Good grief... so you're saying that by using Erlang you need a minimum 20 cores to get a measly 2x speed-up over a single-threaded program, which is probably easier to write too?
Multi-threaded programming is not unique to Erlang, writing parallel code in Java, C#, or even C++ is not exactly rocket surgery, so the comparison to strictly single-threaded code is unfair. I can get a program in any one of those languages to scale up to at least hundreds processors without breaking a sweat. Heck, any web application is automatically multi-threaded with both Java and C#, and with version.NET Framework v4.5 or later it can be trivially made asynchronous as well on top of that! You don't need a single lock or threading primitive of any kind to achieve that. If you want to get more advanced, all three platforms give you easy to use and powerful libraries to make multi-threading safe and efficient.
It's only the crappy open-source platforms written by non-professionals that struggle to scale, which is why Erlang looks so good in comparison to a lot of beginner programmers, and why Node.js is so fascinating to them. Ooo... look... we've invented asynchronous calls, which is brand new and fascinating, because clearly nobody has ever done that before...
The real benefit of Erlang is not speed, but reliability and online maintainability. It's designed for non-stop systems where uptime is more important than performance. This is explicitly stated in its documentation as the primary design goal of the language! Using Erlang to improve the performance of code is asinine. If speed matters, any of the three popular languages will run circles around Erlang.
That fancy 64-processor SGI box of yours? You've lost 90% of its performance to Erlang's inefficiency, so I could get the same performance out of a 6-core computer with an efficient programming language. I have 6 cores in my desktop. I bet it cost a few digits less than your SGI box.
I just did, and the noise is a lot lower, but I can't get the volume as high, and some movies end up too quiet. The only remaining noise is now the cooling fans of my PC, but that's fixable...
Possibly, but I picked this pair of headphones out of over a dozen after an hour of listening tests.
I couldn't find any other headphone in the store (which specialised in nothing other than headphones) that could could go to as low a frequency as these. One of my test tracks was a classical piece with very low frequency double bass notes. I couldn't even hear the instrument at all with most of the others, one other headphone could reproduce it, but all tonality was lost, while with the Sennheiser HD800 I could actually distinguish each note. I compared it to other lower-end Sennheiser models, and a bunch of other brands. It stood out as noticeably better than the others.
I figured it is probably over-priced, but what-the-hey, it costs less than a single speaker for a high-end home audio system, so who cares?
This article is of particular interest to me because just recently I dropped over $1000 on a pair of high-end Sennheiser HD-800 headphones, but now I'm finding the amplifier's background noise is a lot more noticeable. Before, with cheap headphones that didn't have the same dynamic range, it didn't matter, but now it's the limiting factor.
I've got a reasonably decent FLAC collection, some of it classical music in 24/96Khz format, but the background hiss is detracting quite a bit from the potential quality.
What's a good amplifier for headphones that's optimised for a low noise floor instead of power, but isn't over-priced? I don't need a receiver with hundreds of inputs, I need something that takes a single digital input, and outputs the highest possible quality headphone output. Is there anything like that out there?
What I don't understand is why American universities even have campus police?
The last couple of times I've been to an Australian University, including two of the biggest, the closest I've seen were two very bored looking security staff members who spent most of their day making student and staff cards from a tiny little office in the basement. They certainly didn't have guns, and probably didn't even have pepper spray -- none that I could see anyway.
Just how bad is the crime situation on your campuses that you need both a dedicated police force and a security force on top of that?
To a degree, CSIRO here in Australia works like this. Its research is often commercialized and sold, usually via patent licensing. They just recently won $200 million from companies manufacturing WiFi equipment that hadn't properly licensed their patents.
Personally I think that the way a national research organisation should function is that it should give its research away for free to the citizens and corporations of its home country, and charge foreign organisations to use the patents. That way, it would be a socialised boost to the nation's research without the overhead of having to charge people for access, but still self-fund itself to a degree.
Trees are about as effective in doing photosynthesis as any other plant.
That's not true, there's significant differences in efficiency between various species of plants. Most grasses for example are much more efficient than trees, which is why grassland can support huge herds of large animals, but a forest can't.
This is because more than one kind of photosynthesis has evolved, with somewhat different chemical processes. Look up C3 carbon fixation and C4 carbon fixation for the differences.
There is a significant research effort going on looking into ways of taking the genes for the more efficient types of photosynthesis and merging that into less efficient plants. This could be used to make fruit trees grow much faster, or to create algae that can be used to produce alcohol or oil at efficiencies approaching those of solar electric power.
"16 Marketing Managers,HR Directors, and First-Level Help Desk Technicians have decided that routinely testing backups is a waste of effort and not needed at all".
Of course not!
The systems didn't fail yesterday, or the day before that, and I can predict with 99% certainty that nothing will fail tomorrow! Why should we spend so much money on expensive tape library systems and people to manage them, when that money could be redirected towards much more useful things, like iPads for all of the executive staff?
You're young and inexperienced, so what could you possibly know? I can tell you that I've been here for many years, and we've never had a problem that we needed backups for, but we want, err... need iPads today.
What I don't understand is how the US political system has survived this long with this insane ability to throw anything into a proposed bill and go along for the ride, no matter how unrelated to the original topic.
If I could get elected to a position like where that's possible, I would do nothing other than attach the following clause to every bill that's proposed: "Pay the sum of one million US dollars to annually."
Of course, I'd add some clauses for avoiding taxes, ensure that upon my death the monies are paid to my estate, increase the amount to match inflation, grant myself immunity to prosecution, etc...
Sooner or later one of two things will happen:
- I will fail to be re-elected, but I'll be filthy rich for the rest of my life, so I really won't care. - Somebody will wake the fuck up and realise that arbitrary attachments to bills are insane, and stop the madness.
If C# is replacing Java then that would mean companies are also replacing UNIX with Windows as it's the only platform that supports C# (forget Mono). That's definitely not happening.
That is definitely happening! Maybe not where you are, but I'm a consultant that gets to see a wide range of corporations, and everywhere I go I see NetWare and UNIX getting replaced with Windows. It's cheaper than either of those options, and having a single OS family across all servers is a huge win for support and training costs.
The days when you had to have a "big iron" UNIX box to be able to handle an enterprise application workload are over. You can get a Windows compatible server with more CPU cores than you can use for peanuts these days. There just aren't that many business problems that need more than ~256 CPU cores and 2 TB of memory. Keep in mind that a CPU core these days is about 10x faster than a CPU core used to be just a decade ago!
The only area where Microsoft still hasn't won is in reliability, even though they have the technology to do so. Window supports hot patching, but no patch management system that I'm aware of actually applies the patches that way. Windows Server Core needs (supposedly) fewer patches, but still requires a reboot once a month in practice. The have cluster support, but still can't do active-active disks, except for some special cases like Hyper-V virtual disk storage. This is getting less important though as applications these days tend to use load-balanced clusters instead of a single highly reliable server.
There's a really good reason, and it should scare you: they're storing the password in clear-text instead of its hash, so the length is important.
Banks and other such organisations are likely to have some old crappy database running on a mainframe that was never really designed for "web" use. It may not even be possible to add new columns! What they do instead is take an existing unused column that takes a reasonably large text field, and stuff everything in there as a delimiter separated string. This is why they have to enforce a maximum password length, and why many sites also prevent you from using certain characters in passwords.
It's retarded, it's crazy insecure, and ironically it's a problem that affects the most important systems and data.
By definition, you are the geometric centre of the observable universe, which is defined as a sphere centred around you. All other observers are living in the past, and have smaller observable universes that are just subsets of your personal universe. I can talk to you about "my" observable universe, but because of the delay introduced by the speed of light, you can only find out about my observations after some non-zero delay. In that time, your universe would have expanded to include the older, smaller universe that influenced my observations.*
A more interesting point is that since you are an extended object, the universe cannot be a perfect sphere, as there is no well defined mathematical point that can act as the centre. In some sense, the entire volume of your brain is the centre. I've always wondered if that has some sort of metaphysical implications.
*) This may not be 100% true because of the expansion of space-time itself, but I haven't had enough coffee to think it through thoroughly.
There is no way in hell that your Java app uses anything even as remotely complex as DirectX, because a) Java is "lowest common denominator" and hardware acceleration is very new and platform-specific, and b) then it would certainly no longer work across all of those platforms. You'd have had difficulty even using multi-threading across all of those platforms back in the Netscape 2.0 days, because the threading and memory models weren't all that consistent. I've spent a lot of time debugging supposedly WORA multi-threaded Java apps back around 2003 that behaved hugely differently across Windows, Linux, and Solaris.
Comparing a Windows app (game?) that uses hardware acceleration (!) to a business app that's basically just a bunch of "if-else" code and string processing is not exactly fair. Try writing the SAME app on both platforms and see how your portability compares.
In my experience, Microsoft has some of the best backwards compatibility of any vendor out there. Well written C++ and.NET apps on Windows will probably keep working until the heat death of the universe. Meanwhile, Java took a long time to catch on to the fact that the runtime and standard libraries aren't 100% backwards compatible, and that people may actually want to run multiple versions side-by-side. For comparison, every.NET app uses the appropriate runtime automatically.
Back in 2006 I wrote a fairly complex NET app with several interacting components starting with.NET version 2.0 that ran on 32-bit Windows XP/2003. The exact same app works on two processors platforms (x86 and x64), on at least five major editions of Windows, three editions of IIS, and has been upgraded trivially through.NET 3.0, 3.5, and now 4.0. On top of that, the back-end database started on SQL Server 2005 RTM, and went through every major service pack release all the way up to 2012 RC0 without a hitch, despite using.NET stored procedures in the database.
Mind you, Microsoft's.NET isn't perfect either. Their insistence on using external unmanaged code for everything they can does sometimes bite them in the ass. For example, running a business app developed originally on a x86 computer on an x64 machine will sometimes cause it to run as 64-bit, which normally would be fine, except that Windows has a different set of ODBC drivers for x64. The app will run, but it might not be able to connect to its data sources. Oops.
Another good example of the destruction of an empire is Total Annihilation. Released back in the '90s for the first time, it was the first RTS with polygonal units (as opposed to sprites) where you could both create hundreds of units and select and control massive armies. One of the most significant perks of the game was the ability to create new construction prior to being able to afford it, like if you had half the bricks you needed to create your house so you got started before you had the rest of the bricks. If you run out before you're done, that's your problem. For the last 10 years that was fine, then Supreme Commander 2 came out (the 4th iteration of TA) and they removed this keystone element from the game to help simplify the game for console users.
This! A thousand time this!
I can still hardly believe that they started with one of the best RTS franchises out there with the vastly superior gameplay mechanics, took it outside into a dark alley, and shot it in the head.
The sad thing is that SC2 added queuing, probably as a copy of the feature from the TA/SC series, but it has the same problem -- it spends resources immediately for everything that is in the queue.
In both games, you are penalised for thinking ahead! If you want to win, you have to keep your "queue depth" at or near 1, otherwise you're tying up resources that won't be used till much later.
It's one of the few times I've seen a "major feature" in a game that is actually actively detrimental to your chances of winning. It's ridiculous to me that two companies with a long history of RTS game design have made the exact same mistake.
Team Fortress 2 has recently become free-to-play, and it's now suddenly full of 13 year olds with hacks, and there's nothing you can do about it. Before, you'd just get their Steam ID banned, and that would be it. Now, they just create a new free account, log back on right away, and continue griefing like nothing happened.
Mind you, that still takes some effort -- they have to register for a new Steam account, and possibly even change their CPU ID, but as long as there's no financial penalty, some twits are willing to go to that level of effort just to harass other people.
One of these days, I'm going to find one of those people, and I swear it's going to take all of my will to stop myself from indulging my inner Dexter.
It might help to understand where the "typical IT manager" goes wrong by seeing how it can be done right.
One of the first IT jobs I ever had was working for an IT manager of a ~150 user organisation. He was relatively new himself, which wasn't unusual because all of his predecessors were fired one after another. They just couldn't get along with management, couldn't make their needs understood, etc...
This new guy is still there, over a decade later. Why? Because he talked to managers in their own language. Instead of turning up to monthly board meetings in jeans and saying some buzzword-laden crap, he'd turn up in an expensive suit, put on a gorgeous powerpoint presentation which very clearly showed simple charts and graphs of things like "this is going to hit zero in a month, and that's bad because it'll stop our business". Half the time, he didn't even explain that it was disk-space he was talking about, or put numbers on the graph axes. Every month, he'd turn up with nice consistent reports full of simple charts printed in colour onto glossy paper, ending with a simple multiple-choice business decisions with dollar figures and pros and cons.
In the eyes of senior management, he turned IT from a dark pit where money is burned into a clearly separated set of projects and ongoing expenses that made sense to them. Yes, we have twice as many people now, so we're going to need twice as much storage. Obvious if stated right, not so obvious to someone who doesn't even know what "storage" really represents, why it runs out, and who uses it for what.
Here's the thing though: He couldn't solve a computer problem to save his life. That didn't matter, because he just hired competent underlings to do that work.
I found your post interesting, even though it appears to plagiarise this blog post, but I'll give you the benefit of the doubt and assume it's your work.
Anyway, your point is sort-of valid. I too find Bitcoin far more interesting as a protocol than as a real currency, because as a currency it is flawed, but it is a substantial step forward in the application of cryptography to seemingly unrelated problems.
You posit Bitcoin as not-a-currency, but a protocol for the exchange of currency. This doesn't making sense. Let me highlight the errors of logic in your argument by replacing all of your mentions of "dollars" with "valuable goods" and all mentions of "Bitcoin" with "dollars". Your post would now be saying (at first) perfectly valid things about money, and then coming to a conclusion that the end result is not-money, because... it doesn't work as money, so I don't want it used as money. E.g.:
Dollars are a very new technology, even though the concept that it brings to life is decades old. The double spending problem has been solved; this means that it is possible to use a paper certificate to stand in the place of valuable goods and be sure that no one else can spend that paper certificate other than you as long as you hold it. This is an unprecedented paradigm shift, the implications of which are not yet fully understood, and for which the tools do not yet exit to fully take advantage of this new idea.
Compared to exchanging good and services by trying to remember who owes what, dollars are a great improvement. It solves a lot of problems. You either have your money or you don't -- solving the "double spending problem".
Paper currency, which will transform the way you transfer goods and services, needs to be understood on its own terms, and not as a good. Thinking about dollars as a good is as absurd....
Still perfectly true, see? Money is not a good in and of itself. It's a piece of paper, and these days, not even that.
If dollars are a protocol and not goods, then setting up currency exchanges that mimic real world exchanges to trade in it doesn't make any sense.
This is where your argument breaks down. So far, Bitcoin has described something identical to money. Now, there are exchanges for money because there's more than one kind of money. There wouldn't be foreign exchanges if there was only one global currency. You also mixed up other sort of markets (e.g.: stock) with foreign exchange -- they're fundamentally different, because they buy and sell shares, and wait for it... cash is used only as an intermediary -- a protocol to facilitate the exchange.
The same cow is sent to India, whether you use $10,000 or $1. The price of dollars is irrelevant to the good that is being transmitted...
Now, again, that seems sort-of OK on the surface. It's true of dollars too, not just bitcoins. You can "send" a good to someone in another country by sending them enough cash to buy that good locally. This transfers real value, irrespective of the instantaneous "value" of the currency. Makes sense. Except that we know that it doesn't. The dollar would not be useful if its value suddenly dropped to "near zero". This has happened with real currencies before, and almost always results in massive economic damage and the permanent discontinuation of the currency.
The dollar's true nature is as an instant way to transmit the value of goods anywhere in the world. It is not an investment, or the good itself, and holding on to it in the hopes that it will become valuable is like holding on to an email or a PDF in the hopes it will be come valuable in the future; it doesn't make any sense.
You simply trust the postman not to open your letters. You trust the ISP not to read your email.
In your country, maybe, but that's certainly not the common case.
When I was growing up in a communist European country, the postal service would open letters and packages often. First of all, packages were routinely screened for 'contraband', and on top of that, all government agencies were rife with corruption, so theft of mailed goods was also commonplace. Sending anything valuable via post was a risk.
This is still the situation today in many (most?) third-world countries.
Ditto with the internet. You can be certain that the government controls and monitors traffic flowing through ISPs in a long list of countries such as China or Vietnam, and even the good old United States is well known for tapping into international links and using the information to benefit their domestic interests. That may not be a problem if you're a US citizen working for a US company -- it might even be a benefit -- but most of the world doesn't fall into that category.
One major concern is that currently most major 'cloud' services providers are US owned and have data centres in the US. Any major foreign business that engages in international dealings that intersect with US interests (e.g.: resource exports, oil exploration, etc...) ought to think twice before using US IT infrastructure to send or store anything related to the deal.
That sounds like a shockingly inefficient network, I doubt it has anything to do with Windows, and more to do with ingrained poor practices and typical bureaucratic inefficiency.
Switching to anything would have been an automatic improvement simply because it's an opportunity to cleanse the existing system with fire.
I disagree in the instance of androgenic global warming. It's just that science doesn't have all the data on that yet, nor do we know the models are correct. As to the rest, well, pretty much yeah.
I disagree in the instance of evolution. It's just that science doesn't have all the data on that yet, and I find it hard to believe that I'm related to monkeys. As to the rest, well, pretty much yeah.
I disagree in the instance of vaccinations. It's just that science doesn't have all the data on that yet, and I find it hard to believe that some lab geek knows what's best for my kids. As to the rest, well, pretty much yeah.
I may have remember the details wrong, but that doesn't detract from my point. Apparently red light is 750nm, not ~600 (oops). I remember the wavelength discussed as "not much longer". A wavelength of 1100nm is not that much longer than 750m, especially considering that up 10um is still considered to be IR.
One cool thing that I found while googling is this picture of someone doing the same kind of test with an incandescent light bulb, but using IR sensitive cameras to visually demonstrate the effect.
And anyway, you're kind of proving my point: your superior knowledge wins over a nice sounding story. I didn't even have to be wrong on purpose, merely having a fuzzy recollection of events was enough. Nonetheless, you have a method available to you for not just detecting the error, but demonstrating the correct answer. That's capability is just not available in any religion. You're told to believe whatever your religious text and/or people in positions of power say, and there's not much you can say to correct them -- because you have no way of discovering and then demonstrating what is 'correct', even in principle.
The problem is that in the social 'sciences', this is often treated as a 'everybody is right' instead of the approach of the physical sciences: "I'm right, and if you don't believe me, go do the experiment yourself".
That's a monumental difference that a lot of people just fail to grasp, even in serious fields of study. Just read this essay by Richard Feynman where he explains what it means to be properly scientific.
Nonetheless, students questioning their professors is not seen as a problem even in the physical sciences. For example, I had a very vocal disagreement with one of my Physics professors once. I simply refused to believe that what he was saying was possible. He was so impressed that he offered me a research position based on that one interaction.
Of course, this comes with a huge caveat -- I didn't 'just' disagree.
What had happened was that we were studying solid-state lasers, like the type you get in your DVD player or a laser pointer. They are made from crystals of semiconductors, like silicon, germanium, arsenic, etc... He was specifically discussing silicon lasers emitting light at about 650nm. I sat straight up and thought that's crazy -- I've held pure silicon in my hand before, and it looks like metal. Sure, it's a bit dark, but I just couldn't imagine how light that's "just barely infra-red" could go straight through the thing with nearly 0% loss, which is what a laser requires to operate. I argued with him -- surely it's very heavily doped and it's actually a compound of silicon that transmits the light? No. Maybe it's just a very thin surface layer, like transparent gold leaf? No.
The day after that, I was in the lab, and there was a piece of silicon there -- scrap from the chip lab. I took an incandescent lamp that I knew put out most of it's heat energy in the right infra-red range, put my hand in front of it, and then I waved the silicon wafer back and forth between my hand and the light. It's like it wasn't even there -- it blocked none of the IR light. There was no visible light going through, but I could feel the heat on my hand. I compared it to glass and various thicknesses of paper and plastic sheets. Only silicon transmitted all of the IR heat energy. It was like it was made of smoke. Sure, it was a primitive experiment, but very convincing in a I-can-feel-it-with-my-own-hands kind of way.
The next day, we were back in the lecture hall, continuing the topic of silicon lasers, and the lecturer jokingly asked me if I still had problems believing that silicon was transparent to infra-red light. I said no, I tried passing IR light through a piece of silicon in the lab. It doesn't look like it should, but it does.
That change in my position is the very essence of science -- not that disagreeing is bad, but there ought to be a method by which we can all become convinced of the truth and agree on it.
Sadly, the scientific method is not followed rigorously in many fields. Psychology and some areas of medicine come to mind. Just read: Why Most Published Research Findings Are False for an idea of just how far it's possible to stray from the truth because of only small errors in the application of the scientific method.
The Amundsen-Scott South Pole Station in Antarctica is actually a great example, thanks for bringing it up. It's entirely dependent on the outside world for supplies except for air and water. Everything, and I mean everything is shipped in. There's no self-sufficiency to speak of.
We think of Antarctica as an inhospitable place, but it's a tropical paradise compared to anywhere we can land people in space. It has unlimited oxygen and and water, no dangerous radiation, earth normal gravity, unlimited water, vast mineral deposits, and temperatures that can be survived with nothing more than some warm clothing.
Nonetheless, if external support was cut off from the south pole station, then despite having all the existing buildings, infrastructure, machinery, and a staff of hundreds of brilliant scientists and researchers, everyone there would die.
Let me reiterate this: your examples of 'colonies' are all places where the people there are supported by enormous external supply chains, and would die if those supplies are cut off. On Earth, we can keep the supplies going because we can afford to, and because it's worth it -- the relatively low overhead of air freighting in everything is small compared to the valuable science that can be performed in Antarctica, or the money people are willing to spend to climb Everest.
All of these are expeditions, not colonies. They're not self sufficient, and it wouldn't be cost effective to make them self sufficient.
Shipping stuff on Earth is cheap. Air freight to a frozen desert in the middle of nowhere is a negligible overhead when compared to sending stuff to Mars. Even in the wildest, most delusional dreams of space fanatics, there is no way to do it for less than about $100 per pound.
Look around your house -- really look -- and for everything you see, ask yourself: how many pounds is that?. Could anyone afford to live like this if it cost $100 per pound more than it would otherwise? How many pounds of water do you use? How many pounds of air? What does your house weigh?
Try that again with the current, realistic cost of sending things to Mars of $10,000 per pound.
Slashdot Mirror
Good grief... so you're saying that by using Erlang you need a minimum 20 cores to get a measly 2x speed-up over a single-threaded program, which is probably easier to write too?
Multi-threaded programming is not unique to Erlang, writing parallel code in Java, C#, or even C++ is not exactly rocket surgery, so the comparison to strictly single-threaded code is unfair. I can get a program in any one of those languages to scale up to at least hundreds processors without breaking a sweat. Heck, any web application is automatically multi-threaded with both Java and C#, and with version .NET Framework v4.5 or later it can be trivially made asynchronous as well on top of that! You don't need a single lock or threading primitive of any kind to achieve that. If you want to get more advanced, all three platforms give you easy to use and powerful libraries to make multi-threading safe and efficient.
It's only the crappy open-source platforms written by non-professionals that struggle to scale, which is why Erlang looks so good in comparison to a lot of beginner programmers, and why Node.js is so fascinating to them. Ooo... look... we've invented asynchronous calls, which is brand new and fascinating, because clearly nobody has ever done that before...
The real benefit of Erlang is not speed, but reliability and online maintainability. It's designed for non-stop systems where uptime is more important than performance. This is explicitly stated in its documentation as the primary design goal of the language! Using Erlang to improve the performance of code is asinine. If speed matters, any of the three popular languages will run circles around Erlang.
That fancy 64-processor SGI box of yours? You've lost 90% of its performance to Erlang's inefficiency, so I could get the same performance out of a 6-core computer with an efficient programming language. I have 6 cores in my desktop. I bet it cost a few digits less than your SGI box.
I just did, and the noise is a lot lower, but I can't get the volume as high, and some movies end up too quiet. The only remaining noise is now the cooling fans of my PC, but that's fixable...
Possibly, but I picked this pair of headphones out of over a dozen after an hour of listening tests.
I couldn't find any other headphone in the store (which specialised in nothing other than headphones) that could could go to as low a frequency as these. One of my test tracks was a classical piece with very low frequency double bass notes. I couldn't even hear the instrument at all with most of the others, one other headphone could reproduce it, but all tonality was lost, while with the Sennheiser HD800 I could actually distinguish each note. I compared it to other lower-end Sennheiser models, and a bunch of other brands. It stood out as noticeably better than the others.
I figured it is probably over-priced, but what-the-hey, it costs less than a single speaker for a high-end home audio system, so who cares?
This article is of particular interest to me because just recently I dropped over $1000 on a pair of high-end Sennheiser HD-800 headphones, but now I'm finding the amplifier's background noise is a lot more noticeable. Before, with cheap headphones that didn't have the same dynamic range, it didn't matter, but now it's the limiting factor.
I've got a reasonably decent FLAC collection, some of it classical music in 24/96Khz format, but the background hiss is detracting quite a bit from the potential quality.
What's a good amplifier for headphones that's optimised for a low noise floor instead of power, but isn't over-priced? I don't need a receiver with hundreds of inputs, I need something that takes a single digital input, and outputs the highest possible quality headphone output. Is there anything like that out there?
What I don't understand is why American universities even have campus police?
The last couple of times I've been to an Australian University, including two of the biggest, the closest I've seen were two very bored looking security staff members who spent most of their day making student and staff cards from a tiny little office in the basement. They certainly didn't have guns, and probably didn't even have pepper spray -- none that I could see anyway.
Just how bad is the crime situation on your campuses that you need both a dedicated police force and a security force on top of that?
To a degree, CSIRO here in Australia works like this. Its research is often commercialized and sold, usually via patent licensing. They just recently won $200 million from companies manufacturing WiFi equipment that hadn't properly licensed their patents.
Personally I think that the way a national research organisation should function is that it should give its research away for free to the citizens and corporations of its home country, and charge foreign organisations to use the patents. That way, it would be a socialised boost to the nation's research without the overhead of having to charge people for access, but still self-fund itself to a degree.
Trees are about as effective in doing photosynthesis as any other plant.
That's not true, there's significant differences in efficiency between various species of plants. Most grasses for example are much more efficient than trees, which is why grassland can support huge herds of large animals, but a forest can't.
See: Photosynthetic Efficiency, where it has a table of some typical efficiencies:
Plants, typical : 0.1%
Typical crop plants: 1-2%
Sugarcane: 7-8% peak
This is because more than one kind of photosynthesis has evolved, with somewhat different chemical processes. Look up C3 carbon fixation and C4 carbon fixation for the differences.
There is a significant research effort going on looking into ways of taking the genes for the more efficient types of photosynthesis and merging that into less efficient plants. This could be used to make fruit trees grow much faster, or to create algae that can be used to produce alcohol or oil at efficiencies approaching those of solar electric power.
"16 Marketing Managers,HR Directors, and First-Level Help Desk Technicians have decided that routinely testing backups is a waste of effort and not needed at all".
Of course not!
The systems didn't fail yesterday, or the day before that, and I can predict with 99% certainty that nothing will fail tomorrow! Why should we spend so much money on expensive tape library systems and people to manage them, when that money could be redirected towards much more useful things, like iPads for all of the executive staff?
You're young and inexperienced, so what could you possibly know? I can tell you that I've been here for many years, and we've never had a problem that we needed backups for, but we want, err... need iPads today.
What I don't understand is how the US political system has survived this long with this insane ability to throw anything into a proposed bill and go along for the ride, no matter how unrelated to the original topic.
If I could get elected to a position like where that's possible, I would do nothing other than attach the following clause to every bill that's proposed: "Pay the sum of one million US dollars to annually."
Of course, I'd add some clauses for avoiding taxes, ensure that upon my death the monies are paid to my estate, increase the amount to match inflation, grant myself immunity to prosecution, etc...
Sooner or later one of two things will happen:
- I will fail to be re-elected, but I'll be filthy rich for the rest of my life, so I really won't care.
- Somebody will wake the fuck up and realise that arbitrary attachments to bills are insane, and stop the madness.
If C# is replacing Java then that would mean companies are also replacing UNIX with Windows as it's the only platform that supports C# (forget Mono). That's definitely not happening.
That is definitely happening! Maybe not where you are, but I'm a consultant that gets to see a wide range of corporations, and everywhere I go I see NetWare and UNIX getting replaced with Windows. It's cheaper than either of those options, and having a single OS family across all servers is a huge win for support and training costs.
The days when you had to have a "big iron" UNIX box to be able to handle an enterprise application workload are over. You can get a Windows compatible server with more CPU cores than you can use for peanuts these days. There just aren't that many business problems that need more than ~256 CPU cores and 2 TB of memory. Keep in mind that a CPU core these days is about 10x faster than a CPU core used to be just a decade ago!
The only area where Microsoft still hasn't won is in reliability, even though they have the technology to do so. Window supports hot patching, but no patch management system that I'm aware of actually applies the patches that way. Windows Server Core needs (supposedly) fewer patches, but still requires a reboot once a month in practice. The have cluster support, but still can't do active-active disks, except for some special cases like Hyper-V virtual disk storage. This is getting less important though as applications these days tend to use load-balanced clusters instead of a single highly reliable server.
There's a really good reason, and it should scare you: they're storing the password in clear-text instead of its hash, so the length is important.
Banks and other such organisations are likely to have some old crappy database running on a mainframe that was never really designed for "web" use. It may not even be possible to add new columns! What they do instead is take an existing unused column that takes a reasonably large text field, and stuff everything in there as a delimiter separated string. This is why they have to enforce a maximum password length, and why many sites also prevent you from using certain characters in passwords.
It's retarded, it's crazy insecure, and ironically it's a problem that affects the most important systems and data.
By definition, you are the geometric centre of the observable universe, which is defined as a sphere centred around you. All other observers are living in the past, and have smaller observable universes that are just subsets of your personal universe. I can talk to you about "my" observable universe, but because of the delay introduced by the speed of light, you can only find out about my observations after some non-zero delay. In that time, your universe would have expanded to include the older, smaller universe that influenced my observations.*
A more interesting point is that since you are an extended object, the universe cannot be a perfect sphere, as there is no well defined mathematical point that can act as the centre. In some sense, the entire volume of your brain is the centre. I've always wondered if that has some sort of metaphysical implications.
*) This may not be 100% true because of the expansion of space-time itself, but I haven't had enough coffee to think it through thoroughly.
Maybe, but that is a massively unfair comparison.
There is no way in hell that your Java app uses anything even as remotely complex as DirectX, because a) Java is "lowest common denominator" and hardware acceleration is very new and platform-specific, and b) then it would certainly no longer work across all of those platforms. You'd have had difficulty even using multi-threading across all of those platforms back in the Netscape 2.0 days, because the threading and memory models weren't all that consistent. I've spent a lot of time debugging supposedly WORA multi-threaded Java apps back around 2003 that behaved hugely differently across Windows, Linux, and Solaris.
Comparing a Windows app (game?) that uses hardware acceleration (!) to a business app that's basically just a bunch of "if-else" code and string processing is not exactly fair. Try writing the SAME app on both platforms and see how your portability compares.
In my experience, Microsoft has some of the best backwards compatibility of any vendor out there. Well written C++ and .NET apps on Windows will probably keep working until the heat death of the universe. Meanwhile, Java took a long time to catch on to the fact that the runtime and standard libraries aren't 100% backwards compatible, and that people may actually want to run multiple versions side-by-side. For comparison, every .NET app uses the appropriate runtime automatically.
Back in 2006 I wrote a fairly complex NET app with several interacting components starting with .NET version 2.0 that ran on 32-bit Windows XP/2003. The exact same app works on two processors platforms (x86 and x64), on at least five major editions of Windows, three editions of IIS, and has been upgraded trivially through .NET 3.0, 3.5, and now 4.0. On top of that, the back-end database started on SQL Server 2005 RTM, and went through every major service pack release all the way up to 2012 RC0 without a hitch, despite using .NET stored procedures in the database.
Mind you, Microsoft's .NET isn't perfect either. Their insistence on using external unmanaged code for everything they can does sometimes bite them in the ass. For example, running a business app developed originally on a x86 computer on an x64 machine will sometimes cause it to run as 64-bit, which normally would be fine, except that Windows has a different set of ODBC drivers for x64. The app will run, but it might not be able to connect to its data sources. Oops.
Another good example of the destruction of an empire is Total Annihilation. Released back in the '90s for the first time, it was the first RTS with polygonal units (as opposed to sprites) where you could both create hundreds of units and select and control massive armies. One of the most significant perks of the game was the ability to create new construction prior to being able to afford it, like if you had half the bricks you needed to create your house so you got started before you had the rest of the bricks. If you run out before you're done, that's your problem. For the last 10 years that was fine, then Supreme Commander 2 came out (the 4th iteration of TA) and they removed this keystone element from the game to help simplify the game for console users.
This! A thousand time this!
I can still hardly believe that they started with one of the best RTS franchises out there with the vastly superior gameplay mechanics, took it outside into a dark alley, and shot it in the head.
The sad thing is that SC2 added queuing, probably as a copy of the feature from the TA/SC series, but it has the same problem -- it spends resources immediately for everything that is in the queue.
In both games, you are penalised for thinking ahead! If you want to win, you have to keep your "queue depth" at or near 1, otherwise you're tying up resources that won't be used till much later.
It's one of the few times I've seen a "major feature" in a game that is actually actively detrimental to your chances of winning. It's ridiculous to me that two companies with a long history of RTS game design have made the exact same mistake.
Gah... don't get me started!
Team Fortress 2 has recently become free-to-play, and it's now suddenly full of 13 year olds with hacks, and there's nothing you can do about it. Before, you'd just get their Steam ID banned, and that would be it. Now, they just create a new free account, log back on right away, and continue griefing like nothing happened.
Mind you, that still takes some effort -- they have to register for a new Steam account, and possibly even change their CPU ID, but as long as there's no financial penalty, some twits are willing to go to that level of effort just to harass other people.
One of these days, I'm going to find one of those people, and I swear it's going to take all of my will to stop myself from indulging my inner Dexter.
It's the indie stuff that keeps me coming back.
I got Limbo, Braid, and Bastion for about $5 each when they were on sale.
That's ridiculously good value!
That puts some things into perspective.
Imagine trying to build a battery-based energy storage system for intermittent power sources, like wind or solar.
The entire yearly output of batteries from this factory would be able to buffer less than an hour of the power from a 1 GW power plant!
Compared to our ability to generate power, our capability to store it is still quite poor.
It might help to understand where the "typical IT manager" goes wrong by seeing how it can be done right.
One of the first IT jobs I ever had was working for an IT manager of a ~150 user organisation. He was relatively new himself, which wasn't unusual because all of his predecessors were fired one after another. They just couldn't get along with management, couldn't make their needs understood, etc...
This new guy is still there, over a decade later. Why? Because he talked to managers in their own language. Instead of turning up to monthly board meetings in jeans and saying some buzzword-laden crap, he'd turn up in an expensive suit, put on a gorgeous powerpoint presentation which very clearly showed simple charts and graphs of things like "this is going to hit zero in a month, and that's bad because it'll stop our business". Half the time, he didn't even explain that it was disk-space he was talking about, or put numbers on the graph axes. Every month, he'd turn up with nice consistent reports full of simple charts printed in colour onto glossy paper, ending with a simple multiple-choice business decisions with dollar figures and pros and cons.
In the eyes of senior management, he turned IT from a dark pit where money is burned into a clearly separated set of projects and ongoing expenses that made sense to them. Yes, we have twice as many people now, so we're going to need twice as much storage. Obvious if stated right, not so obvious to someone who doesn't even know what "storage" really represents, why it runs out, and who uses it for what.
Here's the thing though: He couldn't solve a computer problem to save his life. That didn't matter, because he just hired competent underlings to do that work.
I found your post interesting, even though it appears to plagiarise this blog post, but I'll give you the benefit of the doubt and assume it's your work.
Anyway, your point is sort-of valid. I too find Bitcoin far more interesting as a protocol than as a real currency, because as a currency it is flawed, but it is a substantial step forward in the application of cryptography to seemingly unrelated problems.
You posit Bitcoin as not-a-currency, but a protocol for the exchange of currency. This doesn't making sense. Let me highlight the errors of logic in your argument by replacing all of your mentions of "dollars" with "valuable goods" and all mentions of "Bitcoin" with "dollars". Your post would now be saying (at first) perfectly valid things about money, and then coming to a conclusion that the end result is not-money, because... it doesn't work as money, so I don't want it used as money. E.g.:
Dollars are a very new technology, even though the concept that it brings to life is decades old. The double spending problem has been solved; this means that it is possible to use a paper certificate to stand in the place of valuable goods and be sure that no one else can spend that paper certificate other than you as long as you hold it. This is an unprecedented paradigm shift, the implications of which are not yet fully understood, and for which the tools do not yet exit to fully take advantage of this new idea.
Compared to exchanging good and services by trying to remember who owes what, dollars are a great improvement. It solves a lot of problems. You either have your money or you don't -- solving the "double spending problem".
Paper currency, which will transform the way you transfer goods and services, needs to be understood on its own terms, and not as a good. Thinking about dollars as a good is as absurd....
Still perfectly true, see? Money is not a good in and of itself. It's a piece of paper, and these days, not even that.
If dollars are a protocol and not goods, then setting up currency exchanges that mimic real world exchanges to trade in it doesn't make any sense.
This is where your argument breaks down. So far, Bitcoin has described something identical to money. Now, there are exchanges for money because there's more than one kind of money. There wouldn't be foreign exchanges if there was only one global currency. You also mixed up other sort of markets (e.g.: stock) with foreign exchange -- they're fundamentally different, because they buy and sell shares, and wait for it... cash is used only as an intermediary -- a protocol to facilitate the exchange.
The same cow is sent to India, whether you use $10,000 or $1. The price of dollars is irrelevant to the good that is being transmitted...
Now, again, that seems sort-of OK on the surface. It's true of dollars too, not just bitcoins. You can "send" a good to someone in another country by sending them enough cash to buy that good locally. This transfers real value, irrespective of the instantaneous "value" of the currency. Makes sense. Except that we know that it doesn't. The dollar would not be useful if its value suddenly dropped to "near zero". This has happened with real currencies before, and almost always results in massive economic damage and the permanent discontinuation of the currency.
The dollar's true nature is as an instant way to transmit the value of goods anywhere in the world. It is not an investment, or the good itself, and holding on to it in the hopes that it will become valuable is like holding on to an email or a PDF in the hopes it will be come valuable in the future; it doesn't make any sense.
You simply trust the postman not to open your letters. You trust the ISP not to read your email.
In your country, maybe, but that's certainly not the common case.
When I was growing up in a communist European country, the postal service would open letters and packages often. First of all, packages were routinely screened for 'contraband', and on top of that, all government agencies were rife with corruption, so theft of mailed goods was also commonplace. Sending anything valuable via post was a risk.
This is still the situation today in many (most?) third-world countries.
Ditto with the internet. You can be certain that the government controls and monitors traffic flowing through ISPs in a long list of countries such as China or Vietnam, and even the good old United States is well known for tapping into international links and using the information to benefit their domestic interests. That may not be a problem if you're a US citizen working for a US company -- it might even be a benefit -- but most of the world doesn't fall into that category.
One major concern is that currently most major 'cloud' services providers are US owned and have data centres in the US. Any major foreign business that engages in international dealings that intersect with US interests (e.g.: resource exports, oil exploration, etc...) ought to think twice before using US IT infrastructure to send or store anything related to the deal.
That sounds like a shockingly inefficient network, I doubt it has anything to do with Windows, and more to do with ingrained poor practices and typical bureaucratic inefficiency.
Switching to anything would have been an automatic improvement simply because it's an opportunity to cleanse the existing system with fire.
I disagree in the instance of androgenic global warming. It's just that science doesn't have all the data on that yet, nor do we know the models are correct. As to the rest, well, pretty much yeah.
I disagree in the instance of evolution. It's just that science doesn't have all the data on that yet, and I find it hard to believe that I'm related to monkeys. As to the rest, well, pretty much yeah.
I disagree in the instance of vaccinations. It's just that science doesn't have all the data on that yet, and I find it hard to believe that some lab geek knows what's best for my kids. As to the rest, well, pretty much yeah.
Do you see a pattern emerging?
I may have remember the details wrong, but that doesn't detract from my point. Apparently red light is 750nm, not ~600 (oops). I remember the wavelength discussed as "not much longer". A wavelength of 1100nm is not that much longer than 750m, especially considering that up 10um is still considered to be IR.
One cool thing that I found while googling is this picture of someone doing the same kind of test with an incandescent light bulb, but using IR sensitive cameras to visually demonstrate the effect.
And anyway, you're kind of proving my point: your superior knowledge wins over a nice sounding story. I didn't even have to be wrong on purpose, merely having a fuzzy recollection of events was enough. Nonetheless, you have a method available to you for not just detecting the error, but demonstrating the correct answer. That's capability is just not available in any religion. You're told to believe whatever your religious text and/or people in positions of power say, and there's not much you can say to correct them -- because you have no way of discovering and then demonstrating what is 'correct', even in principle.
The problem is that in the social 'sciences', this is often treated as a 'everybody is right' instead of the approach of the physical sciences: "I'm right, and if you don't believe me, go do the experiment yourself".
That's a monumental difference that a lot of people just fail to grasp, even in serious fields of study. Just read this essay by Richard Feynman where he explains what it means to be properly scientific.
Nonetheless, students questioning their professors is not seen as a problem even in the physical sciences. For example, I had a very vocal disagreement with one of my Physics professors once. I simply refused to believe that what he was saying was possible. He was so impressed that he offered me a research position based on that one interaction.
Of course, this comes with a huge caveat -- I didn't 'just' disagree.
What had happened was that we were studying solid-state lasers, like the type you get in your DVD player or a laser pointer. They are made from crystals of semiconductors, like silicon, germanium, arsenic, etc... He was specifically discussing silicon lasers emitting light at about 650nm. I sat straight up and thought that's crazy -- I've held pure silicon in my hand before, and it looks like metal. Sure, it's a bit dark, but I just couldn't imagine how light that's "just barely infra-red" could go straight through the thing with nearly 0% loss, which is what a laser requires to operate. I argued with him -- surely it's very heavily doped and it's actually a compound of silicon that transmits the light? No. Maybe it's just a very thin surface layer, like transparent gold leaf? No.
The day after that, I was in the lab, and there was a piece of silicon there -- scrap from the chip lab. I took an incandescent lamp that I knew put out most of it's heat energy in the right infra-red range, put my hand in front of it, and then I waved the silicon wafer back and forth between my hand and the light. It's like it wasn't even there -- it blocked none of the IR light. There was no visible light going through, but I could feel the heat on my hand. I compared it to glass and various thicknesses of paper and plastic sheets. Only silicon transmitted all of the IR heat energy. It was like it was made of smoke. Sure, it was a primitive experiment, but very convincing in a I-can-feel-it-with-my-own-hands kind of way.
The next day, we were back in the lecture hall, continuing the topic of silicon lasers, and the lecturer jokingly asked me if I still had problems believing that silicon was transparent to infra-red light. I said no, I tried passing IR light through a piece of silicon in the lab. It doesn't look like it should, but it does.
That change in my position is the very essence of science -- not that disagreeing is bad, but there ought to be a method by which we can all become convinced of the truth and agree on it.
Sadly, the scientific method is not followed rigorously in many fields. Psychology and some areas of medicine come to mind. Just read: Why Most Published Research Findings Are False for an idea of just how far it's possible to stray from the truth because of only small errors in the application of the scientific method.
The Amundsen-Scott South Pole Station in Antarctica is actually a great example, thanks for bringing it up. It's entirely dependent on the outside world for supplies except for air and water. Everything, and I mean everything is shipped in. There's no self-sufficiency to speak of.
We think of Antarctica as an inhospitable place, but it's a tropical paradise compared to anywhere we can land people in space. It has unlimited oxygen and and water, no dangerous radiation, earth normal gravity, unlimited water, vast mineral deposits, and temperatures that can be survived with nothing more than some warm clothing.
Nonetheless, if external support was cut off from the south pole station, then despite having all the existing buildings, infrastructure, machinery, and a staff of hundreds of brilliant scientists and researchers, everyone there would die.
Let me reiterate this: your examples of 'colonies' are all places where the people there are supported by enormous external supply chains, and would die if those supplies are cut off. On Earth, we can keep the supplies going because we can afford to, and because it's worth it -- the relatively low overhead of air freighting in everything is small compared to the valuable science that can be performed in Antarctica, or the money people are willing to spend to climb Everest.
All of these are expeditions, not colonies. They're not self sufficient, and it wouldn't be cost effective to make them self sufficient.
Shipping stuff on Earth is cheap. Air freight to a frozen desert in the middle of nowhere is a negligible overhead when compared to sending stuff to Mars. Even in the wildest, most delusional dreams of space fanatics, there is no way to do it for less than about $100 per pound.
Look around your house -- really look -- and for everything you see, ask yourself: how many pounds is that?. Could anyone afford to live like this if it cost $100 per pound more than it would otherwise? How many pounds of water do you use? How many pounds of air? What does your house weigh?
Try that again with the current, realistic cost of sending things to Mars of $10,000 per pound.