
The actual reason Australia doesn't have shutdowns like the USA is because bills of supply and appropriation shall not contain matters not related to supply or appropriation. I.e. It's not possible to discuss a budget while tacking on some stupid rider like DACA protections or CHIP. The only thing you can discuss is the budget and they are segregated into multiple documents that it is very unlikely for a single supply bill to shutdown the government. The only reason this happened in 1975 is because the fundamental fight was on the funding of the government and the loans the government was making.
Whereas the USA sees the funding bills as opportunities to wave cocks around and force the other party to pass something unrelated.
Canada does something similar to the US too - we call them "omnibus" bills that seem to do a million unrelated things at once alongside the budget.
Now, that doesn't mean we suffer from the US problem, as in the Westminster system, voters do not elect the prime minister - the prime minister is chosen - first among the party that has the majority, but if that fails then the remaining parties choose a prime minister. So the prime minister and the party or coalition of parties is the same, and they just ram the bill through. In a majority situation, this is simple - you put it up and vote and it passes. In a minority situation, it's a lot dicier.
Thing is, it either passes, or it doesn't. Budget bills are "confidence" matters - if it doesn't pass, it's a general resounding vote that the government is now ineffective and needs to be replaced - either by the opposition (if they can prove they can lead) or another election.
Yes, it's also why minority governments are generally unstable - confidence matters mean everyone else can vote no and force government to fall.
CarPlay isn't entirely free, however. As Markdown inventor and Apple guru Jon Gruber pointed out on Twitter, car manufacturers who wish to officially support Apple products must pay a licensing fee to enter Apple's Made for iPhone (MFi) program, just like any other licensed accessory maker.
Which is only true if the automaker provides a cord ending in a Lightning connector. If they just have a USB socket and tell you to BYO Lightning cable, that fee vanishes as well.
Even so (face it, BMW could just toss you an Apple USB to Lightning cable with the car. It ain't hard), the license fee is really just a once time fee for that item. You go through Apple,. get your stuff approved (ti can be trying, because Apple does do a lot of testing, including security testing) and get the license to use the logo. It's a one time fee and payment, not an ongoing thing (otherwise manufacturers of other things like speakers and headphones would be on the hook for payments as well).
And seriously, $300 to use CarPlay in a $60K+ car? Really, people are so cheap to not add that?
However from a layman's view you might compare this to putting a guy on a train that is already moving at top speed, then later rewarding him for not fucking it up.
You know how freaking hard that is?
Usually someone comes in and wants to dabble with this and that to "put their mark on it" and that leads to all sorts of crap going on and changes for the worse.
To h ave someone able to put all that aside and let things run is quite unnatural. SO yes, the fact it isn't all screwed up by now is amazing. The fact it's sold twice as much since acquisition is amazing.
Also, how many times have you heard of X acquiring Y and Y getting completely messed up? It happens almost always despite promises to the contrary. Thus, if Minecraft is even more successful under Microsoft, it took a lot of restraint in not screwing it up.
And for someone to manage a game studio, I suppose that's an excellent trait - if Microsoft were to acquire a bunch of studios, this guy would likely have a very light touch on things so it likely wouldn't be Microsoft that screwed up.
Wrong. 5S is not included in the cheap batteries, despite being still supported with IOS 11. Only 6 and newer are getting cheap batteries and there is already a 3 month wait time.
Because the 5S is not affected by this. Only the 6 and above have the power management feature and can scale back the processor speed to lower battery consumption and are thus affected by bad batteries.
But really, cheap 5S battery replacements are available quite easily - all those mobile repair shops can trivially change the battery for about $30 or so. And I think iFixit and other online stores can sell you DIY kits for around the same price.
David Ige, the governor of Hawaii has said this has been a "learning experience" for everyone involved, that it will not turn into a witch hunt, and no one will lose their job. In other words, there will be no accountability or consequences, and the same serially incompetent bozos will remain in charge.
You're falling into the "we must fire someone for accountability" trap.
That leads to basically incompetents running your ship - if everyone is deathly afraid of losing their job for making a mistake, you end up with a corporate culture of timidity, cover your ass and hiding mistakes.
The modern method is not to fire the person who pushed the button, but to find out the true reason. This is often' called "The Five Whys" because it literally asks Why over and over again.
Like in this case, given what we know.
Why was a missile alert called? Because someone clicked the link to send it. Why did they click that link? Because they clicked the wrong link - they meant to click the one that produced a test message instead. Why did they click the wrong link? Because the links were presented as an unsorted list, with the test links appearing on some events ahead of the real link, and sometimes afterwards. Why did they click the wrong link? Because when you're looking at a huge list of unsorted links, you tend to focus on the one that matches what you're wanting even though it may not be exactly what you're looking at. Why didn't the software confirm? The software did confirm - it merely asked if they wanted to send the message out. Why didn't he click no? Because the software didn't tell him what link he clicked, just if he was sure. (E.g., you close an app with a dozen documents open, and all you get is "Save file?" instead of it actually telling you what file to save).
Well, there's something you need to fix - the UI sucks and it's really only an accidental mis-click away from saying the president is dead to missiles have been launched.
So the UI has two problems with it - a huge nasty list of unsorted messages that really should be put in order somehow. And perhaps a big ass button that selects test messages from actual messages. And a confirmation dialog that actually confirms what you are going to send. Perhaps if it was a real message, it would ask first "The message you are sending is not a test message. Click OK to continue and have your supervisor access his console to do same" as well as "Send the non-test message 'Missiles are incoming'?"
Firing someone over mistakes doesn't ensure mistakes don't happen (because the person who learned from it will no longer be present). It instills a culture of fear - that if they click the wrong link, they can get fired. So what would take a few minutes now takes 10 people and an hour because the person who is to send the message has to check multiple times they're clicking the right thing. And the underlying cause won't get fixed, leading to more errors in the future
And imagine if (heaven forbid) a real event happens. You have 5 minutes before missiles hits. Do you want 4 of them to be wasted because the person at the desk responsible for sending it to triple check that yes, that's really the intent because if oh my god if there aren't any missiles I'm going to get fired?
It's why no one was fired for Amazon AWS going down last year, or when GitHub suffered a massive meltdown - errors were made, but the root cause turned out to be an opportunity for human error to do bad things accidentally.
Far too often the question asked is "Who" as if firing that person to make a point will fix the problem. It is the dominant question if you want to assign blame and move on, and it is politically popular among the people who are looking for someone to hang. But it turns out doing so doesn't fix underlying structural issues, it just covers it up.
People will just not use the site ads on offer. The video will have an ad spoken about by the presenter as part of their video. The ad can find the content and work direct with the content creator. No need to have site ads and be under the control of new SJW site rules.
People already do that on YouTube - you get lots of videos sponsored by various companies.
The problem is,
1) You need to be an established creator with lots of views, or ad companies won't talk to you 2) It irritates viewers because they are now forced to sit through ads, and YouTube stats have shown the forced ad is the worst regarded ad. Granted, people can skip through them by advancing the play bar so it's a bit better. 3) Your only statistic is views. You don't know how many people actually sat through the ad versus simply skipped ahead. No analytics means a severe restriction on price. 4) Once your video is up, you cannot edit it without starting from scratch. So if your advertiser drops you, you can't edit the advertiser out of your previous videos and add in a new advertiser. Your only option is to upload a new video replacing the old video and start from 0.
So at best, all you have is an upfront per-episode payment from an advertiser, because they know their ad will be up for all eternity and they really have no statistics about it other than views, so they may have to scale their payment so you don't upload crap that gets no views.
Come on, you know that's what they'll end up doing, don't you? Conservative, right-of-center, Euroskeptic or Muslim-critical videos will be deemed "unworthy of monetizing."
It's not Google, it's the advertisers.
Before the adpocalypse happened, YouTube slapped any ad on top of any video and things were good.
The problem is that people started realizing (because the election of a certain president seems to have given legitimacy to viewpoints that at best were controversial, at worse, something generally disallowed (we fought wars against Nazism, for example)) that certain videos were being sponsored with ads. Advertisers got very worried - they're a thin-skinned bunch and it appears that if their ad shows up one of these videos, it's a clear indication that the company supports those viewpoints. So advertisers left in droves - including many big ones.
(Yes, it doesn't help that many of those people think if some webhost or other doesn't take down their content, it's because they support or endorse the content. There's a difference from "I offer service equally (and blindly)" to "I support you". Cloudflare was one of these companies - it only pulled support for DailyStorm after the site promoted that since Cloudflare still provides them service, they agree and support them. No, Cloudflare was trying to be fair and equitable by not caring about the sites that use their services, not that they explicitly endorse or support those sites).
YouTube cracked down harder when again, advertisers discovered their ads were placed on videos that advertisers didn't want, and again, more advertisers pulled out.
So YouTube was forced to demonetize videos that basically could not find ad sponsorship. This is preferable to the alternative, which is YouTube banning or deleting those videos. But it also means people whose content consists of potentially controversial topics are unlikely to ever get monetized because YouTube has no advertiser lined up to support that content. That's why gun videos are hard to get monetized - with all the mass shootings that happen, no "regular" advertiser wants to be seen supporting anything that leads to another Las Vegas or similar incident.
Think of advertisers as people who need trigger warnings, and if your video is likely to cause any of those warnings, it likely won't be monetized. YouTube monetization only happens because of advertisers.
Which is the greater danger, allowing web access in the clear (note that this does not preclude allowing secured access as well) or creating a single point of failure called "Let's Encrypt" such that if it does fail then suddenly the entire world has to start paying money for certificates or finds their sites no longer work properly?
Not only that, but with Let's Encrypt issuing out certificates so sites can phish, it seems like a good way to avoid all the Paypal and other phishing is to block the Let's Encrypt certificate. (they issued like 14,000 phishing certificates)
Of course, we don't do this because Let's Encrypt is sponsored "by the good guys" (Mozilla, EFF, etc). But if it was some other CA, we'd be blocking them ASAP.
It's only a matter of time before they issue a new wave of certificates to phishing and other scammy sites. Not sure how long until people DO start manually blocking Let's Encrypt to get rid of a bunch of problem sites.
The overhead for SSL is not the encryption. Not on a modern CPU it isn't. Any overhead is due to the extra communication steps to set up the connection. But HTTP 1.1 will do a single handshake and reuse the connection.
No, the overhead in SSL has been the management. Especially inside a LAN context where you have to add your own to the trusted root, and maintain it all everywhere and even then you probably run into an odd device or two that won't allow you to install a certificate.
This is the truth, the parent AC doesn't know his gaming history. The Atari 2600 had no DRM and had huge problems with other companies making games for the platform, something that Atari had never anticipated. Nintendo's president believed at the time that this is what killed them. Not lack of royalties, but a flood of low quality games that Atari had no control over.
Whether this is true or not is debatable - they were a little too firmly dedicated to the 2600 and compromised subsequent platforms in its favor
Basically, what happened was EA and Activision happened, formed by a bunch of disgruntled Atari programmers. Basically Atari management refused to let the programmers put their names on the games and get some credit, so they left and formed EA and Activision. Since they were ex-Atari, they had all the inside knowledge on how it worked, so they started making their own games for it.
And make a ton of games they did - they kept cranking it out, because everyone wanted in on video games, so it was the best of times - crank out stuff. But then people came to the realization that most of what they had was... crap - churned out to make a quick buck because everyone was buying up games by the dozen - retailers were ordering hundreds of copies per store, etc. It was a boom time.
Then people realized most of it was crap and shovelware and stopped buying games. Retailers were stuck with thousands of cartridges and returned them in droves. Even worse, retailers were not buying games. Now this did not happen overnight, it basically took a couple of years where the video game industry declined. It became so bad, "video games" were a banned word at many retailers.
And this is where Nintendo comes in. They didn't call their system a video game system, they called it a toy, not to be sold in the now-banned video game section of the store, but where all the toys were. Problem number one - toy sections are girls, or boys. You can guess where Nintendo went, and potentially where we have such a gender imbalance in gaming today. (Check the ads - Atari ads always showed a relatively balanced family - mom, dad, son, daughter, playing their game system. Nintendo, though, showed only boys. No girls, no adults (it was a toy)).
Anyhow, the other reason for it is obvious - few people care about homebrew games on switch, everyone wants pirated games. So cracking the Switch really is for everyone to not pay for games ever again. (And ironically, this time around, it wouldn't be Nintendo strangling 3rd party developers). Face it, that's the real truth behind all the hacking.
If they crap out at 100,000 packets your router is indeed shitty, that's what 100Mb/s networks peak out at.
Most routers are software routed. If you're lucky, they run some form of Linux to do the firewalling, NAT, bridging (to WiFi), and everything else those routers are tasked to do (media serving, etc).
Modern enterprise routers do hardware routing and try to avoid process packet switching as much as possible because it's trivial to overload the main CPU this way.
So you have a relatively wimpy processor that has to handle packets coming in on multiple interfaces and handle all the processings.
And let's not forget most cheapo routers have barely any RAM to work with - just enough for the OS and barely enough for Linux tables and other stuff.
Hell, most cheapo home routers crash because the system simply runs out of RAM. Which could happen for many reasons - perhaps the wifi driver couldn't allocate a chunk to bridge the packet from WiFi to ethernet, LInux runs out of memory doing whatever, etc.
So just because in theory it's possible, you have to remember home routers are process switching and routing and serving up tons of other things at the same time.
Slashdot Mirror
And that's to just use their services. To actually transfer via Bitcoin incurs Bitcoin's transaction fee, which is variable and depends on how long you're willing to wait. This fee occurs every time you transfer anything via Bitcoin and in independent of the exchange or whatever.
The fee goes straight to the miners as a reward for locking that transaction into the blockchain. Remember, miners do more than "make" bitcoin - in the early days they created actual bitcoin itself, but these days, with the reduction in production rate, miners make money locking the transaction into the blockchain.
So the more you pay, the more the miners get, so naturally when looking at all the pending transactions, they will order them from pays the most to least and then hash those into the blockchain. Pay among the top and it's confirmed within minutes. Be completely cheap and it'll take weeks.
At the commuter train/LRT level, there are many systems that are completely automated - no drivers at all. Granted, the switching isn't too complex, but they can be commanded to switch tracks and manage themselves with supervision done at a central control station.
Operating at the full size train level is only a slightly more complex problem, mostly because now the tracks are owned by many people and you really need to get them all to install a common communication an d signalling system. (Right now trains can be monitored remotely).
The problem with Amtrak is also been solved - the technology has long existed and it's nowadays called Positive Control. The train will periodically beep and the driver has to hit a button. If the driver fails to hit the button, the train is brought to a stop automatically. This helps catch distracted and sleepy drivers.
Even speed limit enforcement has been automated away - if the train is coming too fast for the speed limit, the train automatically slows down. The unions generally hate this as it reduces the driver to a monkey. OTOH, the safety record of these systems is quite stellar, and most rail lines only experience it once before the system is rolled out on all the rolling stock.
Though, the Shinkasen drivers of Japan, where they not only have these systems in place but also very rigorous protocols and how they act (they lift their arms up as if to salute, then point at the control they are going to adjust, then adjust it - it looks like part of a military march) don't seem to be monkeying around, but seem to be very professional about it. Even though there is speed limit enforcement, they still can control the speed of the train, and still can get the train to be within a minute of the stated arrival time, even if weather conditions make it so the train has to slow down (it snows in Japan, too, and the trains have to operate in it at reduced speeds).
Well, firefox makes you edit text files to adjust the UI, so if you're a greybeard that should be a pleasant blast from the past. These days though, editing text files to adjust the UI seems very antiquated and a step into 30 years ago when we did it only because there was not enough CPU power or memory to graphically edit UIs so it was a cycle of edit, relaunch, test, quit, ...
I get it', it's more secure when websites can't redo your browser UI on you, but still, having to hunt for third party websites that have the magic incantations to adjust the UI and lack of an editor really put you back.
So... why not take this idea then? That way we can tell facebook to shove it and not pay him, thus blocking his news articles from showing up.
To me, it seems like a great idea, Murdoch requires payment, facebook doesn't pay, Murdoch blocks facebook, everyone benefits.
Or it can do a combination of signal strength and speed and try to figure it out that way. After all an 802.11ac signal is worthless if it's weak but you have a stronger N signal nearby or things like that.
A long long time ago, I had a WiFi card that could connect to an 802.11g AP on the 10th floor from the 3rd through a concrete-and-steel building. This was annoying because I had an 802.11b AP nearby that would because of its proximity and greater signal strength be faster than the 802.11g one would because the distance.
And yes, it actually did work, but wasn't that fast.
Organic parts can be made to last longer. In fact, a boat manufacturer replaced their shaft seals with new ones that lasted 2 to 3 times longer than the existing rubber ones, in testing. Good change, right? Customers can buy seals that last twice as long as the old ones (thus halving the number of tricky seal changes, too, saving money all around since you wouldn't need to dry dock the boat as often, labor, etc).
Problem was, it was soy based I believe, so critters started getting to them and after the summer was over, there were a lot of sunken boats.
Moral is, the soy stuff can genuinely be better from all objective points - it can last longer in the varied conditions a car goes through without the wiring getting brittle and cracking and causing short circuits. It just had an unexpected side effect that critters love to eat it. Even that can be taken care of - perhaps thicker insulation with the application of a bitterant to basically make the wires not tasty at all. (And since bitter is a flavor associated with poisons, it also appeals to the critter's survival instincts that what they're eating is probably bad for them).
And fundamental problems are still fundamental problems. The reason practically every processor has the same issues is because the same optimizations we used to make processors faster had the same fundamental design error.
I mean, either someone designed the core branch predictor block and everyone worldwide copied it for every processor, or everyone implemented it differently, yet it has the same Spectre flaw, implying that the flaw is inherent in the way branch predictors work.
The only way you can guarantee the designs are error free are to abandon everything that makes modern processors fast - OOO, speculation, branch prediction, and plenty more, including potentially pipelining (the fundamental technology everyone is trying to speed up by avoiding pipeline stalls). Go back to the old fetch-and-execute cycle and where memory operands are fully decoded and retrieved prior to even considering fetching the next instruction.
Everyone will hate it, because now your 4GHz processor will be as fast as a 500MHz one.
Some problems are solvable by telecommuting, some problems aren't.
Stuff that involves hardware (which hey, Apple does a lot of) really cannot be done by telecommuting until you get full telepresence robots (and even then, some things really require you to be there - engineer's knack, for example). Expensive test equipment may be utilized, or even just general tools like oscilloscopes. And then there's physical hardware that requires transportation, Apple's already secretive enough that new hardware has to be locked up and multiple layered security (so fat chance bringing it home).
There's also a lot of value in bringing the team together - collaboration can be much faster. Perhaps you come up with a cool idea, but then instead of just walking down the hall and presenting it, you have to type it up in an email or worse yet, the company IRC-like chat room (slack, whatever) and everyone misunderstands you so you spend another hour re-describing it. And heaven forbid you need to draw something...
If you want cutting edge, yes.
If you can stay back a few nodes, not so bad, a few million bucks is needed (masks are expensive at about $100K/each for the older processes), so perhaps a regular 10 metal chip requires a couple million bucks.
And while most of it is autorouted and autoplaced, you still want to hand edit the designs. Remember the reason we're at 10 metal is because for most general random logic, the limiting factor is wiring. The vast majority of transistors in any design is used in memory - caches and the like. The density falls off rapidly when you get to general logic blocks like ALUs and registers and other random logic parts. And because of the low density, you edit the layouts to sprinkle in lots of new transistors that do nothing. This is because when you're fixing bugs, you're going to need extra logic parts, and having them already in the silicon itself means only changing a few masks instead of an entire mask set.
That's why there are steppings - A0 to A1 to A2 are metal-only changes and A0 meant a full mask set, but A1 and A2 only had marginal changes so perhaps 5 masks changed (so instead of spending $2M per pop, you spent an additional $1M). But go from A2 to B0 means you changed everything again, so a complete redesign. The reason for this is you ran out of spares, or more likely, the fix involves speed-sensitive paths and you just cannot route the spares in a way to keep the speed up.
Anyhow, I still don't see how an open design would avoid something like Meltdown or Spectre, because those vulnerabilities came about because of novel ways of using the weaknesses of things we did to make processors so much faster, and given the entrenchment of the designs, everyone implemented the same issues, so it's really an issue going all the way back.
It also means that they are under the control of Congress and abide by all the laws that it passes. But even bigger, they have no representation at all. US territories follow US laws and all that, but have no representation in the House or the Senate, so they don't have a say in any laws that get enacted.
That's why they aren't taxed - perhaps the old refrain "no taxation without representation" sound familiar? It also means those territories are mostly forgettable - they're required to follow US laws even though it can be bad for their own interests (because no one in the House or Senate cares, or knows about, those issues).
That's because the most serious vulnerabilities affect virtualized systems more, and given the popularity of the cloud, they are extremely affected because there is no control over what software is run.
These attacks may seem theoretical, and for a desktop user, they mostly are. Even server users with dedicated machines are generally safe as they own the entire machine and thus control the software on it. Virtualized hosts, not so much, and this ranges from simple VPS hosting companies, to cloud service companies like Amazon and Google. In this case, accessing not-your-memory can be beneficial, like trying to find the encryption keys used by someone sharing your machine. Cloud providers are especially scared because during high peak retail times, encryption keys may be spun to many instances to handle the load, and crafty attackers may simply spin up instances to extract those keys.
Even worse, IBM has only been releasing half-patches, the complete fix is due in mid-February. With the speed that Intel and AMD have released fixes, and OS vendors have too, IBM seems slow off the mark.
And I won't really call it a bug, but an optimization people did that no one realized would open up serious security vulnerabilities. It's a cheat because you're not doing all the proper checks when you should, because if you did, you'd slow down and stall the pipeline. Instead, you begin speculative execution, kick off tasks to fetch operands and other metadata (including security and maps) and hope by the time the instruction needs the information, it's there.. It's why we moved from fetch-and-execute cycles to full OOO execution - the speed up is immense. It's also why everyone did it, and why we're all seeing the result now.
Of course, doing it right avoided the issues, but then your processor ends up slower and thus, less competitive.
And I see them with the huge phones in their hands (hands, plural, because they can't hold it in one hand - too big), so that isn't accurate at all. It's interesting to watch them wrangle a 6" phablet in two hands, which obviously was bought because they needed the screen size than practicality.
Anyhow, apparently the real reason is the iPhone X sales in China is disappointing. Apparently the screen is too small. If your phone requires at least two hands to hold it, it's the right size. If you can hold it in one hand, it's too small. If you require a third hand, bingo, you got the best phone out there.
Apparently size over practicality. Even if it requires two hands to hold it so you can make a phone call (using voice assist, because with two hands holding the phone, there are no spare digits that can reach the numbers).
And yes, they flash the phone everywhere. It is a status symbol, and the iPhone X unfortunately is not "bling-y" enough.
Cars are heavily recycled, to the tun of about 80-90% of a car is recycled back to raw materials. The rest of it is dumped. Car manufacturers have been trying to increase the amount that you can recycle, or reduce the amount that stays stable when dumped (i.e, most plastics). A lot of stuff like fiber reinforced materials cannot be recycled - the fibers make it heavily impure and a pain to remove. It's why recycling fiberglass turns it a rough powder that is then injection molded.
Burning plastic isn't actually all that great - it generally releases a lot of nasty chemicals and solvents in the air that are not good for the environment. It's also hard, because a lot of car wiring has to withstand high temperatures, and often are fireproof, which makes burning them in incinerators even nastier. That's why they switched to more organic materials - burning generally isn't as bad (but it's still horrible - as any crematorium will know, burning dead bodies releases nasty compounds even though we're "natural").
Biodegratable ones can be degraded in special environments that rapidly decompose the materials - and can be designed to last basically forever under normal environments, but put them in special composters they will be gone in weeks, and generally release byproducts that can be captured and used (e.g., methane)
Canada does something similar to the US too - we call them "omnibus" bills that seem to do a million unrelated things at once alongside the budget.
Now, that doesn't mean we suffer from the US problem, as in the Westminster system, voters do not elect the prime minister - the prime minister is chosen - first among the party that has the majority, but if that fails then the remaining parties choose a prime minister. So the prime minister and the party or coalition of parties is the same, and they just ram the bill through. In a majority situation, this is simple - you put it up and vote and it passes. In a minority situation, it's a lot dicier.
Thing is, it either passes, or it doesn't. Budget bills are "confidence" matters - if it doesn't pass, it's a general resounding vote that the government is now ineffective and needs to be replaced - either by the opposition (if they can prove they can lead) or another election.
Yes, it's also why minority governments are generally unstable - confidence matters mean everyone else can vote no and force government to fall.
Even so (face it, BMW could just toss you an Apple USB to Lightning cable with the car. It ain't hard), the license fee is really just a once time fee for that item. You go through Apple,. get your stuff approved (ti can be trying, because Apple does do a lot of testing, including security testing) and get the license to use the logo. It's a one time fee and payment, not an ongoing thing (otherwise manufacturers of other things like speakers and headphones would be on the hook for payments as well).
And seriously, $300 to use CarPlay in a $60K+ car? Really, people are so cheap to not add that?
That's because everyone (including Adobe) uses the exact same RAW library. It happens to be open-source, too.
DCRaw is a raw image handler for basically every camera out there. It's such an impressive piece of code everyone uses it,
You know how freaking hard that is?
Usually someone comes in and wants to dabble with this and that to "put their mark on it" and that leads to all sorts of crap going on and changes for the worse.
To h ave someone able to put all that aside and let things run is quite unnatural. SO yes, the fact it isn't all screwed up by now is amazing. The fact it's sold twice as much since acquisition is amazing.
Also, how many times have you heard of X acquiring Y and Y getting completely messed up? It happens almost always despite promises to the contrary. Thus, if Minecraft is even more successful under Microsoft, it took a lot of restraint in not screwing it up.
And for someone to manage a game studio, I suppose that's an excellent trait - if Microsoft were to acquire a bunch of studios, this guy would likely have a very light touch on things so it likely wouldn't be Microsoft that screwed up.
Because the 5S is not affected by this. Only the 6 and above have the power management feature and can scale back the processor speed to lower battery consumption and are thus affected by bad batteries.
But really, cheap 5S battery replacements are available quite easily - all those mobile repair shops can trivially change the battery for about $30 or so. And I think iFixit and other online stores can sell you DIY kits for around the same price.
You're falling into the "we must fire someone for accountability" trap.
That leads to basically incompetents running your ship - if everyone is deathly afraid of losing their job for making a mistake, you end up with a corporate culture of timidity, cover your ass and hiding mistakes.
The modern method is not to fire the person who pushed the button, but to find out the true reason. This is often' called "The Five Whys" because it literally asks Why over and over again.
Like in this case, given what we know.
Why was a missile alert called? Because someone clicked the link to send it.
Why did they click that link? Because they clicked the wrong link - they meant to click the one that produced a test message instead.
Why did they click the wrong link? Because the links were presented as an unsorted list, with the test links appearing on some events ahead of the real link, and sometimes afterwards.
Why did they click the wrong link? Because when you're looking at a huge list of unsorted links, you tend to focus on the one that matches what you're wanting even though it may not be exactly what you're looking at.
Why didn't the software confirm? The software did confirm - it merely asked if they wanted to send the message out.
Why didn't he click no? Because the software didn't tell him what link he clicked, just if he was sure. (E.g., you close an app with a dozen documents open, and all you get is "Save file?" instead of it actually telling you what file to save).
Well, there's something you need to fix - the UI sucks and it's really only an accidental mis-click away from saying the president is dead to missiles have been launched.
So the UI has two problems with it - a huge nasty list of unsorted messages that really should be put in order somehow. And perhaps a big ass button that selects test messages from actual messages. And a confirmation dialog that actually confirms what you are going to send. Perhaps if it was a real message, it would ask first "The message you are sending is not a test message. Click OK to continue and have your supervisor access his console to do same" as well as "Send the non-test message 'Missiles are incoming'?"
Firing someone over mistakes doesn't ensure mistakes don't happen (because the person who learned from it will no longer be present). It instills a culture of fear - that if they click the wrong link, they can get fired. So what would take a few minutes now takes 10 people and an hour because the person who is to send the message has to check multiple times they're clicking the right thing. And the underlying cause won't get fixed, leading to more errors in the future
And imagine if (heaven forbid) a real event happens. You have 5 minutes before missiles hits. Do you want 4 of them to be wasted because the person at the desk responsible for sending it to triple check that yes, that's really the intent because if oh my god if there aren't any missiles I'm going to get fired?
It's why no one was fired for Amazon AWS going down last year, or when GitHub suffered a massive meltdown - errors were made, but the root cause turned out to be an opportunity for human error to do bad things accidentally.
https://en.wikipedia.org/wiki/...
Far too often the question asked is "Who" as if firing that person to make a point will fix the problem. It is the dominant question if you want to assign blame and move on, and it is politically popular among the people who are looking for someone to hang. But it turns out doing so doesn't fix underlying structural issues, it just covers it up.
People already do that on YouTube - you get lots of videos sponsored by various companies.
The problem is,
1) You need to be an established creator with lots of views, or ad companies won't talk to you
2) It irritates viewers because they are now forced to sit through ads, and YouTube stats have shown the forced ad is the worst regarded ad. Granted, people can skip through them by advancing the play bar so it's a bit better.
3) Your only statistic is views. You don't know how many people actually sat through the ad versus simply skipped ahead. No analytics means a severe restriction on price.
4) Once your video is up, you cannot edit it without starting from scratch. So if your advertiser drops you, you can't edit the advertiser out of your previous videos and add in a new advertiser. Your only option is to upload a new video replacing the old video and start from 0.
So at best, all you have is an upfront per-episode payment from an advertiser, because they know their ad will be up for all eternity and they really have no statistics about it other than views, so they may have to scale their payment so you don't upload crap that gets no views.
It's not Google, it's the advertisers.
Before the adpocalypse happened, YouTube slapped any ad on top of any video and things were good.
The problem is that people started realizing (because the election of a certain president seems to have given legitimacy to viewpoints that at best were controversial, at worse, something generally disallowed (we fought wars against Nazism, for example)) that certain videos were being sponsored with ads. Advertisers got very worried - they're a thin-skinned bunch and it appears that if their ad shows up one of these videos, it's a clear indication that the company supports those viewpoints. So advertisers left in droves - including many big ones.
(Yes, it doesn't help that many of those people think if some webhost or other doesn't take down their content, it's because they support or endorse the content. There's a difference from "I offer service equally (and blindly)" to "I support you". Cloudflare was one of these companies - it only pulled support for DailyStorm after the site promoted that since Cloudflare still provides them service, they agree and support them. No, Cloudflare was trying to be fair and equitable by not caring about the sites that use their services, not that they explicitly endorse or support those sites).
YouTube cracked down harder when again, advertisers discovered their ads were placed on videos that advertisers didn't want, and again, more advertisers pulled out.
So YouTube was forced to demonetize videos that basically could not find ad sponsorship. This is preferable to the alternative, which is YouTube banning or deleting those videos. But it also means people whose content consists of potentially controversial topics are unlikely to ever get monetized because YouTube has no advertiser lined up to support that content. That's why gun videos are hard to get monetized - with all the mass shootings that happen, no "regular" advertiser wants to be seen supporting anything that leads to another Las Vegas or similar incident.
Think of advertisers as people who need trigger warnings, and if your video is likely to cause any of those warnings, it likely won't be monetized. YouTube monetization only happens because of advertisers.
Not only that, but with Let's Encrypt issuing out certificates so sites can phish, it seems like a good way to avoid all the Paypal and other phishing is to block the Let's Encrypt certificate. (they issued like 14,000 phishing certificates)
Of course, we don't do this because Let's Encrypt is sponsored "by the good guys" (Mozilla, EFF, etc). But if it was some other CA, we'd be blocking them ASAP.
It's only a matter of time before they issue a new wave of certificates to phishing and other scammy sites. Not sure how long until people DO start manually blocking Let's Encrypt to get rid of a bunch of problem sites.
No, the overhead in SSL has been the management. Especially inside a LAN context where you have to add your own to the trusted root, and maintain it all everywhere and even then you probably run into an odd device or two that won't allow you to install a certificate.
Basically, what happened was EA and Activision happened, formed by a bunch of disgruntled Atari programmers. Basically Atari management refused to let the programmers put their names on the games and get some credit, so they left and formed EA and Activision. Since they were ex-Atari, they had all the inside knowledge on how it worked, so they started making their own games for it.
And make a ton of games they did - they kept cranking it out, because everyone wanted in on video games, so it was the best of times - crank out stuff. But then people came to the realization that most of what they had was... crap - churned out to make a quick buck because everyone was buying up games by the dozen - retailers were ordering hundreds of copies per store, etc. It was a boom time.
Then people realized most of it was crap and shovelware and stopped buying games. Retailers were stuck with thousands of cartridges and returned them in droves. Even worse, retailers were not buying games. Now this did not happen overnight, it basically took a couple of years where the video game industry declined. It became so bad, "video games" were a banned word at many retailers.
And this is where Nintendo comes in. They didn't call their system a video game system, they called it a toy, not to be sold in the now-banned video game section of the store, but where all the toys were. Problem number one - toy sections are girls, or boys. You can guess where Nintendo went, and potentially where we have such a gender imbalance in gaming today. (Check the ads - Atari ads always showed a relatively balanced family - mom, dad, son, daughter, playing their game system. Nintendo, though, showed only boys. No girls, no adults (it was a toy)).
Anyhow, the other reason for it is obvious - few people care about homebrew games on switch, everyone wants pirated games. So cracking the Switch really is for everyone to not pay for games ever again. (And ironically, this time around, it wouldn't be Nintendo strangling 3rd party developers). Face it, that's the real truth behind all the hacking.
Most routers are software routed. If you're lucky, they run some form of Linux to do the firewalling, NAT, bridging (to WiFi), and everything else those routers are tasked to do (media serving, etc).
Modern enterprise routers do hardware routing and try to avoid process packet switching as much as possible because it's trivial to overload the main CPU this way.
So you have a relatively wimpy processor that has to handle packets coming in on multiple interfaces and handle all the processings.
And let's not forget most cheapo routers have barely any RAM to work with - just enough for the OS and barely enough for Linux tables and other stuff.
Hell, most cheapo home routers crash because the system simply runs out of RAM. Which could happen for many reasons - perhaps the wifi driver couldn't allocate a chunk to bridge the packet from WiFi to ethernet, LInux runs out of memory doing whatever, etc.
So just because in theory it's possible, you have to remember home routers are process switching and routing and serving up tons of other things at the same time.