and PC ROMs no longer implement ASCII, but instead have native video codec support.
For what it's worth, the last new graphics cards to have hardware support for ASCII were released over 15 years ago. In modern cards, it's all done in firmware and with EFI the native interface is a framebuffer with text rendering layered on top. A typical GPU has hardware acceleration for a number of video CODECs, but no acceleration for text (though antialiased text rendering makes heavy use of the compositing engine).
That's nothing to do with demand though, it's entirely as a result of the different costs. Rendering fixed-width text is something that a modern CPU can do without the user noticing even if it takes an interrupt for every character and runs some emulation code in system management mode so there's no point devoting any transistors to it in the hardware.
The problem with text is that it's cheap to host. If all that you're sharing is text then you have enough data in a cheap mobile contract to share it with pretty much anyone who is interested. If you're sharing pictures, then a typical home broadband connection has enough spare upstream bandwidth to share them with pretty much anyone who might be interested (unless you 'go viral' or are DDoS'd). If you're sharing video, then you really want to host them in someone else's connection (unless you're one of the lucky people with 100Mb/s symmetric fibre to their house). That's a big selling point for Facebook - they, like Google, can easily absorb the cost of serving videos (especially as most of them will be watched by 0-1 person and disk space is cheap).
TFA also seems to conflate 'energy' with 'electricity'. Few people use coal for heating these days, but a lot of heating is still gas and heating is a huge proportion of total energy consumption. Even replacing 100% of electricity generation with renewables will leave a lot of gas still being burned.
Balancing the budget is simple arithmetic - don't spend more than what is available.
And that works perfectly, as long as nothing that the government spends money on will have any effect on the size of the economy and therefore on the amount of available tax revenue.
That's not an entirely fair comparison, as Intel is in a lot more markets than AMD. If the world decided to stop buying x86 processors tomorrow, then AMD would be completely sunk (or, close to, depending on how well their ARMv8 chips sell), whereas Intel would just shut down a few divisions. The majority of Intel's R&D spend is on process technology. This is why they're usually a generation ahead of their competition on fabs. AMD, in contrast, outsources the production and so this R&D money shows up as an expense on their balance sheet and is shared with other companies that fab chips in the same foundries as AMD.
Apple was fully capable of working out a deal if they thought it would be of value.
Spoken like someone who has no experience with Oracle's legal department.
The problem with ZFS is that it consumes more hardware resources
You might be surprised. It's quite useable on a laptop and Samsung had a slightly cut-down version of ZFS working well on mobile phones with 4MB of RAM.
Now? It means very little other than effectively guaranteeing a total ordering on files created on a single system. Remember, however, that filesystems typically stay in production for at least 20 years (35 for HFS+ is not that uncommon - UFS is even older) and so a little bit of headroom is always a good idea.
I back up my MacBook Pro to a FreeBSD box using ZFS with compression and deduplication (and snapshot it periodically, because if TimeMachine detects that your backups are corrupted then the only option is to delete them an redo from start, and it's nice to be able to revert just one backup if the last backup broke something). With lz4 compression, the compression ratio for the ZFS filesystem that I use as a backup target is 2.08x - that's a fairly hefty saving. It's harder to measure how much dedup is saving me, because it's a pool-wide property and I have ripped a load of my DVD collection, which probably doesn't contain many duplicate blocks. Currently the dedup ratio is 1.2x, but given that the TimeMachine volume is around 20% of the total space and contains a lot more redundancy than the rest, I wouldn't be surprised if I'm saving a relatively large amount there too.
I probably get more benefit from dedup on the backup disk (TimeMachine copies entire files, even if only a single block has changed) than I would on the laptop, but I'd expect to see similar benefits from compression on the source and the backup, and that 2x compression ratio looks pretty good...
Gzip is completely the wrong algorithm. Modern processors can handle lz4 compression faster than a single SSD can write (around 400MB/s) and decompression a lot faster than one can read (1-2GB/s). lz4 also has quick-fail, so will skip compression when entropy is low (e.g. files that are already compressed or random data), so in your example would add so little latency that it would be hard to measure.
Your experiment is also putting the compression in completely the wrong part of the pipeline. When reading or writing a large file, you will overlap the compression / decompression with the I/O in a real implementation (i.e. each block is decompressed as it arrives, you don't read 500MB and then decompress the whole lot), whereas you have made them sequential. Oh, and lz4 compression and decompression are easy to parallelise, so on the kind of system where 400MB/s write would be a bottleneck, you probably aren't doing all of the compression on a single core.
Try creating two ZFS filesystems in the same pool, one with lz4 and one with no compression, then write the same file to them both. You'll see a very small performance difference, but one that will largely be in the noise in normal use. Now change the compression mode from lz4 to gzip-1 (lowest gzip compression ratio) and you'll see a very noticeable drop in performance.
It wasn't before ZFS was CDDL, ZFS was CDDL from the start. I've heard two conflicting explanations of why Apple dropped support (after publicly announcing it) from different people in their CoreOS team. One of them involved Apple wanting a proprietary license (CDDL is per-file copyleft, so Apple would have had to release changes that they made to any of the ZFS files) with terms that Sun/Oracle wouldn't grant, though I get the impression that the NetApp lawsuit may have been more of an issue.
The general rule of thumb is 10 years between first release that adventurous users trust and solid enough for serious use on important data. ZFS past that point last year. BTRFS will in another three years (or 7 years, depending on how you count). The only exceptions are things like UFS2 and vFAT that are simple extensions to existing filesystems and embedded filesystems that are very simple.
It depends on the touchpad. The ones Apple buys (which I've seen on a couple of other laptops, but not many) are nice because:
They're nice and big.
They're one big button, so there's physical travel when you click.
They're multitouch and so they recognise two-finger click as right click (three-finger click as middle button) and recognise other gestures.
They're a bit inconvenient for games that bind the right or middle button in such a way that you want to be able to press multiple buttons, but for these I'd rather use a mouse than anything built into the laptop case.
I'm guessing that he wants to pressure the Democrats into picking a different candidate by announcing a different candidate, although it's difficult to imagine how they'd do that without looking stupid after Obama called her the 'most qualified candidate ever'. He perhaps hopes that the Democratic Party will tell her that if she does stand and win, then they'll vote along with the Republicans for impeachment as soon as she's elected and force her to hand over the Presidency to whoever she picks as her running mate.
Wrong. I'm comparing modern interpretation of Judaism and Christianity — you got that part right — with the modern interpretation of Islam
You're comparing the interpretation of Judaism and Christianity in mostly secular societies to the interpretation of Islam in strongly religious societies.
LinkedIn seemed to be almost a necessity for job hunting
Perhaps, if you want to work in really bad places. I've never had a LinkedIn profile (and the amount of spam that I get from them is enough to ensure that I never want one) and yet had no shortage of offers last time I was looking for a job. You might want to get someone who has been on the other end of a hiring process to look at your CV. Some of the ones that I see are truly terrible, yet people seem to think that they're sensible to submit for jobs.
Depends a bit on your userbase, but I really like Jekyll. The site it generates is purely static, so you can have a very small attack profile (web server running without write access to any of the filesystem and without access to a database). The source is separated into templates (edit only by people who understand HTML) and content. The content is in Markdown (or one of a few other simple markup languages) and there are a few nice WYSIWYG editors (I like MacDown on OS X, but there are many others). The sources are all stored in a filesystem tree, so you can easily use them with any revision control system (or, if your users can't be persuaded to use one, then with regular FS snapshots).
The only good reason to want to have a database backing a CMS is if you have a lot of dynamic contents (user comments and so on). For a mostly static site, it's overkill.
I guess you missed the Christians burning Muslims and Hindus in parts of India last year then? Oh, and the Christians burning people 600 years ago were doing so because they were the wrong kind of Christian, it wasn't even because they were members of a different religion, it was because they had the audacity to want to read their own holy book. Something that was still going on when I was growing up, with Protestant and Catholics killing each other in Ireland (and, just like the current crop of Islamic terrorists, with a big load of nationalism mixed in). Fortunately, that mostly stopped when some time around late 2001 something happened in New York that suddenly made it unfashionable for New Yorkers to fund terrorists.
There is a crucial difference between atheism and religious belief: the importance of evidence. If a god were to appear and demonstrate their existence in an empirically measurable way then atheists would mentally move them from the category of things that don't exist to things that do, though they would likely no more 'believe' in the god than they believe in a chair. Things that are measurable can be safely assumed to exist (for a purely utilitarian definition of 'exist').
There are an infinite number of things that don't exist. Atheists don't typically go around enumerating all of them any more than a Christian goes around enumerating all of the Egyptian, Norse, Roman, and so on gods that he or she doesn't believe in.
I'm slightly astonished by people who find it easy to disbelieve hundreds of religions and superstitions that have precisely as much evidence supporting them as their own are unable to understand how someone else disbelieves one more than them.
'Religious' people and atheists have more in common than either group has with agnostics. Religious people believe that all religions except for one are false. Atheists believe that all religions are false. Agnostics don't believe that any are false.
Animals do not, generally, behave as amoral rapists, murderers and child molesters
Rape is very common in the animal kingdom. The notion of consent is quite difficult in a species that doesn't have language, but even if you limit the definition of rape to the male holding down the female and forcing himself on her then it's still common. Go and see how ducks mate sometime - three or four of the males hold the female down and take it in turns. Or look at dolphins.
In terms of murder and child molestation, it's fairly common for a male (especially in the large cats) to kill the cubs fathered by another male that they displace.
I suspect that someone in government has seen the big the fines that the EU has collected for violations of the data protection directive and asked the ICO why they're not making the same kinds of contributions to the budget. Large as they are, the fines from the EU are still not enough to be more than the cost of doing business for some of these companies.
I don't want to teach the receptionist how to use Debian
Why not? A receptionist needs to be able to use an address book and calendar, possibly a word processor, and email. These things are basically the same on all major operating systems. There are lots of people that it would be difficult to migrate to a different OS, but the receptionist ought to be one of the easiest.
Helicopters have been used as air ambulances for a long time, but that doesn't mean that we're retiring the fleet of ground-based ones. The helicopters are a lot more expensive to run and have less weight capacity for potentially life-saving equipment in the back. They're only used when the ground-based ambulance can't get there in time (or, more commonly, at all).
We've had the technology to build flying cars for a long time, but without cheap energy few people are going to be able to afford them.
I was unaware we understood how our brains processed data
We understand how neurones work, which is roughly the equivalent to understanding how transistors work and saying that you understand a CPU. We also know that you can approximate any mathematical function with a network of neurones. The accuracy of the approximation depends on the complexity of the network and with modern computers you can simulate very complex networks (not in comparison to an actual brain, but in comparison to the networks built by the people who worked on this stuff 50 years ago). The two big advantages of neural networks over conventional programs are:
If you have a load of input and output from a function, you can train a network to approximate the function even if you don't understand what the function actually is.
It's easy to trade accuracy for performance / power consumption in a neural network. It's hard with most algorithms to take an accurate algorithm and replace it with one that's 70% accurate but uses far less power. For a lot of pattern-matching things, a fairly high false positive or false negative rate is still useful.
There is a huge amount of hype around neural nets at the moment, because an entirely new generation of young computer scientists has learned that they can approximate any function (most of the old ones that knew this have retired or died).
Slashdot Mirror
and PC ROMs no longer implement ASCII, but instead have native video codec support.
For what it's worth, the last new graphics cards to have hardware support for ASCII were released over 15 years ago. In modern cards, it's all done in firmware and with EFI the native interface is a framebuffer with text rendering layered on top. A typical GPU has hardware acceleration for a number of video CODECs, but no acceleration for text (though antialiased text rendering makes heavy use of the compositing engine).
That's nothing to do with demand though, it's entirely as a result of the different costs. Rendering fixed-width text is something that a modern CPU can do without the user noticing even if it takes an interrupt for every character and runs some emulation code in system management mode so there's no point devoting any transistors to it in the hardware.
The problem with text is that it's cheap to host. If all that you're sharing is text then you have enough data in a cheap mobile contract to share it with pretty much anyone who is interested. If you're sharing pictures, then a typical home broadband connection has enough spare upstream bandwidth to share them with pretty much anyone who might be interested (unless you 'go viral' or are DDoS'd). If you're sharing video, then you really want to host them in someone else's connection (unless you're one of the lucky people with 100Mb/s symmetric fibre to their house). That's a big selling point for Facebook - they, like Google, can easily absorb the cost of serving videos (especially as most of them will be watched by 0-1 person and disk space is cheap).
TFA also seems to conflate 'energy' with 'electricity'. Few people use coal for heating these days, but a lot of heating is still gas and heating is a huge proportion of total energy consumption. Even replacing 100% of electricity generation with renewables will leave a lot of gas still being burned.
Balancing the budget is simple arithmetic - don't spend more than what is available.
And that works perfectly, as long as nothing that the government spends money on will have any effect on the size of the economy and therefore on the amount of available tax revenue.
That's not an entirely fair comparison, as Intel is in a lot more markets than AMD. If the world decided to stop buying x86 processors tomorrow, then AMD would be completely sunk (or, close to, depending on how well their ARMv8 chips sell), whereas Intel would just shut down a few divisions. The majority of Intel's R&D spend is on process technology. This is why they're usually a generation ahead of their competition on fabs. AMD, in contrast, outsources the production and so this R&D money shows up as an expense on their balance sheet and is shared with other companies that fab chips in the same foundries as AMD.
I would assume the reasons were more technical.
Apple had ZFS working in the betas.
Apple was fully capable of working out a deal if they thought it would be of value.
Spoken like someone who has no experience with Oracle's legal department.
The problem with ZFS is that it consumes more hardware resources
You might be surprised. It's quite useable on a laptop and Samsung had a slightly cut-down version of ZFS working well on mobile phones with 4MB of RAM.
Now? It means very little other than effectively guaranteeing a total ordering on files created on a single system. Remember, however, that filesystems typically stay in production for at least 20 years (35 for HFS+ is not that uncommon - UFS is even older) and so a little bit of headroom is always a good idea.
I back up my MacBook Pro to a FreeBSD box using ZFS with compression and deduplication (and snapshot it periodically, because if TimeMachine detects that your backups are corrupted then the only option is to delete them an redo from start, and it's nice to be able to revert just one backup if the last backup broke something). With lz4 compression, the compression ratio for the ZFS filesystem that I use as a backup target is 2.08x - that's a fairly hefty saving. It's harder to measure how much dedup is saving me, because it's a pool-wide property and I have ripped a load of my DVD collection, which probably doesn't contain many duplicate blocks. Currently the dedup ratio is 1.2x, but given that the TimeMachine volume is around 20% of the total space and contains a lot more redundancy than the rest, I wouldn't be surprised if I'm saving a relatively large amount there too.
I probably get more benefit from dedup on the backup disk (TimeMachine copies entire files, even if only a single block has changed) than I would on the laptop, but I'd expect to see similar benefits from compression on the source and the backup, and that 2x compression ratio looks pretty good...
Gzip is completely the wrong algorithm. Modern processors can handle lz4 compression faster than a single SSD can write (around 400MB/s) and decompression a lot faster than one can read (1-2GB/s). lz4 also has quick-fail, so will skip compression when entropy is low (e.g. files that are already compressed or random data), so in your example would add so little latency that it would be hard to measure.
Your experiment is also putting the compression in completely the wrong part of the pipeline. When reading or writing a large file, you will overlap the compression / decompression with the I/O in a real implementation (i.e. each block is decompressed as it arrives, you don't read 500MB and then decompress the whole lot), whereas you have made them sequential. Oh, and lz4 compression and decompression are easy to parallelise, so on the kind of system where 400MB/s write would be a bottleneck, you probably aren't doing all of the compression on a single core.
Try creating two ZFS filesystems in the same pool, one with lz4 and one with no compression, then write the same file to them both. You'll see a very small performance difference, but one that will largely be in the noise in normal use. Now change the compression mode from lz4 to gzip-1 (lowest gzip compression ratio) and you'll see a very noticeable drop in performance.
It wasn't before ZFS was CDDL, ZFS was CDDL from the start. I've heard two conflicting explanations of why Apple dropped support (after publicly announcing it) from different people in their CoreOS team. One of them involved Apple wanting a proprietary license (CDDL is per-file copyleft, so Apple would have had to release changes that they made to any of the ZFS files) with terms that Sun/Oracle wouldn't grant, though I get the impression that the NetApp lawsuit may have been more of an issue.
The general rule of thumb is 10 years between first release that adventurous users trust and solid enough for serious use on important data. ZFS past that point last year. BTRFS will in another three years (or 7 years, depending on how you count). The only exceptions are things like UFS2 and vFAT that are simple extensions to existing filesystems and embedded filesystems that are very simple.
They're a bit inconvenient for games that bind the right or middle button in such a way that you want to be able to press multiple buttons, but for these I'd rather use a mouse than anything built into the laptop case.
I'm guessing that he wants to pressure the Democrats into picking a different candidate by announcing a different candidate, although it's difficult to imagine how they'd do that without looking stupid after Obama called her the 'most qualified candidate ever'. He perhaps hopes that the Democratic Party will tell her that if she does stand and win, then they'll vote along with the Republicans for impeachment as soon as she's elected and force her to hand over the Presidency to whoever she picks as her running mate.
Wrong. I'm comparing modern interpretation of Judaism and Christianity — you got that part right — with the modern interpretation of Islam
You're comparing the interpretation of Judaism and Christianity in mostly secular societies to the interpretation of Islam in strongly religious societies.
LinkedIn seemed to be almost a necessity for job hunting
Perhaps, if you want to work in really bad places. I've never had a LinkedIn profile (and the amount of spam that I get from them is enough to ensure that I never want one) and yet had no shortage of offers last time I was looking for a job. You might want to get someone who has been on the other end of a hiring process to look at your CV. Some of the ones that I see are truly terrible, yet people seem to think that they're sensible to submit for jobs.
in 2016 vi is just an alias for vim
That is only true on most Linux systems. Most BSD systems (and a few Linux distros) install nvi as vi.
Depends a bit on your userbase, but I really like Jekyll. The site it generates is purely static, so you can have a very small attack profile (web server running without write access to any of the filesystem and without access to a database). The source is separated into templates (edit only by people who understand HTML) and content. The content is in Markdown (or one of a few other simple markup languages) and there are a few nice WYSIWYG editors (I like MacDown on OS X, but there are many others). The sources are all stored in a filesystem tree, so you can easily use them with any revision control system (or, if your users can't be persuaded to use one, then with regular FS snapshots).
The only good reason to want to have a database backing a CMS is if you have a lot of dynamic contents (user comments and so on). For a mostly static site, it's overkill.
I guess you missed the Christians burning Muslims and Hindus in parts of India last year then? Oh, and the Christians burning people 600 years ago were doing so because they were the wrong kind of Christian, it wasn't even because they were members of a different religion, it was because they had the audacity to want to read their own holy book. Something that was still going on when I was growing up, with Protestant and Catholics killing each other in Ireland (and, just like the current crop of Islamic terrorists, with a big load of nationalism mixed in). Fortunately, that mostly stopped when some time around late 2001 something happened in New York that suddenly made it unfashionable for New Yorkers to fund terrorists.
There is a crucial difference between atheism and religious belief: the importance of evidence. If a god were to appear and demonstrate their existence in an empirically measurable way then atheists would mentally move them from the category of things that don't exist to things that do, though they would likely no more 'believe' in the god than they believe in a chair. Things that are measurable can be safely assumed to exist (for a purely utilitarian definition of 'exist').
There are an infinite number of things that don't exist. Atheists don't typically go around enumerating all of them any more than a Christian goes around enumerating all of the Egyptian, Norse, Roman, and so on gods that he or she doesn't believe in.
I'm slightly astonished by people who find it easy to disbelieve hundreds of religions and superstitions that have precisely as much evidence supporting them as their own are unable to understand how someone else disbelieves one more than them.
'Religious' people and atheists have more in common than either group has with agnostics. Religious people believe that all religions except for one are false. Atheists believe that all religions are false. Agnostics don't believe that any are false.
Animals do not, generally, behave as amoral rapists, murderers and child molesters
Rape is very common in the animal kingdom. The notion of consent is quite difficult in a species that doesn't have language, but even if you limit the definition of rape to the male holding down the female and forcing himself on her then it's still common. Go and see how ducks mate sometime - three or four of the males hold the female down and take it in turns. Or look at dolphins.
In terms of murder and child molestation, it's fairly common for a male (especially in the large cats) to kill the cubs fathered by another male that they displace.
I suspect that someone in government has seen the big the fines that the EU has collected for violations of the data protection directive and asked the ICO why they're not making the same kinds of contributions to the budget. Large as they are, the fines from the EU are still not enough to be more than the cost of doing business for some of these companies.
I don't want to teach the receptionist how to use Debian
Why not? A receptionist needs to be able to use an address book and calendar, possibly a word processor, and email. These things are basically the same on all major operating systems. There are lots of people that it would be difficult to migrate to a different OS, but the receptionist ought to be one of the easiest.
Helicopters have been used as air ambulances for a long time, but that doesn't mean that we're retiring the fleet of ground-based ones. The helicopters are a lot more expensive to run and have less weight capacity for potentially life-saving equipment in the back. They're only used when the ground-based ambulance can't get there in time (or, more commonly, at all).
We've had the technology to build flying cars for a long time, but without cheap energy few people are going to be able to afford them.
I was unaware we understood how our brains processed data
We understand how neurones work, which is roughly the equivalent to understanding how transistors work and saying that you understand a CPU. We also know that you can approximate any mathematical function with a network of neurones. The accuracy of the approximation depends on the complexity of the network and with modern computers you can simulate very complex networks (not in comparison to an actual brain, but in comparison to the networks built by the people who worked on this stuff 50 years ago). The two big advantages of neural networks over conventional programs are:
There is a huge amount of hype around neural nets at the moment, because an entirely new generation of young computer scientists has learned that they can approximate any function (most of the old ones that knew this have retired or died).