And if you dont run it thru the debugger and STEP thru it you are just guessing what it will do.
Your inexperience is showing. Quite apart from the fact that you should make your assumptions explicit in your code (if the argument must not be null, put code in to check it!) you have the problem that there's a lot of code where attaching a debugger is impossible because it runs in a tricky environment. You can try to test by putting the code in an environment mostly populated by mocked components, but you can't be sure that you've got things right in such a situation: more often than not, the problem comes when components undergo a subtle change unannounced. (I had to deal with such a problem a few weeks ago. Everything worked within itself, but failed to function correctly when integrated; the problem was a change to the encoding of one of the files used in the interface itself. The whole shebang was rather security-aware too, so debugging was super-difficult.)
Many time I step thru my code to find some assumption I was making that is invalid.
You can write code that compiles with 0 warnings on the highest levels, can get thru the most stringent of lint checks, passed dozens of code reviews, pair wise coded, etc, etc etc. But until you run it and step thru and see you will never know.
Single-stepping is crappy way to debug code. Far better to have a consistent fundamental model and prove that your code adheres to it. It's a technique that tends towards the mathematical/logical end of things, but it works far better than simply trying to step through in that it can catch problems where required code is missing. (OTOH, stepping can help with cases where you've got a really nasty abstraction-busting bug like an endian-specific stack smash. I've seen some weird stuff in my time, and I now mostly pick my languages to avoid the worst disasters.)
Best complement I get from fellow engieres "your code is easy to read and when it screws up it does it in a way I can tell what is wrong"
My favorite is "I can't see how else you could possibly do it" when I know there's loads of ways which are all worse; means that the model I've picked is actively guiding people away from the dangers.
A hot TSA agent is like a cold blue-giant star. It just doesn't happen.
It depends on what level the AC is set to. Some airports are really annoyingly warm, especially in summer, but spring and autumn are worse because then I'm likely to have warm clothing with me (I needed it when I left home and can't put it in my luggage until I get to my final destination).
Oh, you meant the other meaning of "hot". Never mind.
It does, but not a consolidated one like the US has and constitutionality is genuinely a matter for experts only.
That said, one of the general requirements on the authorities (including both the police and social services) is that the actions they take are proportionate to the type of risk that they are intended to prevent. This is a very far-reaching principle that enormously constrains the power of the state to act in an arbitrary fashion. I believe there's also a right for action to be taken reasonably promptly; dragging things out leaves the state open to having to pay compensation (plus the legal costs of the challenge; UK courts usually award costs provided those costs are themselves reasonable and proportionate). That all adds up to the authorities having to be able to show that this was the right action to take (given the facts, which we don't have) and to having to get on with the investigation or drop the order.
So how do you deal with servers that have more than one purpose?
If you've got too many purposes on a single machine, that's an indication of potential problems anyway. Split up into separate VMs (with their own names) and the physical machine then has a clear purpose: hosting VMs.
At least in Europe, software patents aren't valid.
You are wrong. Software patents are valid in Europe, but they've got to show that they're doing something genuinely new. If someone's doing something that's really 10–20 years ahead, it's really not a problem. The problem is when someone's getting a patent for work that is only 2 months ahead (or worse, years behind). Those sorts of patents (whether for software or otherwise) are a real blight as they end up as combination of a blight on the state of the art and a bunch of landmines.
(What's the collective noun for such a thing? A blighted field of landmines?)
I would worry if important projects with large budgets and generous timeframes switched from Java to e.g. Ruby, but this won't happen.
You're more likely to see large projects use multiple languages (e.g., Java to make components, JRuby to stitch them together). This makes a lot of sense because it is using languages for the purposes they were designed for; Building high-level apps in just low-level languages is a PITA because of the amount of work involved, and building low-level apps with high-level languages is a very peculiar kind of crazy, but using things for their strengths is just smart. By using such combinations, large projects can achieve much more than they would otherwise (or what would have been a large project could be brought within the grasp of a small team).
I suspect that the real story of programming languages over the past few decades has been the slow death of monolingualism. These days, I only really see it with noobs and diehard C++/Lisp aficionados.
I suspect the porn company is not liability limited and probably has lots of cash. Sue them, and let them sort it out with Flickr.
Sue both of them. It should also be possible to work out some kind of libel claim in there ("They were alleging I was a copyright pirate, your honour, yet I have enormous respect for all forms of IP as it is a key part of how I earn my living!") Now, Gorman's a UK citizen and is resident in the UK, so using a UK court to pursue a libel claim is A-OK. Heh heh...
A grand a year is assuming a class of at least about 250 students and all of them buying the book (instead of borrowing it from the library or sharing). And that's for at least a month's worth of work - at the speed most lecturers write it's 2-3 months. Hardly a good return on investment.
I've been involved with writing a book (admittedly a tech book, not a college book) and given the sheer amount of work involved, I surely was not doing it to get rich. Heck, I wasn't doing to pay my yearly beer bill and I'm a light drinker. (It's certainly nowhere close to what I spend on caffeine!) It's only a tiny proportion of people who can ever earn a basic living from writing books; I don't appear to be one of them either.
And the grand a year would not last all that long. Books go obsolete and students trade them. Preventing that requires more work, worsening the rate of return...
I saw an analysis of their Terms of Service somewhere, indicating that they will only compensate up to the value of the service paid. So, if your service was $100/mo, they'd only compensate you for the downtime you experienced, or up to that month's service charge of $100.
But such limitations would not necessarily apply if they (or some of their employees) were engaging in malfeasance. Criminal activity by a trusted party changes the game quite substantially. Was this the case this time? I have no idea, but <hypothentically> if an employee were to see that someone is using a server for BTC mining and was to use this knowledge plus their own privileges to steal, it would leave the firm at least partially on the hook. </hypothetically> Whether or not a system is correctly and sanely used and adequately protected doesn't make it right to steal from it, though it might affect compensation levels (and whether any insurers involved give you anything; they usually include clauses to require you to take proper steps to protect your property).
The best applications on the early internet were about connecting people to each other.
That's changed? Not the way I see it. Yes, the way in which people connect to each other may well have changed, but the fundamental thing that wins is still there: connecting people together.
Personally I think it is better if people without an axe to grind work out how to use taxes in a socially beneficial way and politicians only get to vote on it.
I understand what you are saying, but you're completely off-base there. What to tax and what to spend the tax income on are right at the very core of politics. This has been the case for as long as there have been governments at all. It predates the US. It predates democracy. It predates money. It might predate cities (hard to say for sure; records that far back are sketchy but the earliest surviving records appear to be for tax purposes).
Most of us would love to use a viable and convenient mass transit system, and thus use as little fuel as you are able, but it's simply not practical.
The big problem is that (most) US cities are built in a spread out fashion. They sprawl. That means that mass transit is disadvantaged because that works best when serving greater concentrations of people, though the neighborhoods themselves can be dispersed as driving (or running trains/trams) between them is fairly efficient. European cities tend to be more compact (though the US has more super-tall buildings, median building height is greater in Europe) in part because cars became affordable later and in part because land values are higher; this has resulted in a different structure of city layout, and that's pretty slow to change.
Change in the US will be very painful (though it's not as bad as you make out).
(Obama only makes $400k, CEOs making that typically oversee less then 1% of the Fed $Trillion budget)
There's a lot of incidental benefits to that job though. For example, while he's in office he never has to pay for a hotel room in DC for himself or his family, and the furnishing of those rooms will be done at the public expense with a substantial budget available to do it. Yes, it sounds silly but that sort of thing adds up.
In general though, the big difference between the public and private sectors is a different balance between pay and benefits. The private sector tends to be more focussed on pay, the public more on non-pay benefits. This is true in many countries.
Corporate America loves to outsource. Not because it's efficient or cheap, but because it provides someone to blame!
Outsource the network to one firm, the generator to another, the HVAC to a third. Hire temp contract lackeys to staff the place, and rent-a-cops to "guard" it. Then, when something goes wrong, blame them. If it's a big enough issue fire them and replace them with the next batch of people who won't be trained, won't care, and will eventually screw up.
They're forgetting that the one thing they cannot outsource is the overall responsibility for having things working enough to support their business, for if they get rid of that then they've eliminated the need for them to exist at all (and their supplier will simply cut them out of the equation with no ill-consequences). If things keep failing horribly because the people they're outsourcing to suck, it's Corporate America's fault for outsourcing to the wrong people (or outsourcing at all).
Mind you, it might be better to deal with problems through conventional insurance than trying to make the system infallible.
Is that carbon neutral? Why haven't I seen cars that run on ego?
The conversion rate of ego to useful energy is poor. Luckily, the team have Branson about so they've got near-infinite amounts of ego on site, but redistributing to the rest of the world is impractical. (If we could substitute ego for hyperbole and hot air, we'd be able to make some use of politicians...)
If I need high availability, I would use Rackspace, Amazon, Azure, and I'd ensure that I have a plan to deal with a major outage with any of the providers. Each have APIs,
Unless the API is proprietary (or just non standard) and the cloud operator introduces some systemic fault* into their services. What then?
Warm up the lawyers.
Seriously, if you've got a critical part of your business based on a single supplier, you're vulnerable to problems and so should have some other means of protecting yourself. Cloud computing is just yet another way that this can happen, but it's not particularly special in this regard. At the commoditized service end (i.e., IaaS) you can protect yourself by having multiple suppliers. For a specialized service, you have fewer options (but can potentially gain more benefit from that service, of course).
[...] we're talking about more than half a billion per year. [...] If it saves all 200 lives and all 17,000 injuries, that would be a good price.
Not really. That's $2.5M per life minimum. OK, you can probably also factor in all those injuries as well which improves things, but you're unlikely to hit that target anyway. (After all, some of those deaths/injuries will be actually caused by inattention, not inability to see, though the stats won't be able to pick that up.) It would really help to know what actuaries think of the value of these devices, as they've got a good idea what the real average value of a human life is (sad that we need such a concept, but we do) and on the effectiveness of the measure.
'cloud solutions for customers on their own premises'
You keep using that word. I don't think it means what you think it means....
An awful lot of larger businesses are internally structured as a group of smaller businesses, with "contracts" between business units. (Yes, they're not formally contracts, but the main difference on a practical level is that it is the CEO of the overall company who is the ultimate decider of last resort, not the law.) Within such a setting, an "internal" cloud can still make a lot of sense and the technologies used to implement a cloud are highly relevant in any case. It's very common to have needs that are well met by a virtualized server slice or a segregated piece of diskstore.
That said, Sturgeon's Law applies thoroughly to the cloud, and the 90% includes all the marketing and corporate statements.
Primaries themselves are a dirty trick on democracy.
I thought that primaries were a technique to allow parties to choose their candidate for an election. Ultimately, a party should have the right to choose their candidate in any way they want, though it is presumably advisable for them to choose someone that they think will win the election. OTOH, what happens all too often is that a party instead chooses someone who represents what their activists want most, so leading to total failure in the wider election. In the UK, this was particularly observed in the Labour party at the start of the 80s; they picked candidates who were far too left wing, and so weren't in power until the latter half of the 90s. Looks like this is happening in the US with the Reps now (except they're picking candidates who are too right wing); the activists are voting for irrelevance and they won't/can't believe it either.
Speaking as someone who doesn't agree with the Reps on many things, make popcorn on election night!
I'll try to explain it simply: The government finances scientific research, with tax money. That research is conducted by scientists, then other sicentists review the research for flaws, and finally the research is published in scientific journals. Elsevier is the editor of several such journals.
Publisher, not editor.
Elsevier and other publishers do not pay scientists who do research and they do not pay the scientists who review the original research either. They don't pay anybody. They only pay the publication of journals (i.e. printing). And then they sell those journals for a very, very expensive sum of money. I don't have the prices, but it's so expensive that only universities buy these journals (even public libraries can't afford them). So Elsevier and other publishers like them make a ton of money through the work of others.
On the other hand, it does cost quite a bit to print a journal. It's on high-quality paper, not the normal rubbish, and it's done with a very high quality print mechanism. A copy of a journal is supposed to be able to survive for at least a century with only minimal effort at maintenance. A journal paper is for life (and beyond), not just for Christmas.
Which isn't to say that the current situation is right either. Too many rights have been signed over in the past. Prices are not necessarily right either (though long-term preservation of data is also surprisingly expensive if you're actually serious).
Now get this: The government wanted research that it finances to be available to the public. Your tax money pays for research, therefore you should have access to that research - makes sense, right? Well Elsevier had a problem with that. Publishers such as them try to keep the research for themselves, in order to force universities and public services to buy their journals. So Elsevier pushed the Research Works Act. As the summary says, this act would make it illegal for the government to say "we'll pay for this research, but on the condition that the results are made public". Yes, I know how crazy it sounds but no, there's no mistake. It's like you paying an artist to make a painting, and then being forced to pay a publisher (on top of the artist) in order to receive the painting.
But getting access to the work isn't free. It might be very cheap, it might be paid by someone else, but it's not going to be free. For some works of art, it would be the costs involved in getting to see it that would dominate (there's lots of public art in NYC, but it's not very cheap to access from Colorado). For others, the reproduction costs might be covered through your internet access charges or through advertisements.
For science, access is somewhat different. In particular, there's three key parts: the data, the method and the results. There's a habit of making both method and results available through journals (which isn't a bad method, though it is slow) and conferences (quicker, far more ephemeral) that have been peer-reviewed (a good thing!) Now you also see people putting what they do online, but usually then without peer review (discounting the writing of nasty comments in Twitter) so there's very little filtering at all. The problem with a lack of filtering is that it means there's no inclination for people to actually prove their argument; research shouldn't be a bunch of advertorials!
The third part is the data. We're still learning how to effectively put that online. (Just piling it up in a big heap on a disk with no backups and piss-poor search is not a good way, and Google aren't good at indexing this sort of thing as it isn't link-heavy.)
[...]
And why are expensive scientific journals an issue now? Because of the Internet and advances in computers. Before, these journals had to be printed. If you haven't seen these journals before, trust me, they're huge and there's usually a new one to print each mont
Headlight glare is a real issue for me here in the UK where presumably our headlights already meet EU standards.
It seems to me that things have got worse (in the UK; not enough experience with night driving elsewhere to comment there) over the past decade or so, like manufacturers have focussed on making the lights brighter without thinking enough about where that light is actually going. But I've no objective evidence, just my fallible memories and possibly-changing eyesight.
Unlimited access to internet is much cheaper than university administrative fees for such things as, you know, internet connection.Your concept of "free" is quite distorted.
The fees won't just be paying for the connection. There's a whole bunch of services too, including not just email but also things like having people around to help when things go wrong. That really costs and the value is invisible until you really need it.
But it's their connection! Therefore, they are exempt from all criticism and he should do nothing if he disagrees with their policies.
It's the university's connection, not the particular administrator of the IT department of the university's connection. That said, if it's an order that's come down from On High (it could be) then it's going to be hard to change since admins don't usually like to directly challenge nonsense from formal superiors.
Another thing to check is whether the restriction is the same for all connections. We (speaking as someone who works in university IT) have a whitelisted wired network with very few restrictions (SMTP being the main one, for obvious reasons) and multiple wireless networks, all of which require authentication of some form in order to let you onto the majority of the internet. (Some of the wireless nets direct via a webpage form, others use a cryptographic identity as that's much more convenient for logging in with mobile devices.) Our main requirement overall is that only authorized users use the network (we've no problem with temporary auth for visitors) as that ensures that people are aware that they've got to obey the AUP (summary: don't break the law, don't be a total dick). As long as they're known, we don't care what they do (and for sure couldn't log it! We're really short of storage space.)
Slashdot Mirror
And if you dont run it thru the debugger and STEP thru it you are just guessing what it will do.
Your inexperience is showing. Quite apart from the fact that you should make your assumptions explicit in your code (if the argument must not be null, put code in to check it!) you have the problem that there's a lot of code where attaching a debugger is impossible because it runs in a tricky environment. You can try to test by putting the code in an environment mostly populated by mocked components, but you can't be sure that you've got things right in such a situation: more often than not, the problem comes when components undergo a subtle change unannounced. (I had to deal with such a problem a few weeks ago. Everything worked within itself, but failed to function correctly when integrated; the problem was a change to the encoding of one of the files used in the interface itself. The whole shebang was rather security-aware too, so debugging was super-difficult.)
Many time I step thru my code to find some assumption I was making that is invalid.
You can write code that compiles with 0 warnings on the highest levels, can get thru the most stringent of lint checks, passed dozens of code reviews, pair wise coded, etc, etc etc. But until you run it and step thru and see you will never know.
Single-stepping is crappy way to debug code. Far better to have a consistent fundamental model and prove that your code adheres to it. It's a technique that tends towards the mathematical/logical end of things, but it works far better than simply trying to step through in that it can catch problems where required code is missing. (OTOH, stepping can help with cases where you've got a really nasty abstraction-busting bug like an endian-specific stack smash. I've seen some weird stuff in my time, and I now mostly pick my languages to avoid the worst disasters.)
Best complement I get from fellow engieres "your code is easy to read and when it screws up it does it in a way I can tell what is wrong"
My favorite is "I can't see how else you could possibly do it" when I know there's loads of ways which are all worse; means that the model I've picked is actively guiding people away from the dangers.
A hot TSA agent is like a cold blue-giant star. It just doesn't happen.
It depends on what level the AC is set to. Some airports are really annoyingly warm, especially in summer, but spring and autumn are worse because then I'm likely to have warm clothing with me (I needed it when I left home and can't put it in my luggage until I get to my final destination).
Oh, you meant the other meaning of "hot". Never mind.
Pretty sure the UK doesn't have a constitution.
It does, but not a consolidated one like the US has and constitutionality is genuinely a matter for experts only.
That said, one of the general requirements on the authorities (including both the police and social services) is that the actions they take are proportionate to the type of risk that they are intended to prevent. This is a very far-reaching principle that enormously constrains the power of the state to act in an arbitrary fashion. I believe there's also a right for action to be taken reasonably promptly; dragging things out leaves the state open to having to pay compensation (plus the legal costs of the challenge; UK courts usually award costs provided those costs are themselves reasonable and proportionate). That all adds up to the authorities having to be able to show that this was the right action to take (given the facts, which we don't have) and to having to get on with the investigation or drop the order.
So how do you deal with servers that have more than one purpose?
If you've got too many purposes on a single machine, that's an indication of potential problems anyway. Split up into separate VMs (with their own names) and the physical machine then has a clear purpose: hosting VMs.
At least in Europe, software patents aren't valid.
You are wrong. Software patents are valid in Europe, but they've got to show that they're doing something genuinely new. If someone's doing something that's really 10–20 years ahead, it's really not a problem. The problem is when someone's getting a patent for work that is only 2 months ahead (or worse, years behind). Those sorts of patents (whether for software or otherwise) are a real blight as they end up as combination of a blight on the state of the art and a bunch of landmines.
(What's the collective noun for such a thing? A blighted field of landmines?)
I would worry if important projects with large budgets and generous timeframes switched from Java to e.g. Ruby, but this won't happen.
You're more likely to see large projects use multiple languages (e.g., Java to make components, JRuby to stitch them together). This makes a lot of sense because it is using languages for the purposes they were designed for; Building high-level apps in just low-level languages is a PITA because of the amount of work involved, and building low-level apps with high-level languages is a very peculiar kind of crazy, but using things for their strengths is just smart. By using such combinations, large projects can achieve much more than they would otherwise (or what would have been a large project could be brought within the grasp of a small team).
I suspect that the real story of programming languages over the past few decades has been the slow death of monolingualism. These days, I only really see it with noobs and diehard C++/Lisp aficionados.
"Capsicum" is the British/Australian English term for the fruit known in American as the bell pepper.
That's because it's the Latin name of the genus including bell peppers and chili peppers.
he'd be in charge of dupe-checking and ensuring all WaPo blog blurbs are high quality and accurate
So, the quality of the WaPo is going to go up a lot...
I suspect the porn company is not liability limited and probably has lots of cash. Sue them, and let them sort it out with Flickr.
Sue both of them. It should also be possible to work out some kind of libel claim in there ("They were alleging I was a copyright pirate, your honour, yet I have enormous respect for all forms of IP as it is a key part of how I earn my living!") Now, Gorman's a UK citizen and is resident in the UK, so using a UK court to pursue a libel claim is A-OK. Heh heh...
A grand a year is assuming a class of at least about 250 students and all of them buying the book (instead of borrowing it from the library or sharing). And that's for at least a month's worth of work - at the speed most lecturers write it's 2-3 months. Hardly a good return on investment.
I've been involved with writing a book (admittedly a tech book, not a college book) and given the sheer amount of work involved, I surely was not doing it to get rich. Heck, I wasn't doing to pay my yearly beer bill and I'm a light drinker. (It's certainly nowhere close to what I spend on caffeine!) It's only a tiny proportion of people who can ever earn a basic living from writing books; I don't appear to be one of them either.
And the grand a year would not last all that long. Books go obsolete and students trade them. Preventing that requires more work, worsening the rate of return...
I saw an analysis of their Terms of Service somewhere, indicating that they will only compensate up to the value of the service paid. So, if your service was $100/mo, they'd only compensate you for the downtime you experienced, or up to that month's service charge of $100.
But such limitations would not necessarily apply if they (or some of their employees) were engaging in malfeasance. Criminal activity by a trusted party changes the game quite substantially. Was this the case this time? I have no idea, but <hypothentically> if an employee were to see that someone is using a server for BTC mining and was to use this knowledge plus their own privileges to steal, it would leave the firm at least partially on the hook. </hypothetically> Whether or not a system is correctly and sanely used and adequately protected doesn't make it right to steal from it, though it might affect compensation levels (and whether any insurers involved give you anything; they usually include clauses to require you to take proper steps to protect your property).
The best applications on the early internet were about connecting people to each other.
That's changed? Not the way I see it. Yes, the way in which people connect to each other may well have changed, but the fundamental thing that wins is still there: connecting people together.
Personally I think it is better if people without an axe to grind work out how to use taxes in a socially beneficial way and politicians only get to vote on it.
I understand what you are saying, but you're completely off-base there. What to tax and what to spend the tax income on are right at the very core of politics. This has been the case for as long as there have been governments at all. It predates the US. It predates democracy. It predates money. It might predate cities (hard to say for sure; records that far back are sketchy but the earliest surviving records appear to be for tax purposes).
Most of us would love to use a viable and convenient mass transit system, and thus use as little fuel as you are able, but it's simply not practical.
The big problem is that (most) US cities are built in a spread out fashion. They sprawl. That means that mass transit is disadvantaged because that works best when serving greater concentrations of people, though the neighborhoods themselves can be dispersed as driving (or running trains/trams) between them is fairly efficient. European cities tend to be more compact (though the US has more super-tall buildings, median building height is greater in Europe) in part because cars became affordable later and in part because land values are higher; this has resulted in a different structure of city layout, and that's pretty slow to change.
Change in the US will be very painful (though it's not as bad as you make out).
(Obama only makes $400k, CEOs making that typically oversee less then 1% of the Fed $Trillion budget)
There's a lot of incidental benefits to that job though. For example, while he's in office he never has to pay for a hotel room in DC for himself or his family, and the furnishing of those rooms will be done at the public expense with a substantial budget available to do it. Yes, it sounds silly but that sort of thing adds up.
In general though, the big difference between the public and private sectors is a different balance between pay and benefits. The private sector tends to be more focussed on pay, the public more on non-pay benefits. This is true in many countries.
Corporate America loves to outsource. Not because it's efficient or cheap, but because it provides someone to blame!
Outsource the network to one firm, the generator to another, the HVAC to a third. Hire temp contract lackeys to staff the place, and rent-a-cops to "guard" it. Then, when something goes wrong, blame them. If it's a big enough issue fire them and replace them with the next batch of people who won't be trained, won't care, and will eventually screw up.
They're forgetting that the one thing they cannot outsource is the overall responsibility for having things working enough to support their business, for if they get rid of that then they've eliminated the need for them to exist at all (and their supplier will simply cut them out of the equation with no ill-consequences). If things keep failing horribly because the people they're outsourcing to suck, it's Corporate America's fault for outsourcing to the wrong people (or outsourcing at all).
Mind you, it might be better to deal with problems through conventional insurance than trying to make the system infallible.
Is that carbon neutral? Why haven't I seen cars that run on ego?
The conversion rate of ego to useful energy is poor. Luckily, the team have Branson about so they've got near-infinite amounts of ego on site, but redistributing to the rest of the world is impractical. (If we could substitute ego for hyperbole and hot air, we'd be able to make some use of politicians...)
If I need high availability, I would use Rackspace, Amazon, Azure, and I'd ensure that I have a plan to deal with a major outage with any of the providers. Each have APIs,
Unless the API is proprietary (or just non standard) and the cloud operator introduces some systemic fault* into their services. What then?
Warm up the lawyers.
Seriously, if you've got a critical part of your business based on a single supplier, you're vulnerable to problems and so should have some other means of protecting yourself. Cloud computing is just yet another way that this can happen, but it's not particularly special in this regard. At the commoditized service end (i.e., IaaS) you can protect yourself by having multiple suppliers. For a specialized service, you have fewer options (but can potentially gain more benefit from that service, of course).
[...] we're talking about more than half a billion per year. [...] If it saves all 200 lives and all 17,000 injuries, that would be a good price.
Not really. That's $2.5M per life minimum. OK, you can probably also factor in all those injuries as well which improves things, but you're unlikely to hit that target anyway. (After all, some of those deaths/injuries will be actually caused by inattention, not inability to see, though the stats won't be able to pick that up.) It would really help to know what actuaries think of the value of these devices, as they've got a good idea what the real average value of a human life is (sad that we need such a concept, but we do) and on the effectiveness of the measure.
'cloud solutions for customers on their own premises'
You keep using that word. I don't think it means what you think it means....
An awful lot of larger businesses are internally structured as a group of smaller businesses, with "contracts" between business units. (Yes, they're not formally contracts, but the main difference on a practical level is that it is the CEO of the overall company who is the ultimate decider of last resort, not the law.) Within such a setting, an "internal" cloud can still make a lot of sense and the technologies used to implement a cloud are highly relevant in any case. It's very common to have needs that are well met by a virtualized server slice or a segregated piece of diskstore.
That said, Sturgeon's Law applies thoroughly to the cloud, and the 90% includes all the marketing and corporate statements.
Primaries themselves are a dirty trick on democracy.
I thought that primaries were a technique to allow parties to choose their candidate for an election. Ultimately, a party should have the right to choose their candidate in any way they want, though it is presumably advisable for them to choose someone that they think will win the election. OTOH, what happens all too often is that a party instead chooses someone who represents what their activists want most, so leading to total failure in the wider election. In the UK, this was particularly observed in the Labour party at the start of the 80s; they picked candidates who were far too left wing, and so weren't in power until the latter half of the 90s. Looks like this is happening in the US with the Reps now (except they're picking candidates who are too right wing); the activists are voting for irrelevance and they won't/can't believe it either.
Speaking as someone who doesn't agree with the Reps on many things, make popcorn on election night!
I'll try to explain it simply:
The government finances scientific research, with tax money.
That research is conducted by scientists, then other sicentists review the research for flaws, and finally the research is published in scientific journals. Elsevier is the editor of several such journals.
Publisher, not editor.
Elsevier and other publishers do not pay scientists who do research and they do not pay the scientists who review the original research either. They don't pay anybody. They only pay the publication of journals (i.e. printing). And then they sell those journals for a very, very expensive sum of money. I don't have the prices, but it's so expensive that only universities buy these journals (even public libraries can't afford them).
So Elsevier and other publishers like them make a ton of money through the work of others.
On the other hand, it does cost quite a bit to print a journal. It's on high-quality paper, not the normal rubbish, and it's done with a very high quality print mechanism. A copy of a journal is supposed to be able to survive for at least a century with only minimal effort at maintenance. A journal paper is for life (and beyond), not just for Christmas.
Which isn't to say that the current situation is right either. Too many rights have been signed over in the past. Prices are not necessarily right either (though long-term preservation of data is also surprisingly expensive if you're actually serious).
Now get this:
The government wanted research that it finances to be available to the public. Your tax money pays for research, therefore you should have access to that research - makes sense, right?
Well Elsevier had a problem with that. Publishers such as them try to keep the research for themselves, in order to force universities and public services to buy their journals. So Elsevier pushed the Research Works Act.
As the summary says, this act would make it illegal for the government to say "we'll pay for this research, but on the condition that the results are made public". Yes, I know how crazy it sounds but no, there's no mistake.
It's like you paying an artist to make a painting, and then being forced to pay a publisher (on top of the artist) in order to receive the painting.
But getting access to the work isn't free. It might be very cheap, it might be paid by someone else, but it's not going to be free. For some works of art, it would be the costs involved in getting to see it that would dominate (there's lots of public art in NYC, but it's not very cheap to access from Colorado). For others, the reproduction costs might be covered through your internet access charges or through advertisements.
For science, access is somewhat different. In particular, there's three key parts: the data, the method and the results. There's a habit of making both method and results available through journals (which isn't a bad method, though it is slow) and conferences (quicker, far more ephemeral) that have been peer-reviewed (a good thing!) Now you also see people putting what they do online, but usually then without peer review (discounting the writing of nasty comments in Twitter) so there's very little filtering at all. The problem with a lack of filtering is that it means there's no inclination for people to actually prove their argument; research shouldn't be a bunch of advertorials!
The third part is the data. We're still learning how to effectively put that online. (Just piling it up in a big heap on a disk with no backups and piss-poor search is not a good way, and Google aren't good at indexing this sort of thing as it isn't link-heavy.)
[...]
And why are expensive scientific journals an issue now? Because of the Internet and advances in computers. Before, these journals had to be printed. If you haven't seen these journals before, trust me, they're huge and there's usually a new one to print each mont
Headlight glare is a real issue for me here in the UK where presumably our headlights already meet EU standards.
It seems to me that things have got worse (in the UK; not enough experience with night driving elsewhere to comment there) over the past decade or so, like manufacturers have focussed on making the lights brighter without thinking enough about where that light is actually going. But I've no objective evidence, just my fallible memories and possibly-changing eyesight.
Unlimited access to internet is much cheaper than university administrative fees for such things as, you know, internet connection.Your concept of "free" is quite distorted.
The fees won't just be paying for the connection. There's a whole bunch of services too, including not just email but also things like having people around to help when things go wrong. That really costs and the value is invisible until you really need it.
But it's their connection! Therefore, they are exempt from all criticism and he should do nothing if he disagrees with their policies.
It's the university's connection, not the particular administrator of the IT department of the university's connection. That said, if it's an order that's come down from On High (it could be) then it's going to be hard to change since admins don't usually like to directly challenge nonsense from formal superiors.
Another thing to check is whether the restriction is the same for all connections. We (speaking as someone who works in university IT) have a whitelisted wired network with very few restrictions (SMTP being the main one, for obvious reasons) and multiple wireless networks, all of which require authentication of some form in order to let you onto the majority of the internet. (Some of the wireless nets direct via a webpage form, others use a cryptographic identity as that's much more convenient for logging in with mobile devices.) Our main requirement overall is that only authorized users use the network (we've no problem with temporary auth for visitors) as that ensures that people are aware that they've got to obey the AUP (summary: don't break the law, don't be a total dick). As long as they're known, we don't care what they do (and for sure couldn't log it! We're really short of storage space.)