I've read claims that the school policy was to keep the machines on 24/7 and the tech department's proposal to automate shutting down the machines overnight (to the tune of saving $90K/year) was rejected. If that's true, the school is standing on one foot (having shot themselves in the other).
Yes, but the amount of power actually consumed by the machine depends on the workload. A truly idle machine consumes less. The exact figures depend on the system specifics; a measurement tool should be used under the different workload scenarios to figure out what the actual cost is. The other things to take into account are bandwidth costs (for one machine running SETI@home this is trivial, but for 5000 it's not so small) and the costs from a shortened replacement cycle; as systems work more, they wear out faster.
Though yes, the school board were douches to say that systems should be on 24/7. Throwing away $90k/year like that is just... stupid. (Doesn't mean that the other guy was in the right. Sometimes it's just a battle of the assholes.)
Scripts are fine for small mundane tasks, they're NOT good for building applications.
A lot of the very largest applications are scripted (in various languages). If you've got a choice between writing a million lines of script or a hundred million lines of C, it's a real no-brainer.
Well, in fact the massive apps use a mix of different languages, typically crafting components in a low-level language that the high-level scripting language composes to create the overall application. That tends to be the most productive approach; the alternative "One True Language" method is silly.
But then would *you* run arbitrary native binary code off the web?
No, but lots of people do. Witness the vast numbers of idiots who get caught by trojans masquerading as cute cat screensavers or whatever the latest meme is.
[District heating is] fairly common in Northern Europe (excepting England, for some reason).
It's usually used for industrial purposes in England rather than domestic or commercial. Part of this is due to the fact that large power plants tend to be located well outside cities (as an anti-pollution measure and because land values out in the boonies are lower).
Rule #1 of cloud computing: "Do not trust the cloud".
Rule #1 is "You get the level of service that you pay for."
(OK, rule #1 is really "Don't get caught" but you know what I mean.)
Why is Google even able to review the content? Content should be encrypted.
Because they're a very cheap provider. Pay more and you'll get the sort of service that you're asking for. Do you want the money or do you want the privacy?
Things like RSA encryption are good examples of patents.
... which was so novel, it had already been independently developed.
And it's a good thing that the number theory it relies on wasn't patented by Euler.
The innovative thing was using the pieces together to create a piece of software that could encrypt and decrypt messages without using an explicit shared secret. That greatly changed things, and it's things like that which merit patenting (or other protection). Whether it was RSA or a predecessor that did it is not so important. Whoever it was, allowing a patent on the idea was reasonable.
The problem with software patents is that they're too often awarded without making clear just what is going on ("making the method patent"; it's the original definition of the word) and are too often awarded for trivial things, things that would actually be obvious extensions of the idea to someone who is a practitioner in the art of programming. Those issues (plus general problems like submarine patents) turn reasonable protection of real innovation that advances the state of human thought and productivity into a tool of oppression. Had the patent office only awarded a small number of software patents on the really innovative stuff, we wouldn't be having this argument now.
But in an application where the lights stay on all night, regular CFL bulbs work fine in the coldest days of winter.
Why are you having lights on outside all night? To give the local wildlife something to see by? To make it easier for burglars to find their way across your property? You could just switch the lights off during the hours when you're asleep (using a cheap timeswitch) and save a bunch at practically zero cost to yourself.
That's only one of the goals. Not having to replace broken bulbs so often is another good goal, and we can justify that on pure convenience for ourselves. (You could also make points about reduced cost of transport, etc., but you'd just be skirting round the central fact of it just being better on the laziness factor.)
Sadly if we start shutting down corrupt businesses we will shut down the American economy.
If that's so, then the economy deserves to be shut down. There's no reason at all to ever tolerate crooked businesses. (And no, I don't think that shutting them down will break the economy; most people - even most business owners - are honest.)
Off the top of my head, your applications will run on OS X without having to add Aqua specific extensions, people won't ignore your application just because it's "java" and you won't get a plethora of users moaning that it's "slow".
The main issue for OSX GUI apps vs. GUI apps on other platforms is that apps tend to be laid out differently on OSX. So while you can reuse your layouts from other platforms, they look out of place. (By contrast, Windows and Linux tend to use much more similar GUI layout strategies; their GUI heritages diverged much more recently.)
This kind of applications won't have the hardware catching up to let you replace C, Asembly and VHDL with Ruby or Java for decades yet.
I have contacts in the embedded safety-critical world (though admittedly they're in a domain where there's lots of power available). They're using high-level languages. I wonder whether it is a good idea for them to do so, but they seem to be happy and productive so the risk-reward tradeoff must seem OK to them.
I suspect that if you only ever use low-level languages then you only ever really deal with low-level details, but a program in total - including runtime and libs - will always have both low-level and high-level concerns. High-level languages let you get to working on the high-level problems sooner; the low-level stuff is left to the runtime and libs which are (usually) well-written enough to take care of it without messing. (For a simple example, why would you write your own sorting code when you could just use one from a library where they've picked a good algorithm and dealt with the complex edge-cases already? There are cases where you would, but most coders aren't writing those things.)
On the desktop? Sure, who cares, though again a lot of java apps are slow to start and sluggish to respond.
That's the basic startup cost of the java runtime, which is a bloated pig. But if you compare a complex app startup then the costs aren't that different (opening databases and initializing security contexts can take quite a while...)
Java does best for domains where it can stay resident, such as a complex server. Writing those in C is a real nightmare - if one module gets its memory management wrong then the whole lot goes up in smoke and you've got almost no hope to find out where, especially for rare errors - whereas it's a lot more practical in a managed language like Java or C# (or Python or...) I'll reserve judgement there on C++ (some people think it is suited, others don't).
You do know that reference counting is a (simple and limited) form of garbage collection?
You do know that refcounts work fine as long as you don't have reference loops? They even tend to give you better memory release behavior, though the release code has to be written carefully; the obvious recursive algorithm can blow up on large trees by blowing the stack. (Of course, as long as you know that this is possible it's pretty easy to avoid.)
The gain of "real" GC is that it handles the complex cases better, but you pay for that. It's often worthwhile paying that "tax".
As with most things, the USA is still using the same system as the UK, and it's only mainland Europe that has them reversed. I'm not sure 'God' has anything to do with it.
Of course He has something to do with it. As everyone of note knows, God is an Englishman. (There's even some evidence.)
I strongly suspect that the manufacturing capacity of Europe is rather a lot smaller than the total EU-wide demand for consumer goods (eg. childs toys). It follows that if the EC were to receive such a letter, they couldn't very well respond by saying "Fine. We'll embargo all your products" for very long - they'd drive the prices for a lot of items up so high that the politicians in member states would have Hell to pay.
Ah, but remember that they can blame it on those horrible Chinese and dastardly unelected Commission so they look blameless. It's very useful for the national-level politicians to have someone else to take the unpopular decisions for them so they can focus on taking decisions that people will vote for...
The UK allows a couple of extra channels (12 and 13) that I believe you can't use in the US. It might be necessary to change some settings to allow use of these channels.
However, few APs actually use these channels, so it's unlikely to be much of an issue.
It's especially not an issue because he's not bringing an access point (that would be a silly thing to bring for a two-week trip). Almost all APs are on channels 1, 6 and 11, the same as virtually everywhere else in the world; if you find another in use, it's almost worth posting a story on/. about it...
UK is 230V, as is most of the rest of Europe. At one time it was 240V, but voltage rates have been harmonized. Hence the reason many devices have been required to support 220-240V to facilitate that harmonization.
Only nominally. In practice, what's happened is that devices have had to cope with wider ranges of voltages than before. For switched-mode power supplies, this is not an issue (they run hotter on the higher voltage than in North America, but not so much that it's a big issue) and you won't be bringing the straight resistive-load devices where this would be a problem. After all, why would you bring a kettle when you can get one for hardly anything in any supermarket? Hotels that have razor sockets in the bathroom will also allow you to control the voltage delivered there.
Slashdot Mirror
I've read claims that the school policy was to keep the machines on 24/7 and the tech department's proposal to automate shutting down the machines overnight (to the tune of saving $90K/year) was rejected. If that's true, the school is standing on one foot (having shot themselves in the other).
Yes, but the amount of power actually consumed by the machine depends on the workload. A truly idle machine consumes less. The exact figures depend on the system specifics; a measurement tool should be used under the different workload scenarios to figure out what the actual cost is. The other things to take into account are bandwidth costs (for one machine running SETI@home this is trivial, but for 5000 it's not so small) and the costs from a shortened replacement cycle; as systems work more, they wear out faster.
Though yes, the school board were douches to say that systems should be on 24/7. Throwing away $90k/year like that is just... stupid. (Doesn't mean that the other guy was in the right. Sometimes it's just a battle of the assholes.)
Scripts are fine for small mundane tasks, they're NOT good for building applications.
A lot of the very largest applications are scripted (in various languages). If you've got a choice between writing a million lines of script or a hundred million lines of C, it's a real no-brainer.
Well, in fact the massive apps use a mix of different languages, typically crafting components in a low-level language that the high-level scripting language composes to create the overall application. That tends to be the most productive approach; the alternative "One True Language" method is silly.
But then would *you* run arbitrary native binary code off the web?
No, but lots of people do. Witness the vast numbers of idiots who get caught by trojans masquerading as cute cat screensavers or whatever the latest meme is.
[District heating is] fairly common in Northern Europe (excepting England, for some reason).
It's usually used for industrial purposes in England rather than domestic or commercial. Part of this is due to the fact that large power plants tend to be located well outside cities (as an anti-pollution measure and because land values out in the boonies are lower).
but you don't generally tell children that CEOs are trusted authority figures who deserve their respect and obedience.
Well, unless their name is Steve Jobs.
Doesn't count. Apple are a religion.
Rule #1 of cloud computing: "Do not trust the cloud".
Rule #1 is "You get the level of service that you pay for."
(OK, rule #1 is really "Don't get caught" but you know what I mean.)
Why is Google even able to review the content? Content should be encrypted.
Because they're a very cheap provider. Pay more and you'll get the sort of service that you're asking for. Do you want the money or do you want the privacy?
Things like RSA encryption are good examples of patents.
... which was so novel, it had already been independently developed.
And it's a good thing that the number theory it relies on wasn't patented by Euler.
The innovative thing was using the pieces together to create a piece of software that could encrypt and decrypt messages without using an explicit shared secret. That greatly changed things, and it's things like that which merit patenting (or other protection). Whether it was RSA or a predecessor that did it is not so important. Whoever it was, allowing a patent on the idea was reasonable.
The problem with software patents is that they're too often awarded without making clear just what is going on ("making the method patent"; it's the original definition of the word) and are too often awarded for trivial things, things that would actually be obvious extensions of the idea to someone who is a practitioner in the art of programming. Those issues (plus general problems like submarine patents) turn reasonable protection of real innovation that advances the state of human thought and productivity into a tool of oppression. Had the patent office only awarded a small number of software patents on the really innovative stuff, we wouldn't be having this argument now.
it doesn't even need an author with a cheesy villan laugh.
But I like the laugh. I've practiced it so much. :-(
we live in unpredictable London (hot-ish in summer, fscking cold in winter)
This is London, Ontario? The other one doesn't get cold in winter; hardly ever drops below freezing.
But in an application where the lights stay on all night, regular CFL bulbs work fine in the coldest days of winter.
Why are you having lights on outside all night? To give the local wildlife something to see by? To make it easier for burglars to find their way across your property? You could just switch the lights off during the hours when you're asleep (using a cheap timeswitch) and save a bunch at practically zero cost to yourself.
Remember the goal is lower CO2 emissions.
That's only one of the goals. Not having to replace broken bulbs so often is another good goal, and we can justify that on pure convenience for ourselves. (You could also make points about reduced cost of transport, etc., but you'd just be skirting round the central fact of it just being better on the laziness factor.)
If you actually pay to have a single CFL delivered to your door, does that count against your carbon footprint?
You buy lightbulbs one at a time?
Isk is also the in-game currency of EVE Online.
Look there! Up, in the sky! It's a bird! It's a plane!
W h o o s h !
No, it's the joke going over your head.
I really wouldn't steak 'isk' if I were you. The Icelandic Kronor isn't worth nearly as much as it use to be.
The question is WHY THE FUCK DID IT MAKE SLASHDOT?
Let's see...
Looks like we've got an answer.
Sadly if we start shutting down corrupt businesses we will shut down the American economy.
If that's so, then the economy deserves to be shut down. There's no reason at all to ever tolerate crooked businesses. (And no, I don't think that shutting them down will break the economy; most people - even most business owners - are honest.)
Off the top of my head, your applications will run on OS X without having to add Aqua specific extensions, people won't ignore your application just because it's "java" and you won't get a plethora of users moaning that it's "slow".
The main issue for OSX GUI apps vs. GUI apps on other platforms is that apps tend to be laid out differently on OSX. So while you can reuse your layouts from other platforms, they look out of place. (By contrast, Windows and Linux tend to use much more similar GUI layout strategies; their GUI heritages diverged much more recently.)
There's one thing that computer code and natural language text have in common: For both, confusing "writing" with "typing" is moronic.
Sounds like you're not management material!
This kind of applications won't have the hardware catching up to let you replace C, Asembly and VHDL with Ruby or Java for decades yet.
I have contacts in the embedded safety-critical world (though admittedly they're in a domain where there's lots of power available). They're using high-level languages. I wonder whether it is a good idea for them to do so, but they seem to be happy and productive so the risk-reward tradeoff must seem OK to them.
I suspect that if you only ever use low-level languages then you only ever really deal with low-level details, but a program in total - including runtime and libs - will always have both low-level and high-level concerns. High-level languages let you get to working on the high-level problems sooner; the low-level stuff is left to the runtime and libs which are (usually) well-written enough to take care of it without messing. (For a simple example, why would you write your own sorting code when you could just use one from a library where they've picked a good algorithm and dealt with the complex edge-cases already? There are cases where you would, but most coders aren't writing those things.)
On the desktop? Sure, who cares, though again a lot of java apps are slow to start and sluggish to respond.
That's the basic startup cost of the java runtime, which is a bloated pig. But if you compare a complex app startup then the costs aren't that different (opening databases and initializing security contexts can take quite a while...)
Java does best for domains where it can stay resident, such as a complex server. Writing those in C is a real nightmare - if one module gets its memory management wrong then the whole lot goes up in smoke and you've got almost no hope to find out where, especially for rare errors - whereas it's a lot more practical in a managed language like Java or C# (or Python or ...) I'll reserve judgement there on C++ (some people think it is suited, others don't).
You do know that reference counting is a (simple and limited) form of garbage collection?
You do know that refcounts work fine as long as you don't have reference loops? They even tend to give you better memory release behavior, though the release code has to be written carefully; the obvious recursive algorithm can blow up on large trees by blowing the stack. (Of course, as long as you know that this is possible it's pretty easy to avoid.)
The gain of "real" GC is that it handles the complex cases better, but you pay for that. It's often worthwhile paying that "tax".
As with most things, the USA is still using the same system as the UK, and it's only mainland Europe that has them reversed. I'm not sure 'God' has anything to do with it.
Of course He has something to do with it. As everyone of note knows, God is an Englishman. (There's even some evidence.)
I strongly suspect that the manufacturing capacity of Europe is rather a lot smaller than the total EU-wide demand for consumer goods (eg. childs toys). It follows that if the EC were to receive such a letter, they couldn't very well respond by saying "Fine. We'll embargo all your products" for very long - they'd drive the prices for a lot of items up so high that the politicians in member states would have Hell to pay.
Ah, but remember that they can blame it on those horrible Chinese and dastardly unelected Commission so they look blameless. It's very useful for the national-level politicians to have someone else to take the unpopular decisions for them so they can focus on taking decisions that people will vote for...
The UK allows a couple of extra channels (12 and 13) that I believe you can't use in the US. It might be necessary to change some settings to allow use of these channels.
However, few APs actually use these channels, so it's unlikely to be much of an issue.
It's especially not an issue because he's not bringing an access point (that would be a silly thing to bring for a two-week trip). Almost all APs are on channels 1, 6 and 11, the same as virtually everywhere else in the world; if you find another in use, it's almost worth posting a story on /. about it...
UK is 230V, as is most of the rest of Europe. At one time it was 240V, but voltage rates have been harmonized. Hence the reason many devices have been required to support 220-240V to facilitate that harmonization.
Only nominally. In practice, what's happened is that devices have had to cope with wider ranges of voltages than before. For switched-mode power supplies, this is not an issue (they run hotter on the higher voltage than in North America, but not so much that it's a big issue) and you won't be bringing the straight resistive-load devices where this would be a problem. After all, why would you bring a kettle when you can get one for hardly anything in any supermarket? Hotels that have razor sockets in the bathroom will also allow you to control the voltage delivered there.