Am I free to compile, without modification, a copy of the code that I receive and give it to a less-technical friend? (GPLv2: Only if I give him either a copy of the source code or a written offer good for 3 [I think] years to provide him with the source code on demand. GPLv3: yes, if I also give him a link to where he can download the source).
Am I free to link against it in a proprietary program and call a single function that consumes a string and produces a string as output? (No).
Am I free to write an BSD-licensed wrapper around the library that runs in a separate process and receives a string from stdin and writes the result to stdout, publish that, and use it from my program? (Not 100% sure, but the FSF lawyers believe that the answer is yes)
Am I free to create some well-defined interfaces, ship a proprietary program that uses them and can load another module, wrap the GPL'd library in some BSDL code that exposes these interfaces, and have my program load it at run time (Yes, probably, though not tested in court - lots of lawyers agree that this one is fine though).
Am I free to ship a proprietary program that can optionally load a GPL'd library and use its functionality directly, as long as I don't distribute the GPL'd code? (Maybe, depending on the copyright status of the interfaces that I use, which Oracle vs Google has now made a lot more murky. Probably 50:50 which way a court would go on this one.)
Europeans love to hate on American companies. Just see all of the court cases the European Union brings against American companies (Apple and Microsoft being the first two to spring to mind).
Have you looked at the list of court cases the European Union brings against companies? I'm guessing not, and I'm also guessing that you read about these court cases only in the tech press where they're reporting only on cases brought against big tech firms where American companies tend to dominate, and not against other markets where EU companies dominate.
That was the thing that made me cringe whenever I read about the Munich migration. Trying to move everything at once was a political statement, not a practical one. Windows should be the last proprietary software product that you abandon, and when you do it should be easy. First, move all of your back-end stuff to alternatives that use open protocols and work with different clients. Then move the clients for these over. Then start moving to LibreOffice or OpenOffice - have both installed, but mandate that new documents must be in the OpenOffice file formats. Then move to having MS Office on one machine per office that's used for legacy documents that don't open correctly. Move all of your other apps over to portable alternatives. When it comes time to replace Windows, you aren't running any Windows-only software and if you pick a DE with a Windows-like theme most of your uses won't even notice.
Our big build machines have been 24 or 32 core for quite a few years, so neither of these gives us a huge performance improvement. We'll evaluate them when we get around to buying more, but from what I've seen they just mean that our next upgrade will be cheaper, not significantly faster.
No, but I can imagine doing it on a bluetooth keyboard and a phone. I wrote about 20-30 articles on my old Nokia 770 and ThinkOutside folding keyboard. The 770 is pretty limited, but was able to run an xterm and vim quite happily. For a couple of summers, I'd wake up, stroll across the park and along the beach to a cafe overlooking the sea, read and drink coffee for half an hour to an hour, and then get out the keyboard and machine and write for an hour or two. The keyboard and 770 would fit in my pockets, so were more convenient for me to take than my laptop. My phone is vastly more powerful than my 770, and smaller.
On the other hand, remote terminal usage has become a lot better. When I first used vim over ssh, its tendency to completely redraw parts of the screen made it noticeably slow, even with a machine not far from the other end of my dial-up link. nvi was a lot more useable. Over Christmas, I was using vim on a machine in a different country via SSH and even with pretty crappy WiFi at my end it was fine - and the rebuild times on the 24-core machine with 256GB of RAM that I was ssh'd into made it a much better experience than working on my laptop.
I remember shitty laptop keyboards. I remember ones with so little travel that your fingers hurt after 20 minutes of typing. I remember ones with a spring right in the middle and a really crappy mechanism so if you hit them slightly off centre they'd bend and not register a key press.
I haven't seen a shitty keyboard on any laptop for about 10 years. There are a few really nice ones but most, including the Macs, have been good enough for a long time.
I haven't used a Mac with the OLED bar, but some of my colleagues have them. If you're in the terminal, they'll show the function keys (though that's configurable and a few command-line apps do modify the display). For most other things, they show context info that is more useful than having to remember what F5 does in this particular application (for example, in XCode they'll show things like 'run' and 'debug').
In the corporate office we have PC's on 5 year replacement cycles
How long have you been doing 5-year replacements? We used to do 3-year replacements, but that's been gradually extended. My work machine is now over four years old and is eligible for replacement, but there's nothing really compelling to replace it with. The same is true for everything from laptops to our big build servers. On our old one, I tried running poudriere and rebuilding the entire FreeBSD ports collection. It took 24 hours, but the last 4 hours were spent downloading the Vega Strike game data files from a very slow upstream source. On the newer machines, it's closer to 16 hours, but that's not really a compelling upgrade - for most things, we get a bigger return from buying more machines, rather than replacing old ones (we can never have enough continuous integration machines, for example).
Gold is a good conductor and doesn't corrode in air. These two things mean that there is a real demand for gold for electrical connectors and jewellery. It's relatively scarce, so the demand is high in proportion to the supply. If gold were cheaper, then we'd plate a lot more things in gold (e.g. pretty much every electrical connector - gold isn't quite as good a conductor as copper, but it's a much better conductor than copper oxide). That gives a lower bound on the price of gold: if it were plentiful then we'd use it for a lot more and its price would drop to close to that of copper or aluminium, but not to zero.
Salt hasn't been a viable currency for quite a long time, but that doesn't mean that people don't still trade salt in futures markets.
It can take a long time though, and while you're riding the upwards curve and selling slowly you can make a lot of money. The thing that people always forget when they read about the tulip bubble and he wall street crash is that as many people became rich as went bankrupt. It's a zero-sum game, so every dollar someone loses will be won by someone else. In many ways, it resembles a poker game where both players are bluffing. Eventually either one will fold or they'll call and whoever has the higher card will win.
I have not speculated on bitcoin, because I don't have any confidence that I can predict the inflection point well enough to find a greater fool before it does.
The test had little to nothing to do with what she learned in high school or what she's learning in college right now
Note that this is intentional. The SAT is intended as an aptitude test. As such, it is intended to measure your ability as independent of knowledge and learned skills as possible. This is obviously impossible, but tests like the SAT and IQ tests try to get as close as they can. Unfortunately, it is often possible with such tests to learn for a particular style of test (and you can't significantly change the style without compromising reliability). There's some research that indicates that you get much more useful information by making people take a lot of these tests and comparing their best and worst marks, but that is not normally practical.
Maybe they own Apple stock. If they're clever, then they bought a load of Apple shares, announced that Apple devices were too secure for them to be able to crack but that all of their competitors' devices weren't, and then waited for the media to pick this up before selling their shares.
Please don't make arguments like that, because that's exactly the same flawed logic that climate change deniers use. All this tells you is that something in your local climate has changed. The global warming argues that pumping large amounts of energy into the atmosphere as a result of an increase in the greenhouse effect will result in either a new equilibrium condition or a more chaotic global climate. Both going from never snowing to snowing each year or going from snowing each year to never snowing are consistent with this hypothesis, but for local climates they're also explainable by a number of other mechanisms. To either support or contradict a hypothesis like global warming, you need to look at global data.
A better analogy would be a tobacco company. Can a smoker sue a tobacco company that covered up reports of health damage as a result of smoking and continued to sell their products without any health warnings for decades after they became aware that they cause cancer and death? The answer, by the way, is yes and they have done so successfully.
Who needs to vote, when candidates can't win without spending a lot of money and you can financially support both candidates on the tacit understanding that they either act in your interest or will lose your support next round?
Microsoft does this with Azure. Their EU operation is run by a separate company, so they cannot be compelled to hand over data to US law enforcement because they have a complete separation and no one in Microsoft has access to the data in the first place (at least, in theory). I believe the other big cloud providers have similar things, because otherwise they can't sell hosting to any company that needs to comply with the GDPR (i.e. anyone in the EU).
Maybe where you live, but in the UK the analysis of recent election results tells the exact opposite story. Older people were far more swayed by appeals to emotion than to logic.
It's difficult to do. The gerrymandered districts will have a similar number of people in them, but that's not the problem. Imagine you have 10,000 people voting for teams A and B, with a 50:50 split. It's possible to arrange 10 districts where each one will have 500 A voters and 500 B voters, but it's also possible to have 8 districts with 600 A voters and 400 B voters and two districts with 100 A voters and 900 B voters. In the first configuration, you'll end up with some tough campaigns and probably average 5 A seats and 5 B seats. In the second, you'll always have 8 A seats and 2 B seats. That's the essence of gerrymandering: you distribute your voters to maximise the number of seats.
You can automate creating constituencies with even numbers, but that's also biased. The real solution is to move away from single-representative constituencies and closer to an electoral system where every vote counts.
How many of those users are actually committing code? I've gone to SF a few times recently to look at the code for orphaned projects and see if they're worth reviving, but that's about it. I'd love to see some serious competition to GitHub, but I'm not convinced that SF is in a position to provide it. GitHub at least has a solid business model (get people hooked on the free service, then sell them various degrees of hosted service).
The main advantage for a project using GitHub is the network effect. I have a couple of things hosted there and get a lot of interaction with users via the issue trackers, because the barrier to entry is tiny. Similarly, these days I am far more likely to report a bug if the project is on GitHub, because I already have an account and it's easy.
Back when those weavers got replaced there were actual shortages of wealth. Lots of people had to die because there wasn't enough food to support the birth rate.
That's not really true. This coincided with the agricultural revolution, which dramatically increased the food supply. There were all sorts of distribution problems, but there was more food than was needed. Things like the Irish Potato Famine were a bit later, but had causes that didn't relate to available food production ability.
We'll always accept them. We have a couple of machines specifically set up for this kind of thing, one with a USB analyser to look for low-level attacks and a RPi to look for software attacks. It's always interesting to see what you get on a free USB stick (police agency or otherwise).
Slashdot Mirror
And so they released GPL'd software, which Red Hat and Canonical bundled up into packages and charged people for?
Am I free to compile, without modification, a copy of the code that I receive and give it to a less-technical friend? (GPLv2: Only if I give him either a copy of the source code or a written offer good for 3 [I think] years to provide him with the source code on demand. GPLv3: yes, if I also give him a link to where he can download the source).
Am I free to link against it in a proprietary program and call a single function that consumes a string and produces a string as output? (No).
Am I free to write an BSD-licensed wrapper around the library that runs in a separate process and receives a string from stdin and writes the result to stdout, publish that, and use it from my program? (Not 100% sure, but the FSF lawyers believe that the answer is yes)
Am I free to create some well-defined interfaces, ship a proprietary program that uses them and can load another module, wrap the GPL'd library in some BSDL code that exposes these interfaces, and have my program load it at run time (Yes, probably, though not tested in court - lots of lawyers agree that this one is fine though).
Am I free to ship a proprietary program that can optionally load a GPL'd library and use its functionality directly, as long as I don't distribute the GPL'd code? (Maybe, depending on the copyright status of the interfaces that I use, which Oracle vs Google has now made a lot more murky. Probably 50:50 which way a court would go on this one.)
Yup, the GPL is very simple.
Europeans love to hate on American companies. Just see all of the court cases the European Union brings against American companies (Apple and Microsoft being the first two to spring to mind).
Have you looked at the list of court cases the European Union brings against companies? I'm guessing not, and I'm also guessing that you read about these court cases only in the tech press where they're reporting only on cases brought against big tech firms where American companies tend to dominate, and not against other markets where EU companies dominate.
That was the thing that made me cringe whenever I read about the Munich migration. Trying to move everything at once was a political statement, not a practical one. Windows should be the last proprietary software product that you abandon, and when you do it should be easy. First, move all of your back-end stuff to alternatives that use open protocols and work with different clients. Then move the clients for these over. Then start moving to LibreOffice or OpenOffice - have both installed, but mandate that new documents must be in the OpenOffice file formats. Then move to having MS Office on one machine per office that's used for legacy documents that don't open correctly. Move all of your other apps over to portable alternatives. When it comes time to replace Windows, you aren't running any Windows-only software and if you pick a DE with a Windows-like theme most of your uses won't even notice.
Our big build machines have been 24 or 32 core for quite a few years, so neither of these gives us a huge performance improvement. We'll evaluate them when we get around to buying more, but from what I've seen they just mean that our next upgrade will be cheaper, not significantly faster.
No, but I can imagine doing it on a bluetooth keyboard and a phone. I wrote about 20-30 articles on my old Nokia 770 and ThinkOutside folding keyboard. The 770 is pretty limited, but was able to run an xterm and vim quite happily. For a couple of summers, I'd wake up, stroll across the park and along the beach to a cafe overlooking the sea, read and drink coffee for half an hour to an hour, and then get out the keyboard and machine and write for an hour or two. The keyboard and 770 would fit in my pockets, so were more convenient for me to take than my laptop. My phone is vastly more powerful than my 770, and smaller.
On the other hand, remote terminal usage has become a lot better. When I first used vim over ssh, its tendency to completely redraw parts of the screen made it noticeably slow, even with a machine not far from the other end of my dial-up link. nvi was a lot more useable. Over Christmas, I was using vim on a machine in a different country via SSH and even with pretty crappy WiFi at my end it was fine - and the rebuild times on the 24-core machine with 256GB of RAM that I was ssh'd into made it a much better experience than working on my laptop.
I remember shitty laptop keyboards. I remember ones with so little travel that your fingers hurt after 20 minutes of typing. I remember ones with a spring right in the middle and a really crappy mechanism so if you hit them slightly off centre they'd bend and not register a key press.
I haven't seen a shitty keyboard on any laptop for about 10 years. There are a few really nice ones but most, including the Macs, have been good enough for a long time.
I haven't used a Mac with the OLED bar, but some of my colleagues have them. If you're in the terminal, they'll show the function keys (though that's configurable and a few command-line apps do modify the display). For most other things, they show context info that is more useful than having to remember what F5 does in this particular application (for example, in XCode they'll show things like 'run' and 'debug').
In the corporate office we have PC's on 5 year replacement cycles
How long have you been doing 5-year replacements? We used to do 3-year replacements, but that's been gradually extended. My work machine is now over four years old and is eligible for replacement, but there's nothing really compelling to replace it with. The same is true for everything from laptops to our big build servers. On our old one, I tried running poudriere and rebuilding the entire FreeBSD ports collection. It took 24 hours, but the last 4 hours were spent downloading the Vega Strike game data files from a very slow upstream source. On the newer machines, it's closer to 16 hours, but that's not really a compelling upgrade - for most things, we get a bigger return from buying more machines, rather than replacing old ones (we can never have enough continuous integration machines, for example).
Gold is a good conductor and doesn't corrode in air. These two things mean that there is a real demand for gold for electrical connectors and jewellery. It's relatively scarce, so the demand is high in proportion to the supply. If gold were cheaper, then we'd plate a lot more things in gold (e.g. pretty much every electrical connector - gold isn't quite as good a conductor as copper, but it's a much better conductor than copper oxide). That gives a lower bound on the price of gold: if it were plentiful then we'd use it for a lot more and its price would drop to close to that of copper or aluminium, but not to zero.
Salt hasn't been a viable currency for quite a long time, but that doesn't mean that people don't still trade salt in futures markets.
It can take a long time though, and while you're riding the upwards curve and selling slowly you can make a lot of money. The thing that people always forget when they read about the tulip bubble and he wall street crash is that as many people became rich as went bankrupt. It's a zero-sum game, so every dollar someone loses will be won by someone else. In many ways, it resembles a poker game where both players are bluffing. Eventually either one will fold or they'll call and whoever has the higher card will win.
I have not speculated on bitcoin, because I don't have any confidence that I can predict the inflection point well enough to find a greater fool before it does.
The test had little to nothing to do with what she learned in high school or what she's learning in college right now
Note that this is intentional. The SAT is intended as an aptitude test. As such, it is intended to measure your ability as independent of knowledge and learned skills as possible. This is obviously impossible, but tests like the SAT and IQ tests try to get as close as they can. Unfortunately, it is often possible with such tests to learn for a particular style of test (and you can't significantly change the style without compromising reliability). There's some research that indicates that you get much more useful information by making people take a lot of these tests and comparing their best and worst marks, but that is not normally practical.
I'd be interested to know which country that is. Holland?
The EU, by passing the GDPR.
Maybe they own Apple stock. If they're clever, then they bought a load of Apple shares, announced that Apple devices were too secure for them to be able to crack but that all of their competitors' devices weren't, and then waited for the media to pick this up before selling their shares.
Nope, they simply told Microsoft that no company could legally do business with Microsoft if they operated their own datacenter.
Please don't make arguments like that, because that's exactly the same flawed logic that climate change deniers use. All this tells you is that something in your local climate has changed. The global warming argues that pumping large amounts of energy into the atmosphere as a result of an increase in the greenhouse effect will result in either a new equilibrium condition or a more chaotic global climate. Both going from never snowing to snowing each year or going from snowing each year to never snowing are consistent with this hypothesis, but for local climates they're also explainable by a number of other mechanisms. To either support or contradict a hypothesis like global warming, you need to look at global data.
A better analogy would be a tobacco company. Can a smoker sue a tobacco company that covered up reports of health damage as a result of smoking and continued to sell their products without any health warnings for decades after they became aware that they cause cancer and death? The answer, by the way, is yes and they have done so successfully.
Who needs to vote, when candidates can't win without spending a lot of money and you can financially support both candidates on the tacit understanding that they either act in your interest or will lose your support next round?
Microsoft does this with Azure. Their EU operation is run by a separate company, so they cannot be compelled to hand over data to US law enforcement because they have a complete separation and no one in Microsoft has access to the data in the first place (at least, in theory). I believe the other big cloud providers have similar things, because otherwise they can't sell hosting to any company that needs to comply with the GDPR (i.e. anyone in the EU).
Many seniors engage in evidence-based-reasoning
Maybe where you live, but in the UK the analysis of recent election results tells the exact opposite story. Older people were far more swayed by appeals to emotion than to logic.
It's difficult to do. The gerrymandered districts will have a similar number of people in them, but that's not the problem. Imagine you have 10,000 people voting for teams A and B, with a 50:50 split. It's possible to arrange 10 districts where each one will have 500 A voters and 500 B voters, but it's also possible to have 8 districts with 600 A voters and 400 B voters and two districts with 100 A voters and 900 B voters. In the first configuration, you'll end up with some tough campaigns and probably average 5 A seats and 5 B seats. In the second, you'll always have 8 A seats and 2 B seats. That's the essence of gerrymandering: you distribute your voters to maximise the number of seats.
You can automate creating constituencies with even numbers, but that's also biased. The real solution is to move away from single-representative constituencies and closer to an electoral system where every vote counts.
How many of those users are actually committing code? I've gone to SF a few times recently to look at the code for orphaned projects and see if they're worth reviving, but that's about it. I'd love to see some serious competition to GitHub, but I'm not convinced that SF is in a position to provide it. GitHub at least has a solid business model (get people hooked on the free service, then sell them various degrees of hosted service).
The main advantage for a project using GitHub is the network effect. I have a couple of things hosted there and get a lot of interaction with users via the issue trackers, because the barrier to entry is tiny. Similarly, these days I am far more likely to report a bug if the project is on GitHub, because I already have an account and it's easy.
Back when those weavers got replaced there were actual shortages of wealth. Lots of people had to die because there wasn't enough food to support the birth rate.
That's not really true. This coincided with the agricultural revolution, which dramatically increased the food supply. There were all sorts of distribution problems, but there was more food than was needed. Things like the Irish Potato Famine were a bit later, but had causes that didn't relate to available food production ability.
We'll always accept them. We have a couple of machines specifically set up for this kind of thing, one with a USB analyser to look for low-level attacks and a RPi to look for software attacks. It's always interesting to see what you get on a free USB stick (police agency or otherwise).