A large part of the Ubuntu users doesn't pay Canonical a cent. And that is probably a major factor in its popularity.
A better comparison would be Debian vs Ubuntu installs on servers. Although the different release cycles may be relevant there: Debian has long and unpredictable cycles, while Ubuntu has a release every 6 months. On the other hand, Debian releases deserve the predicate "stable" while Ubuntu releases have had some rough edges.
Advertisers are sensitive to media outrage, manufactured or not. And YouTube is sensitive to advertisers, since those are their major source of income.
If YouTube actually wanted to do something about pedophiles abusing their platform via comments, they would either identify and ban the accounts making those comments or, if that is too much work or ineffective, disable comments altogether on content featuring children. Instead, they are demonetizing videos, in other words: not showing ads on them. They're pretending to protect children, but all that move protects is the advertisers.
This is not a case of the exchange getting robbed though, it is the currency itself that got attacked. The problem with proof-of-work is that if someone manages to control over half the mining power, they get to decide which transactions happen and which do not. The theory was that there would never be a single party in that position, but apparently that theory doesn't apply to the less popular coins.
It was on the news here: the police announced it at this moment on purpose, because several people getting arrested recently made other criminals suspect someone in their mids was leaking to the police and they were planning violent actions against them.
I don't know why Red Hat is deprecating it, but my experience with btrfs was so poor that I've dropped it.
The first problem I ran into is that free space is not automatically reclaimed. I'd get processes unable to create files when "df" reported 75 GB free. It turns out you have to regularly run a cleanup command that moves data around so the free space is combined and usable again. So not only do you have to defragment your disk in 2018, if you don't, it doesn't just slow down but becomes unusable for everyday operations.
I figured I'd set up a cron job to defragment the disk every week. But then the problem is that while the defragmentation is running, the PC becomes so unresponsive that it's barely usable. This is a problem with Linux in general, that heavy I/O will make the system unresponsive. But usually it doesn't happen unless you're using a lot of swap space.
The final drop was when the system became so unresponsive that it wouldn't even react to the command I issued to abort the defragmentation in progress. I wanted to actually use the PC, so I did an emergency unmount and reboot with SysRq. On the reboot the brtfs file system was corrupted and wouldn't mount. Eventually I got it to mount read-only and managed to get most of the data off, but I never got it to mount read-write anymore.
Apart from the inconvenience, what really scared me off btrfs was the state of the recovery tools. The official pages firmly tell you to only run fsck as a last resort. There are a whole lot of other tools that can each recover from specific problems, but unless you know how btrfs works internally and can figure out from a vague error message which part of it is broken, all you can do is try one tool after another, wait a long time as it scans your entire partition and see nothing improve.
So all in all, brtfs has some interesting features, but despite being around for quite some years, it's nowhere near a mature file system.
Sorry, number, not digit. In any case, the patch level is the last number in the series and since 3.0 the version number was shortened from 4 numbers to 3 numbers.
A lack of driver source code could be a problem upgrading the kernel from for example 4.4 to 4.9, since internal kernel interfaces may have changed. However, upgrading the kernel from 4.4.7 to 4.4.123 will only include bug fixes with no interface changes and the vast majority of consumer electronics manufacturers aren't doing those upgrades either.
Different things are called "AI" these days. Thinking machines are still very primitive, there hasn't been a breakthrough there. What is starting to show results are statistical models on large data sets, that are constructed at least partially automatically.
But these depend on having a large data set, which means that whatever company has the most data will likely have the best model, which in turn attracts more users, allowing them to collect more data. So it's hard for a newcomer to gain traction.
In any case, the point I think the authors were making (the wrong article is linked so I can't easily verify it) is that AI will make it harder for developing nations to improve their economies, rather than making their current economies worse. So it's more a case of AI undermining evolutionary development than it being a revolution in itself.
You mean Angry Birds? Flappy Bird is a game that I think would suffer from high latency. The importance of latency differs a lot per game: I once played Final Fantasy using a TV capture card to turn my PC into a makeshift monitor and I didn't notice the higher latency. But as soon as I tried Frequency (a rhythm game), I was behind on all the beats, while usually I get at least half of them right.
In any case, I agree that there will be considerable extra latency, but you might underestimate how much latency people are willing to cope with, in certain games.
Note that when it comes to online games where multiple avatars walk around in the same world, input events already have to travel to the server and back. So moving the rendering will make some of the latency hiding techniques weaker, but the worst case latency (when prediction is wrong) is still the same.
This is Take Two, who make hundreds of millions a year with GTA Online. If you want games without microtransaction bullshit, you're not the kind of customer they're looking for.
Most big console titles run at 30 fps, so they don't have super low latency anyway. Having a data center within a couple of thousand kilometers might be good enough. Also latency hiding tricks like input prediction could be used.
Their store calls it "buy" and the price is set at a level far above the rent price: everything suggests you can buy movies. Don't blame the victim for not reading the fine print on what seems like a very ordinary consumer purchase.
How do unchecked exceptions fix the things you mentioned? Sure, it's a pain to do proper error handling, but I have never seen any programming language in which error handling is effortless. At least with checked exceptions, the compiler will alert you to unhandled situations.
Deciding that you won't bother to do proper error handling is fine if you're working on some tool that you only use yourself. But if your code is going to be part of a large complex system, you're trading a bit of development speed now for a lot of pain later. "One in a million" events can happen several times a day, especially if the way your code is used changes over time.
Adult content has a high percentage of disputes (probably from guys who's wives/girlfriends notice the charge).
That doesn't apply to Patreon though, since Patreon charges just show up as "Patreon" on the credit card statement; the specific creators that received the money are listed in an e-mail instead.
The gaming authority said that under current laws, if the items won have a monetary value, a loot box is gambling and gambling is tightly regulated in the Netherlands. They also said that items that cannot be transferred to other people don't have an monetary value, so even though they saw a lot of similarities between gambling and loot boxes containing non-transferrable items, those loot boxes are not in violation of the law. As a result, Steam blocked the selling of in-game items, while the loot boxes remain.
Generics exist because of static typing; there is no need for them as a language feature when you have dynamic typing.
Python has decorators which can do more or less the same things that annotations can.
Static typing and execution speed are the main differences that matter, in my opinion. Static typing can be an advantage or disadvantage, depending on what you're building. Recent versions of Python have optional type declarations that can be used for static code checking, but I haven't yet tried those in practice, so I don't know how useful they are in bridging the gap between static and dynamic typing.
Execution speed is usually much better (like a factor 4 to 10) in Java, although if you use a lot of library calls or I/O, the interpreter is not the deciding factor in the speed of a Python program and then it can run just as fast as Java.
Many games already have matchmaking algorithms that try to pit players with similar skill level against each other. It probably doesn't matter whether a skill difference is caused by practice or by using a superior input method.
When bribery is illegal, there is risk involved for both the one paying and the one receiving the bribe. That risk will raise the price. Also hiding the transaction will take effort, which adds to the cost. So even though you can't eliminate all bribery, making it illegal will likely reduce it, if only because the same favors are going to be more expensive.
They mean it's not gambling by the legal definition of gambling in various countries. It does feed upon the same psychological mechanisms that gambling does, but they don't care about ethics when there is so much money to be made.
It depends on what you call the "problem". If the problem is being considered gambling under current law, then making the items non-transferable would solve it. If the problem is an addiction risk for players, then there are several more properties mentioned in the report, like illusion of skill, near misses and the opening being accompanied by visual and sound effects.
Here is the full report (English version, PDF), for anyone interested in the details.
Slashdot Mirror
A large part of the Ubuntu users doesn't pay Canonical a cent. And that is probably a major factor in its popularity.
A better comparison would be Debian vs Ubuntu installs on servers. Although the different release cycles may be relevant there: Debian has long and unpredictable cycles, while Ubuntu has a release every 6 months. On the other hand, Debian releases deserve the predicate "stable" while Ubuntu releases have had some rough edges.
Advertisers are sensitive to media outrage, manufactured or not. And YouTube is sensitive to advertisers, since those are their major source of income.
If YouTube actually wanted to do something about pedophiles abusing their platform via comments, they would either identify and ban the accounts making those comments or, if that is too much work or ineffective, disable comments altogether on content featuring children. Instead, they are demonetizing videos, in other words: not showing ads on them. They're pretending to protect children, but all that move protects is the advertisers.
This is not a case of the exchange getting robbed though, it is the currency itself that got attacked. The problem with proof-of-work is that if someone manages to control over half the mining power, they get to decide which transactions happen and which do not. The theory was that there would never be a single party in that position, but apparently that theory doesn't apply to the less popular coins.
Maybe it's time to port Wine to Windows, so WIndows 10 users can run Win32 applications...
It was on the news here: the police announced it at this moment on purpose, because several people getting arrested recently made other criminals suspect someone in their mids was leaking to the police and they were planning violent actions against them.
I checked my old fstab and it only had the "noatime" option specified. Maybe that is the difference then?
Trim was set up as a cron job, but even if it wasn't, that would lead to poor free space management at the SSD level, not at the file system level.
I'm still not eager to go back due to the mess that the recovery tools are.
I don't know why Red Hat is deprecating it, but my experience with btrfs was so poor that I've dropped it.
The first problem I ran into is that free space is not automatically reclaimed. I'd get processes unable to create files when "df" reported 75 GB free. It turns out you have to regularly run a cleanup command that moves data around so the free space is combined and usable again. So not only do you have to defragment your disk in 2018, if you don't, it doesn't just slow down but becomes unusable for everyday operations.
I figured I'd set up a cron job to defragment the disk every week. But then the problem is that while the defragmentation is running, the PC becomes so unresponsive that it's barely usable. This is a problem with Linux in general, that heavy I/O will make the system unresponsive. But usually it doesn't happen unless you're using a lot of swap space.
The final drop was when the system became so unresponsive that it wouldn't even react to the command I issued to abort the defragmentation in progress. I wanted to actually use the PC, so I did an emergency unmount and reboot with SysRq. On the reboot the brtfs file system was corrupted and wouldn't mount. Eventually I got it to mount read-only and managed to get most of the data off, but I never got it to mount read-write anymore.
Apart from the inconvenience, what really scared me off btrfs was the state of the recovery tools. The official pages firmly tell you to only run fsck as a last resort. There are a whole lot of other tools that can each recover from specific problems, but unless you know how btrfs works internally and can figure out from a vague error message which part of it is broken, all you can do is try one tool after another, wait a long time as it scans your entire partition and see nothing improve.
So all in all, brtfs has some interesting features, but despite being around for quite some years, it's nowhere near a mature file system.
Sorry, number, not digit. In any case, the patch level is the last number in the series and since 3.0 the version number was shortened from 4 numbers to 3 numbers.
The third digit since 3.0 is comparable to the fourth digit before 3.0.
A lack of driver source code could be a problem upgrading the kernel from for example 4.4 to 4.9, since internal kernel interfaces may have changed. However, upgrading the kernel from 4.4.7 to 4.4.123 will only include bug fixes with no interface changes and the vast majority of consumer electronics manufacturers aren't doing those upgrades either.
I think this may be the actual article.
Different things are called "AI" these days. Thinking machines are still very primitive, there hasn't been a breakthrough there. What is starting to show results are statistical models on large data sets, that are constructed at least partially automatically.
But these depend on having a large data set, which means that whatever company has the most data will likely have the best model, which in turn attracts more users, allowing them to collect more data. So it's hard for a newcomer to gain traction.
In any case, the point I think the authors were making (the wrong article is linked so I can't easily verify it) is that AI will make it harder for developing nations to improve their economies, rather than making their current economies worse. So it's more a case of AI undermining evolutionary development than it being a revolution in itself.
You mean Angry Birds? Flappy Bird is a game that I think would suffer from high latency. The importance of latency differs a lot per game: I once played Final Fantasy using a TV capture card to turn my PC into a makeshift monitor and I didn't notice the higher latency. But as soon as I tried Frequency (a rhythm game), I was behind on all the beats, while usually I get at least half of them right.
In any case, I agree that there will be considerable extra latency, but you might underestimate how much latency people are willing to cope with, in certain games.
Note that when it comes to online games where multiple avatars walk around in the same world, input events already have to travel to the server and back. So moving the rendering will make some of the latency hiding techniques weaker, but the worst case latency (when prediction is wrong) is still the same.
This is Take Two, who make hundreds of millions a year with GTA Online. If you want games without microtransaction bullshit, you're not the kind of customer they're looking for.
Most big console titles run at 30 fps, so they don't have super low latency anyway. Having a data center within a couple of thousand kilometers might be good enough. Also latency hiding tricks like input prediction could be used.
Their store calls it "buy" and the price is set at a level far above the rent price: everything suggests you can buy movies. Don't blame the victim for not reading the fine print on what seems like a very ordinary consumer purchase.
From the way he handled North Korea and the EU, it will probably go like this:
How do unchecked exceptions fix the things you mentioned? Sure, it's a pain to do proper error handling, but I have never seen any programming language in which error handling is effortless. At least with checked exceptions, the compiler will alert you to unhandled situations.
Deciding that you won't bother to do proper error handling is fine if you're working on some tool that you only use yourself. But if your code is going to be part of a large complex system, you're trading a bit of development speed now for a lot of pain later. "One in a million" events can happen several times a day, especially if the way your code is used changes over time.
Adult content has a high percentage of disputes (probably from guys who's wives/girlfriends notice the charge).
That doesn't apply to Patreon though, since Patreon charges just show up as "Patreon" on the credit card statement; the specific creators that received the money are listed in an e-mail instead.
The gaming authority said that under current laws, if the items won have a monetary value, a loot box is gambling and gambling is tightly regulated in the Netherlands. They also said that items that cannot be transferred to other people don't have an monetary value, so even though they saw a lot of similarities between gambling and loot boxes containing non-transferrable items, those loot boxes are not in violation of the law. As a result, Steam blocked the selling of in-game items, while the loot boxes remain.
Generics exist because of static typing; there is no need for them as a language feature when you have dynamic typing.
Python has decorators which can do more or less the same things that annotations can.
Static typing and execution speed are the main differences that matter, in my opinion. Static typing can be an advantage or disadvantage, depending on what you're building. Recent versions of Python have optional type declarations that can be used for static code checking, but I haven't yet tried those in practice, so I don't know how useful they are in bridging the gap between static and dynamic typing.
Execution speed is usually much better (like a factor 4 to 10) in Java, although if you use a lot of library calls or I/O, the interpreter is not the deciding factor in the speed of a Python program and then it can run just as fast as Java.
Many games already have matchmaking algorithms that try to pit players with similar skill level against each other. It probably doesn't matter whether a skill difference is caused by practice or by using a superior input method.
When bribery is illegal, there is risk involved for both the one paying and the one receiving the bribe. That risk will raise the price. Also hiding the transaction will take effort, which adds to the cost. So even though you can't eliminate all bribery, making it illegal will likely reduce it, if only because the same favors are going to be more expensive.
They mean it's not gambling by the legal definition of gambling in various countries. It does feed upon the same psychological mechanisms that gambling does, but they don't care about ethics when there is so much money to be made.
It depends on what you call the "problem". If the problem is being considered gambling under current law, then making the items non-transferable would solve it. If the problem is an addiction risk for players, then there are several more properties mentioned in the report, like illusion of skill, near misses and the opening being accompanied by visual and sound effects.
Here is the full report (English version, PDF), for anyone interested in the details.