When I was at school (in the UK), I'd get through 20-60 rounds a week on the range. Only during term time, so probably about 1-2K/year. This was mostly bolt-action target rifle ammunition, in a scenario where accuracy is the most important and speed is not a concern. When firing things like assault rifles and lever-action carbines I could easily shoot 5-10 times as much ammunition in the same amount of time. If someone has a job where being able to shoot both quickly and accurately is a requirement, I'd expect them to be practicing for more time than I was and using more ammunition per practice session, so 10K rounds per year seems like an absolute lower bound, and probably an underestimation by an order of magnitude. For 60K employees, that works out at a minimum of 600M rounds just for practice, but presumably some are going to be using other kinds of ammunition and some are not going to be armed (DHS secretaries and accountants probably don't spend much time on the range...)
The RPi is now producing 4,000 a week and shipping as many as are coming off the production line. Availability can be a bit spotty, but I had someone drop four on my desk yesterday, so they're definitely around...
The point of the RPi seems to be missed on many of the 'competitors' though. It's cheap, has a load of GPIO pins, and runs a general purpose OS. It's designed as a modern BBC Micro - something that schools can use to teach programming and can use to control things like robots, not as a low-end Linux computer. The fact that it runs Linux is entirely incidental (and, actually, a compromise: the original plan was to have it boot directly into a REPL programming environment, just like the BBC).
I just set up a map using OpenLayers for an event, with OSM data. I didn't have to register anything, I didn't have to create an account, and I didn't have to host anything myself. And, unlike the Google map of the same area, the OSM data contains accurate cycle paths for where the visitors are likely to be going and has all of the buildings correctly labelled. So, uh, why would I use Google Maps?
Oh, and if I only wanted to place a single pin in a map (your 'how to find us' example), then this is trivial without even knowing any JavaScript. The export tab on the OSM front page will generate the HTML for an iframe that you can just paste into your web page.
If the tables were reversed and a female programmer declared a function called insanely_large_dick() do you think any males would throw up their hands in the air and quit their jobs in IT because they felt threatened and/or harassed?
This argument has been made a couple of times, but it's misleading. For a function, I would certainly expect her to be reprimanded because function names are expected to be explanatory. Magic constants, in contrast, just have to be recognisable. 0x81680085 is easily recognisable and memorable. Picking an 8-digit hex number that meets those requirements is hard. 0xDEADBEEF, which Microsoft also uses, is probably the best example, because it immediately jumps out of a hex dump.
As a vegetarian, I am offended that Microsoft's implementation marks free'd memory with 0xDEADBEEF! I demand that this be changed to something that won't offend me! Or Hindus! Or people with irrational dislike of certain numbers when represented in hex!
Also, if conditional prefixes were so great, why's ARM64 eliminating them.
ARMv8 is not eliminating them, it's reducing the number of instructions that have them. Conditional instructions are useful because you can eliminate branches and so keep the pipeline full. For example, consider this contrived example:
if (a < b) a++;
On ARMv7 and earlier, this would be a conditional add. The pipeline would always be full, the add would always be executed, but the result would only be retired if the condition is true. On MIPS, it would be a branch (complete with the insanity known as branch delay slots, which if you look at the diassembly of most MIPS code typically means with a nop, so you get to waste some i-cache as well) and if it's mispredicted then you get a pipeline stall.
On ARMv8, you don't have a conditional add, but you do have a conditional register-register move and you have twice as many registers. The compiler would still issue the add instruction and then would do a conditional move to put it in the result register. From the compiler perspective, this means that you can lower PHI nodes from your SSA representation directly to conditional moves in a lot of cases.
Basically, 32-bit ARM is designed for assembly writers, ARMv8 is designed for compilers. As a compiler writer, it's hands-down the best ISA I've worked with, although I would prefer to write assembly by hand for ARMv7. I wouldn't want to do either with MIPS, although I currently am working on MIPS-based CPU with some extra extensions.
One disadvantage MIPS has is that it has a completely batshit insane architecture full of things that kind-of vaguely made sense in a research project and have been a liability for over a decade.
I'm currently working the MIPS back end in LLVM and every day brings a new WTF. ARM, at least, has a mostly sane architecture and in the 64-bit variant has removed a load of stuff that doesn't make sense with newer pipeline designs (e.g. store multiple, conditional adds, and so on). MIPS just accumulates legacy cruft. It's not quite as bad as x86 in that regard, but it's also an astonishingly verbose instruction set. Where ARM or x86 can do something in 2-3 instructions, MIPS is lucky to manage in half a dozen. It's not a coincidence that the decline of MIPS on the high end coincided with instruction cache utilisation being important. Oh, and every MIPS vendor adds some extra incompatible changes to the instruction set just to make life even more fun.
It's not so much consumer electronics. The main place for Contiki is sensor networks. For example, consider a building with electricity usage and IR / motion sensors in every light fitting. You may have a few thousand sensors in a single building, and you want to make them as cheap as possible.
The problem that Contiki will have in the next few years is that the price of the IC is increasingly limited by the cost of the package at the low end. This means that in a few years you'll be able to buy a 100MHz ARMv7 core with 1MB of RAM and Flash on die for approximately the same cost as a cheap microcontroller. At this point, you may as well use something like a slimmed-down BSD kernel than an embedded OS.
Last time I was in the US, it was very easy to find SIM-only plans that were considerably cheaper than the equivalent with a 'free' phone. There was an article in the NYT a couple of years ago that compared them and came to the conclusion that the best value plan with a 'free' phone was effectively a loan with a 20% APR. The worst value ones were over 100% APR. In short, it was cheaper to buy the phone on your credit card, only pay the interest every month, and get a SIM-only deal, get no one would ever suggest doing that seriously because it's obvious that it's a stupidly expensive way of buying a phone.
The marketing department of Apple, Inc. did not need to "sell" their wares as much as their peers in other companies (like Nokia or RIM, for example), as Mr. Jobs himself had done most of the selling.
Ah, that must be why Apple posters seem to be everywhere, if you turned on a TV any time in the last five years you had a good chance of seeing at least one Apple advert, and every major film in the last decade or so has had gratuitous Apple product placement.
Ever watched a Formula 1 race where someone died? I only remember two, and both interrupted the race and had a load of depressing safety committees after them. On the other hand, a spectacular crash that the driver then walks away from makes great television...
There are a few other issues. Astronaut training is expensive. The figure I saw was about $4m. Even if the astronaut is willing to work for free and there is an inexhaustible supply of potential replacements, the cost (and time) of getting a new applicant up to standard is significant. The other factor is PR. It's really really bad publicity if the public sees astronauts die in fireballs. The USSR worked around this by only admitting to successful missions: any that failed were never publicly announced. That's a lot more difficult for NASA to do...
Are you basing that on some actual data, or just the whining of people on Slashdot? I know someone who had his third delivered a couple of weeks ago, and he didn't order it until after his first had arrived, so the delays can't be that huge...
It depends on the supermarket. When I go and visit my mother, for example, two of the chains have local supermarkets, but four will deliver to her house. The other two are have a warehouse about an hour's drive away. Their drivers leave in the morning with a van load of boxes for delivery, on a route that takes them through a load of small villages and a few big towns.
As to the jobs, fewer are needed in this scenario. For a real supermarket, you have people moving things from the warehouse shelves to the customer-facing shelves, people standing behind food counters and serving individuals, and people at the checkouts. These are all replaced by - fewer - people who just take things off the warehouse shelves and put them in boxes and then others that deliver them.
And, while mobile, OSMAnd (also using OSM data) does a good job of providing cycling directions. Not sure about the rest of the UK, but in Cambridge the OSM data on cycle maps is vastly better than the Google data. It's a shame Google doesn't contribute to OSM rather than building their own inferior proprietary dataset.
For grocery stores, if you think shopping online is going to replace the local supermarket, you're insane and need psychiatric intervention. How are you going to buy ice cream online? Unless they do some really fast delivery, it'll be melted.
I don't know about where you live, but all of the major supermarkets in the UK provide local delivery, including Asda (Walmart). And, yes, you can order icecream and it works fine. They ship everything in vans that have a freezer compartment and the frozen items arrive significantly colder than if you'd taken them home in a car. I've done pretty much all of my supermarket shopping online for the last 5 years or so, and I know a lot of other people who do as well. The delivery charges are as low as £3, which may be slightly more expensive than driving to the shop and back, but it isn't if you value your time: you can do the entire shop online in less time than it takes most people to get to the supermarket.
I may be overly pessimistic, but I suspect that a lot of people would object to something that is effectively a scramjet flying at an altitude of under a metre...
I've not seen American Ninja Warrior, but the vast majority of American remakes of popular import shows get the same treatment. First, you start of with something that's popular. Then you assume that Americans are too stupid to understand any culture other than the perfectly homogenised average of the USA as portrayed by Hollywood. Then you remove everything that doesn't conform to this. And then you end up with something that even the target audience thinks is dumb.
There's also the localisation issue. What TV standard do you support? Presumably something digital, but is that DVB-t, used in most of the world, ASTC, used in North America, ISDB-T, used in most of South America, or DTMB, used in China? Do you include support for all of them? And then when you sell in a country like the UK that has a TV license, you effectively tack that onto the price tag if the person doesn't already have a TV.
Don't fall into that trap. Who appoints them? I don't know
And there's the reason why it's not worth reading the rest of your post. Why not educate yourself? The European Commission is appointed by the European Council, which is comprised of the heads of state of the various members. The indirection between them and the people you elect is huge. In the UK, for example, you vote for a Member of Parliament (MP). The party with the most MPs selects the Prime Minister (modulo coalitions), who is then the UK representative in the European Council. He, along with the other members, is responsible for appointing the members of the European Commission. So, my influence on the Commission is that I vote for someone who may have a vote for the person that has a vote to appoint the person who is supposed to represent me. In contrast, I have 4 MEPs who are supposed to represent me, of whom one is someone I respect and who I can rely on to act in my interests and the others presumably act in the interests of other members of my constituency.
Universities are not money sinks because they are (mostly) publicly funded with these purposes:
Creating an educated pool of researchers.
Creating an educated pool of knowledgeable workers.
Creating a pool of knowledge for the general good.
This is not the case with commercial research. Commercial research institutions fall into two categories. There are companies like SRI that are contracted by organisations like DARPA and sometimes by big companies to look at specific problems. These are evaluated by how happy their customers are with their work. Then there are in-house research groups like MSR, IBM Research and PARC. These exist to produce the seed research that the rest of the company can then turn into products. They are evaluated based on their knowledge transfer to the rest of the organisation.
PARC was, by all accounts, a huge success. It did blue-sky research, much of which was nonsense, but enough of which produced good results that it completely funded the rest of the research. And, of course, if you could identify which 10% would be successful in advance, you'd be doing development not research, so the failures are just the cost of doing business for a research institution (although PARC had an astonishingly high hit rate).
MSR is doing very well at producing ideas (and it's a fun place to hang out, full of bright people working on interesting projects), but MS really sucks at pulling ideas in from there. You're more likely to see an idea from a university or a startup being used as the basis for something in a Microsoft product than something from MSR.
Oh, and if you look at where people go when they leave MSR, it's usually into universities or other commercial research labs. Very few of them go into MS, so you can't even justify the investment as training.
Part of the problem is that MSR sites are often geographically separate from the rest of MS, so there's no simple way for ideas to flow from one to the other. MS could gain a lot by rotating their research staff through their product groups for a couple of weeks each year (or sending their engineers on sabbatical to MSR periodically).
The problem is that most OS X upgrades contain a number of usability regressions and half a dozen really useful features. It's always (well, except with 10.6) a bit difficult to decide whether the benefits outweigh the pain.
Slashdot Mirror
Un, no. Pay-as-you-go means pre-pay. It's the exact opposite of receiving a bill at the end of the month.
When I was at school (in the UK), I'd get through 20-60 rounds a week on the range. Only during term time, so probably about 1-2K/year. This was mostly bolt-action target rifle ammunition, in a scenario where accuracy is the most important and speed is not a concern. When firing things like assault rifles and lever-action carbines I could easily shoot 5-10 times as much ammunition in the same amount of time. If someone has a job where being able to shoot both quickly and accurately is a requirement, I'd expect them to be practicing for more time than I was and using more ammunition per practice session, so 10K rounds per year seems like an absolute lower bound, and probably an underestimation by an order of magnitude. For 60K employees, that works out at a minimum of 600M rounds just for practice, but presumably some are going to be using other kinds of ammunition and some are not going to be armed (DHS secretaries and accountants probably don't spend much time on the range...)
The RPi is now producing 4,000 a week and shipping as many as are coming off the production line. Availability can be a bit spotty, but I had someone drop four on my desk yesterday, so they're definitely around...
The point of the RPi seems to be missed on many of the 'competitors' though. It's cheap, has a load of GPIO pins, and runs a general purpose OS. It's designed as a modern BBC Micro - something that schools can use to teach programming and can use to control things like robots, not as a low-end Linux computer. The fact that it runs Linux is entirely incidental (and, actually, a compromise: the original plan was to have it boot directly into a REPL programming environment, just like the BBC).
I just set up a map using OpenLayers for an event, with OSM data. I didn't have to register anything, I didn't have to create an account, and I didn't have to host anything myself. And, unlike the Google map of the same area, the OSM data contains accurate cycle paths for where the visitors are likely to be going and has all of the buildings correctly labelled. So, uh, why would I use Google Maps?
Oh, and if I only wanted to place a single pin in a map (your 'how to find us' example), then this is trivial without even knowing any JavaScript. The export tab on the OSM front page will generate the HTML for an iframe that you can just paste into your web page.
If the tables were reversed and a female programmer declared a function called insanely_large_dick() do you think any males would throw up their hands in the air and quit their jobs in IT because they felt threatened and/or harassed?
This argument has been made a couple of times, but it's misleading. For a function, I would certainly expect her to be reprimanded because function names are expected to be explanatory. Magic constants, in contrast, just have to be recognisable. 0x81680085 is easily recognisable and memorable. Picking an 8-digit hex number that meets those requirements is hard. 0xDEADBEEF, which Microsoft also uses, is probably the best example, because it immediately jumps out of a hex dump.
As a vegetarian, I am offended that Microsoft's implementation marks free'd memory with 0xDEADBEEF! I demand that this be changed to something that won't offend me! Or Hindus! Or people with irrational dislike of certain numbers when represented in hex!
Also, if conditional prefixes were so great, why's ARM64 eliminating them.
ARMv8 is not eliminating them, it's reducing the number of instructions that have them. Conditional instructions are useful because you can eliminate branches and so keep the pipeline full. For example, consider this contrived example:
On ARMv7 and earlier, this would be a conditional add. The pipeline would always be full, the add would always be executed, but the result would only be retired if the condition is true. On MIPS, it would be a branch (complete with the insanity known as branch delay slots, which if you look at the diassembly of most MIPS code typically means with a nop, so you get to waste some i-cache as well) and if it's mispredicted then you get a pipeline stall.
On ARMv8, you don't have a conditional add, but you do have a conditional register-register move and you have twice as many registers. The compiler would still issue the add instruction and then would do a conditional move to put it in the result register. From the compiler perspective, this means that you can lower PHI nodes from your SSA representation directly to conditional moves in a lot of cases.
Basically, 32-bit ARM is designed for assembly writers, ARMv8 is designed for compilers. As a compiler writer, it's hands-down the best ISA I've worked with, although I would prefer to write assembly by hand for ARMv7. I wouldn't want to do either with MIPS, although I currently am working on MIPS-based CPU with some extra extensions.
One disadvantage MIPS has is that it has a completely batshit insane architecture full of things that kind-of vaguely made sense in a research project and have been a liability for over a decade.
I'm currently working the MIPS back end in LLVM and every day brings a new WTF. ARM, at least, has a mostly sane architecture and in the 64-bit variant has removed a load of stuff that doesn't make sense with newer pipeline designs (e.g. store multiple, conditional adds, and so on). MIPS just accumulates legacy cruft. It's not quite as bad as x86 in that regard, but it's also an astonishingly verbose instruction set. Where ARM or x86 can do something in 2-3 instructions, MIPS is lucky to manage in half a dozen. It's not a coincidence that the decline of MIPS on the high end coincided with instruction cache utilisation being important. Oh, and every MIPS vendor adds some extra incompatible changes to the instruction set just to make life even more fun.
It's not so much consumer electronics. The main place for Contiki is sensor networks. For example, consider a building with electricity usage and IR / motion sensors in every light fitting. You may have a few thousand sensors in a single building, and you want to make them as cheap as possible.
The problem that Contiki will have in the next few years is that the price of the IC is increasingly limited by the cost of the package at the low end. This means that in a few years you'll be able to buy a 100MHz ARMv7 core with 1MB of RAM and Flash on die for approximately the same cost as a cheap microcontroller. At this point, you may as well use something like a slimmed-down BSD kernel than an embedded OS.
Last time I was in the US, it was very easy to find SIM-only plans that were considerably cheaper than the equivalent with a 'free' phone. There was an article in the NYT a couple of years ago that compared them and came to the conclusion that the best value plan with a 'free' phone was effectively a loan with a 20% APR. The worst value ones were over 100% APR. In short, it was cheaper to buy the phone on your credit card, only pay the interest every month, and get a SIM-only deal, get no one would ever suggest doing that seriously because it's obvious that it's a stupidly expensive way of buying a phone.
The marketing department of Apple, Inc. did not need to "sell" their wares as much as their peers in other companies (like Nokia or RIM, for example), as Mr. Jobs himself had done most of the selling.
Ah, that must be why Apple posters seem to be everywhere, if you turned on a TV any time in the last five years you had a good chance of seeing at least one Apple advert, and every major film in the last decade or so has had gratuitous Apple product placement.
Ever watched a Formula 1 race where someone died? I only remember two, and both interrupted the race and had a load of depressing safety committees after them. On the other hand, a spectacular crash that the driver then walks away from makes great television...
There are a few other issues. Astronaut training is expensive. The figure I saw was about $4m. Even if the astronaut is willing to work for free and there is an inexhaustible supply of potential replacements, the cost (and time) of getting a new applicant up to standard is significant. The other factor is PR. It's really really bad publicity if the public sees astronauts die in fireballs. The USSR worked around this by only admitting to successful missions: any that failed were never publicly announced. That's a lot more difficult for NASA to do...
Are you basing that on some actual data, or just the whining of people on Slashdot? I know someone who had his third delivered a couple of weeks ago, and he didn't order it until after his first had arrived, so the delays can't be that huge...
It depends on the supermarket. When I go and visit my mother, for example, two of the chains have local supermarkets, but four will deliver to her house. The other two are have a warehouse about an hour's drive away. Their drivers leave in the morning with a van load of boxes for delivery, on a route that takes them through a load of small villages and a few big towns.
As to the jobs, fewer are needed in this scenario. For a real supermarket, you have people moving things from the warehouse shelves to the customer-facing shelves, people standing behind food counters and serving individuals, and people at the checkouts. These are all replaced by - fewer - people who just take things off the warehouse shelves and put them in boxes and then others that deliver them.
And, while mobile, OSMAnd (also using OSM data) does a good job of providing cycling directions. Not sure about the rest of the UK, but in Cambridge the OSM data on cycle maps is vastly better than the Google data. It's a shame Google doesn't contribute to OSM rather than building their own inferior proprietary dataset.
For grocery stores, if you think shopping online is going to replace the local supermarket, you're insane and need psychiatric intervention. How are you going to buy ice cream online? Unless they do some really fast delivery, it'll be melted.
I don't know about where you live, but all of the major supermarkets in the UK provide local delivery, including Asda (Walmart). And, yes, you can order icecream and it works fine. They ship everything in vans that have a freezer compartment and the frozen items arrive significantly colder than if you'd taken them home in a car. I've done pretty much all of my supermarket shopping online for the last 5 years or so, and I know a lot of other people who do as well. The delivery charges are as low as £3, which may be slightly more expensive than driving to the shop and back, but it isn't if you value your time: you can do the entire shop online in less time than it takes most people to get to the supermarket.
I may be overly pessimistic, but I suspect that a lot of people would object to something that is effectively a scramjet flying at an altitude of under a metre...
A computer is not a toaster. It can't be
Clearly you never used a first-generation Athlon.
I've seen some strange spellings on Slashdot, but spelling IBM c-r-a-y is probably the weirdest.
I've not seen American Ninja Warrior, but the vast majority of American remakes of popular import shows get the same treatment. First, you start of with something that's popular. Then you assume that Americans are too stupid to understand any culture other than the perfectly homogenised average of the USA as portrayed by Hollywood. Then you remove everything that doesn't conform to this. And then you end up with something that even the target audience thinks is dumb.
There's also the localisation issue. What TV standard do you support? Presumably something digital, but is that DVB-t, used in most of the world, ASTC, used in North America, ISDB-T, used in most of South America, or DTMB, used in China? Do you include support for all of them? And then when you sell in a country like the UK that has a TV license, you effectively tack that onto the price tag if the person doesn't already have a TV.
Don't fall into that trap. Who appoints them? I don't know
And there's the reason why it's not worth reading the rest of your post. Why not educate yourself? The European Commission is appointed by the European Council, which is comprised of the heads of state of the various members. The indirection between them and the people you elect is huge. In the UK, for example, you vote for a Member of Parliament (MP). The party with the most MPs selects the Prime Minister (modulo coalitions), who is then the UK representative in the European Council. He, along with the other members, is responsible for appointing the members of the European Commission. So, my influence on the Commission is that I vote for someone who may have a vote for the person that has a vote to appoint the person who is supposed to represent me. In contrast, I have 4 MEPs who are supposed to represent me, of whom one is someone I respect and who I can rely on to act in my interests and the others presumably act in the interests of other members of my constituency.
This is not the case with commercial research. Commercial research institutions fall into two categories. There are companies like SRI that are contracted by organisations like DARPA and sometimes by big companies to look at specific problems. These are evaluated by how happy their customers are with their work. Then there are in-house research groups like MSR, IBM Research and PARC. These exist to produce the seed research that the rest of the company can then turn into products. They are evaluated based on their knowledge transfer to the rest of the organisation.
PARC was, by all accounts, a huge success. It did blue-sky research, much of which was nonsense, but enough of which produced good results that it completely funded the rest of the research. And, of course, if you could identify which 10% would be successful in advance, you'd be doing development not research, so the failures are just the cost of doing business for a research institution (although PARC had an astonishingly high hit rate).
MSR is doing very well at producing ideas (and it's a fun place to hang out, full of bright people working on interesting projects), but MS really sucks at pulling ideas in from there. You're more likely to see an idea from a university or a startup being used as the basis for something in a Microsoft product than something from MSR.
Oh, and if you look at where people go when they leave MSR, it's usually into universities or other commercial research labs. Very few of them go into MS, so you can't even justify the investment as training.
Part of the problem is that MSR sites are often geographically separate from the rest of MS, so there's no simple way for ideas to flow from one to the other. MS could gain a lot by rotating their research staff through their product groups for a couple of weeks each year (or sending their engineers on sabbatical to MSR periodically).
The problem is that most OS X upgrades contain a number of usability regressions and half a dozen really useful features. It's always (well, except with 10.6) a bit difficult to decide whether the benefits outweigh the pain.