There's an old addage in the IT industry about a server that crashed every night at 7pm. It crashed because the janitor unplugged the server and plugged in his vacuum.
Reminds me of a problem I had with NT for a while. I would come in to work in the morning and a large fraction of the time see the blue screen of death. Eventually I figure out that it happened when Norton Anti-Virus was starting an autoscan. Every time the autoscan started, up came the blue screen. If I ran it manually it behaved fine.
Many people don't realize that the relatively modern addition of variable speed windshield wipers were invented in the early part of the previous century. I forget the exact year. Now, however, that the patent has expired this is a standard feature on most automobiles.
It appeared on most automobiles long before the patent expired. A family friend sat on the jury when it went to trial with one of the big auto companies. According to this: http://www.inc.com/magazine/19971201/1374.html he's collected about $30M so far from Ford and Chrysler.
Cows can stand in a variety of poses allowing their center of mass to be in a variety of position with respect to their hooves
Yeah, the best time to get them is when they're doing "tree" pose and are just balanced over one leg, or "standing bow" when they're on one leg and are hanging most of their mass way out front.
With an M2 you can probably sink the boat pretty easily even if you miss all the people in it.
Unless it's an armored (and actually armored, not just mild steel) boat, an M2 on board the ship, firing from elevation, is going to make the small boat a very unpleasant place to be.
The story made CNN-- I saw it on the monitor in the airport on Saturday evening. They did point out that malicious programs could use the DRM to hide in your system.
According to CNN Sony claims that the future version won't cloak the files, but nothing about how to deal with it if you're already infected with it.
About half the fatal "accidents" involve alcohol, often not consumed by the people who die in the accident. Just eliminating drunk driving would save ~20,000 lives/year. The problem seems to be that americans like to drive drunk and so won't support laws that actually keep drunk drivers off the road.
Of the remaining half (and this part is educated speculation) a large fraction are caused by things like excess speed, sleep deprivation, and distractions like trying to do other things that require attention while driving.
But yeah, I spend a lot more time avoiding bad drivers on the road, and very little worrying about terrorist attacks.
Google is working closely with several major libraries to do this. Do you think those libraries aren't actively involved? It's not like Google is sneaking in with a photocopier and making illicit copies- the universities are active participants and have, no doubt, had their lawyers look over everything pretty carefully.
I'm sure one of the incentives for their participation is that they'll get backups of their entire libraries at little or no cost to them. A fire in a library can be devastating, and would likely leave the libraries with zero copies of many of the books. In addition to this, the libraries are getting indexes of the contents that are *way* better than the old subject cards in the old card catalogs. Without the G-factor, this is a library user's wet dream. With it, it offers the same index to everyone in the world and leaves it to individuals to get ahold of the books.
The books were all legally obtained by the libraries Google is collaborating with.
The copy Google is making isn't "sitting on the shelves". There's a small window to it in the "card catalog" that lets you see the couple of sentences (or maybe pages) relevant to what you're searching on.
Re:You are confusing two issues
on
Reining in Google
·
· Score: 5, Insightful
By serving up snippets they are able to make sales of advertising. Therefore Google are using entire copies of copyrighted material for commercial benefit. This is so far from fair use it's not debatable.
Yes, it is debatable. They aren't republishing the entire book.
If their potentially infringing use decreases the value of the original (i.e. makes a user of gBooks not want to get a copy of the original) then it's probably infringing. If their use increases the likelihood that someone will want to get the original, then it's probably not infringing.
Book reviews depend on this-- they excerpt things all the time without needing permission. The whole point of book reviews is to help evaluate whether you want to get a book or not. They also make a profit from republishing snippets, but it's certainly not infringing. Authors don't see a penny of that unless it encourages someone to go out and buy the book.
What Google is doing is providing customized excerpts based on what you're looking for. In all probability it will increase sales for books that it indexes-- there are plenty of times that I haven't bought a book online because I couldn't tell if it included what I wanted. Then I check it out in person at a bookstore if it's available. If it's not, then I just don't buy it.
And as a side note, bookstores already provide a service that allows you to do what people are afraid Google will enable, and possibly for less effort. If you're a reasonably quick reader it's not hard to read an entire book (or at least as much as you want) in a single bookstore visit. There are plenty of books that I've read in the bookstore that I didn't buy.
Making full copies may be irrelevant- libraries routinely make full copies of damaged books that are out of print in order to preserve their content.
Copyright wasn't intended to give authors absolute control over the ability to profit from their content. It strikes a balance (at least it used to) between protecting authors' ability to profit and the benefit to society of having things made publicly available.
I'll spell this out even more clearly. I have written book X recently...[lots of stuff cut out for space]... And there's certain no law that compels me to accept free advertising in recompense for abuse of my copyright.
Ok, so suppose your Book X is on the shelf at the library and I decide to read it there and write a review of it. I pull out a couple of excerpts and include them in the review. I sell this review to some book of reviews, who then publish it with ads. They and I have both profited from your book, and you don't see a penny (except from the single sale to the library). That's free advertising for your book, and there's nothing you can say or do about it- it falls under fair use. I can even say your book sucks in the review, and it's still fair use.
In fact, I just did something like that in this post-- if you read "The Fine Print" at the top of all the replies, all your comments belong to you. I just excerpted your copyrighted material (I could even have fairly included the whole thing, but it would have been annoying for other readers) to write a negative review of it, and Slashdot profits from it.
On the mac if you exited to to debugger and it said it died in "Jackson" it meant that you had so munged the memory that it looked like a Jackson Pollock painting.
But pointers rule. They're hidden in other languages (pass by reference vs. pass by value), but they're incredibly useful and fun, too.
Memory management is just something you have to treat like parentheses.
Perl is great for instant gratification, but the varieties of syntax for saying exactly the same thing can produce more than a little confusion, along with the overloading of symbols in strange ways (e.g. things with different meanings inside square brackets vs. outside them). Python might be good though- I haven't gotten around to looking at it because there's a bunch of stuff in CPAN that does stuff I really want, and it's easier to improve my Perl skills than try to translate or regenerate all that in Python.
First, start with a full-featured interpreted language like Python, Perl, Ruby, or Scheme.... (much deleted)...Now you should learn Java. The main reason for this is that you will need to learn C, or at least a C-like language
Yikes! I think I'd recommend doing it in the opposite order. Certainly Perl is not a good first language (though it may be a good last language)--it's not terribly readable, and even when you're fairly experienced it can take more than a little digging through reference books to figure out what some code is maybe supposed to do. I can't really comment on Python, Ruby, or Scheme.
C is a good first language, because there's really not much to the core of it and it's very close to the machine. Java might not be bad either-- I went through some tutorials to learn it but never got around to trying to do anything with it (didn't have anything I wanted to program in it). It deals pretty well with the things that make C++ icky. If you're going to go OO, it's probably a good place to start.
As another poster wrote, once you have a good sense of how to program and have developed a style, Perl is probably a decent way to go. It takes care of all the icky background stuff that you otherwise have to do yourself in something like C, but if you have a decent background in a low level language it's easier to understand what Perl is hiding from you and where you might write a couple lines that take the lifetime of the universe to execute.
My favorite is writing text editors and calculators.
Some things that are more fun to write and play with are cellular automata. Start with implementing John Conway's life, and then try changing the rules or extending it into 3-D.
Another fun one is CoreWars (http://www.corewars.org/). It requires that you write a machine emulator that executes code in a small (~10ish) instruction assembly language with maybe 4 addressing modes and the option to split your program into multiple threads. You also have to write an assembler (or run your emulator as an interpreter), and back in the day you also had to write your own program editor. It's also nice, but not critical, to have a graphical display. I haven't kept up with the current specification, but it wasn't too hard to implement the original spec, and it then gives you a simple assembler playground to mess around in, too.
The OP may or may not have whole house AC. If there is whole house AC unit is probably more efficient and quieter than a small unit for local cooling. The small unit also has to dump its waste heat somewhere, and that's likely into the house. So now the house AC has to cool the waste heat of the little AC, which will be more than the heat of the servers.
Louvers probably won't let that much noise back in, and a little baffling can probably take care of it if it's more than the OP wants. I've been in plenty of recording studios, and spent more than a little time doing acoustic isolation (more for lab purposes with sensitive instruments but some for recording). It probably doesn't have to be studio quality, and I've seen a simple plywood box (with various holes for venting and feedthroughs) do a fine job of reducing noise from a loud pump or compressor without going to a lot of trouble to make baffles at the ports.
And if the OP is actually using them as servers rather than consoles, whenever the door is closed the monitors are off and the light is off. hubs and switches for in the house are going to be in the neighborhood of 20 W each.
If the AC is in the closet it has to be one of the little mobile "local" AC units that has an exhaust vent tube that goes somewhere else to take all the heat it generates away, otherwise it's just a heater. It would be a minor violation of thermodynamics to expect the AC to exhaust the waste heat into the closet and try to remove that heat with the AC. A window type AC unit won't work (unless the back is hanging out the closet door, defeating the purpose because it's going to make more noise than the servers).
1200-1300 W should be no problem for a vent with a fan pulling air out, and another vent for makeup air from the house.
As long as you're going to have a vent hole anyway
Since the OP said that the closet used to house the furnace, there's probably already a chimney anyway. If it was an old furnace it actually lost some heat out the chimney, some of the fanciest new ones have cold exhaust, but either way, there's probably a duct ~5-6" in diameter that goes to the outside. It may have been plugged or blocked when the furnace was removed, but it's probably pretty easy to get to.
Since the servers probably don't produce all that much heat, either just let the heat exhaust itself passively by putting some louvers in the door to supply makeup air, or maybe add a fan to force air out the duct. Either way, it probably doesn't require an air conditioner, which would probably use several times as much power as the servers.
They started out with a squad level WWII game (Combat Mission) that takes place on the western front and uses a really interesting (and effective) WEGO game play model. The game is everything that Avalon Hill's Squad Leader boardgame wanted to be but couldn't, because of the immense complexity of the Squad Leader rules.
Battlefront/Big Time Software started out distributing only via online purchase (with CD sent in the mail) and were spectacularly successful for an indie game. Despite selling smaller numbers, they seemed quite happy with the financial returns (the principals in the company have long experience as game developers for other companies), and have released a couple of sequels, plus published games for several other developers, and are working on a new, more powerful game engine.
To top it all off, they release for Mac and PC at the same time.
A couple of the things in their formula for development that I think made a big difference: 1) the guys developing it are game players, as well as developers, and developed a game they wanted to play, first and foremost. 2) they developed a great game first, and worried about the eye candy later. Eye candy might help sales up front (wow! you can see where the bolts on that truck were rounded with a wrong sized wrench!) but game play and repeat playability is what keeps the game selling. 3) they developed a community on their message boards and really listened and responded to comments and questions. During the beta days they were very active on the boards. As it got closer to release time they were less active, but when they showed up they gave really good information about what was going on. They've continued like this for subsequent releases. 4) they didn't promise what they didn't intend to deliver. If they weren't going to put something in that people wanted, they generally said so, and often explained why. 5) they had great advance stuff to show off the game. They showed bits from an actual game, with comments by the players, even at the alpha stage. They released a fully functional beta for free, with a couple scenarios, but no editor. The beta had some bugs, and some things that just weren't quite right, but I ordered in advance after I realized that even if all they did was ship the beta plus a scenario editor I was going to enjoy it for a long time. Even with only two scenarios the two player play was good enough that people played them for months against various opponents and never tired of it. They got a ton of good feedback from the beta, and took advantage of all of it to improve the game.
Right now, the people are cheaper and do more than robots.
Huh?
Space is has been pretty well commercialized out to LEO already, and it's almost all robotic. Telecom sats? Robotic. Satellite imagery? Robotic. The satellites that bring you TV (yeah, they fall under telecom)? Robotic.
If corporations needed to put people in space, they would. So far, everything commercial except tourism hasn't required corporations to put people in space. And the people who want to sell tickets on tourist flights? They're either using foreign launchers or developing their own. NASA limits space tourism to their own astronauts and the occasional senator or teacher, but if you just want to buy a ride to space you're not going to get NASA support, and you probably can't get Boeing or Lockheed to sell you a ride either. NASA will even be a little peevish about it when the Russians sell you a ticket.
If you want to go beyond LEO it's *way* cheaper to use robots--it's expensive to lift a lot of mass out of earth orbit, and you can make very capable robots quite small. People don't get very small (except steve martin), and need a lot more than a solar array and a battery to keep them alive for 5 years.
The robotic missions are generally substantially less expensive. The problem with manned missions is the amount of expense that goes into keeping the people alive. The people themselves are massive, and the life support systems still more massive, and the redundant systems still more massive. All that mass adds up to $$$, especially since things have to be higher reliability because it looks *really* bad when people die in space, but when all you've lost is some metal and electronics people don't get as upset.
It's much cheaper to use robots, even if you build two on the expectation that you'll lose one.
Except that using the moon blocking the light (as in an eclipse) isn't a good analogy for a nuller. The nulling interferometer doesn't have to put in anything to block the light-- it adjusts the relative phase of light on two different paths so that the on-axis light cancels out, but the off axis light doesn't. There are different instruments that work more like an eclipse, where a stop is used to block the startlight but not the planet light.
The technology required to combine two light beams in a coherent way is wa-a-a-y more expensive than a "cheap" telescope.
On the plus side, if you want a large aperture to get high resolution, an interferometer with a long baseline will be cheaper than a telescope with a similar resolution. There's a lot less glass to grind, and unless you're trying to make a Fizeau interferometer, you need high precision in fewer places.
But you know what? Science and engineering are hard. That's the honest truth.
That reminds me of when I was in TA training-- we actually had a pretty rigorous training for the TAs for the service courses, and the courses themselves were designed much better than the stereotype, and I watched a lot of people learn physics that they might not have otherwise. The prof who was teaching the course was talking about how the students feel much of the time in the class, and said something along the lines of "They get really frustrated, and they don't like it. They're not like us, we go into physics because we enjoy being frustrated."
That's not entirely why we do it, but when it comes down to actually doing science and working at the edge of what's known and not known (and which may or may not be of interest to anyone besides the researcher in question) you spend a lot of time being frustrated. If you aren't, then you're probably not at the edge. But that's where the fun is.
I was on the fence whether to mod or post on this thread, but you just tipped me to post.
You're right that engineering schools in general aren't conducive to learning much liberal arts/history/whatever (though some may do a decent job of it). Science curricula, however do allow for more of the liberal artsy stuff, and will let you go into engineering later if you want, or something liberal artsy (with maybe a technical twist) later. I did physics (and eventually went on to a PhD in it) and managed to study abroad, take history and lit classes, be involved in extracurricular stuff that I'm still glad I did 20 years later. My current job is on the line between science and engineering (tips back and forth), but also occasionally benefits a lot from my liberal education.
A friend did a similar thing, doing chemistry and art history, and uses knowledge of both as a preservationist.
Another friend did biochemistry, also managed to spend a year abroad as an undergrad, is extremely well rounded in science and literature , and went on to a PhD in physics and is now a professor of engineering.
So my advice, if you like science and engineering and technical things, but also like the "soft" stuff, is to go science. Some schools even (in my opinion correctly) put science in the same college as literature and arts, rather than with engineering. Science (the real deal, with calculus and all) is as much a part of a good liberal arts education as art and literature are. If you go with a non-science major, getting into an engineering job or grad school could be hard, and if you go into engineering it could be tough to get into a non-engineering field. If you do a science major with a strong emphasis in a non-science thing, you can probably go either way.
If you want my opinion as to what science will be hot for a long time, it would be neuroscience, but you'll be better at it if you do it in a physics or chemistry (or electrical engineering) department rather than a dedicated neuroscience dept.
Slashdot Mirror
There's an old addage in the IT industry about a server that crashed every night at 7pm. It crashed because the janitor unplugged the server and plugged in his vacuum.
Reminds me of a problem I had with NT for a while. I would come in to work in the morning and a large fraction of the time see the blue screen of death. Eventually I figure out that it happened when Norton Anti-Virus was starting an autoscan. Every time the autoscan started, up came the blue screen. If I ran it manually it behaved fine.
The vote would be unflagged if the voter provided a photo id at any point after the vote.
Except then you no longer have secret balloting if you can connect people back to their votes after they've been cast.
Many people don't realize that the relatively modern addition of variable speed windshield wipers were invented in the early part of the previous century. I forget the exact year.
Now, however, that the patent has expired this is a standard feature on most automobiles.
It appeared on most automobiles long before the patent expired. A family friend sat on the jury when it went to trial with one of the big auto companies. According to this: http://www.inc.com/magazine/19971201/1374.html he's collected about $30M so far from Ford and Chrysler.
Cows can stand in a variety of poses allowing their center of mass to be in a variety of position with respect to their hooves
Yeah, the best time to get them is when they're doing "tree" pose and are just balanced over one leg, or "standing bow" when they're on one leg and are hanging most of their mass way out front.
With an M2 you can probably sink the boat pretty easily even if you miss all the people in it.
Unless it's an armored (and actually armored, not just mild steel) boat, an M2 on board the ship, firing from elevation, is going to make the small boat a very unpleasant place to be.
The story made CNN-- I saw it on the monitor in the airport on Saturday evening. They did point out that malicious programs could use the DRM to hide in your system.
According to CNN Sony claims that the future version won't cloak the files, but nothing about how to deal with it if you're already infected with it.
Yes, car accidents are by definition accidents.
Sort of.
About half the fatal "accidents" involve alcohol, often not consumed by the people who die in the accident. Just eliminating drunk driving would save ~20,000 lives/year. The problem seems to be that americans like to drive drunk and so won't support laws that actually keep drunk drivers off the road.
Of the remaining half (and this part is educated speculation) a large fraction are caused by things like excess speed, sleep deprivation, and distractions like trying to do other things that require attention while driving.
But yeah, I spend a lot more time avoiding bad drivers on the road, and very little worrying about terrorist attacks.
Google is working closely with several major libraries to do this. Do you think those libraries aren't actively involved? It's not like Google is sneaking in with a photocopier and making illicit copies- the universities are active participants and have, no doubt, had their lawyers look over everything pretty carefully.
I'm sure one of the incentives for their participation is that they'll get backups of their entire libraries at little or no cost to them. A fire in a library can be devastating, and would likely leave the libraries with zero copies of many of the books. In addition to this, the libraries are getting indexes of the contents that are *way* better than the old subject cards in the old card catalogs. Without the G-factor, this is a library user's wet dream. With it, it offers the same index to everyone in the world and leaves it to individuals to get ahold of the books.
The books were all legally obtained by the libraries Google is collaborating with.
The copy Google is making isn't "sitting on the shelves". There's a small window to it in the "card catalog" that lets you see the couple of sentences (or maybe pages) relevant to what you're searching on.
By serving up snippets they are able to make sales of advertising. Therefore Google are using entire copies of copyrighted material for commercial benefit. This is so far from fair use it's not debatable.
Yes, it is debatable. They aren't republishing the entire book.
If their potentially infringing use decreases the value of the original (i.e. makes a user of gBooks not want to get a copy of the original) then it's probably infringing. If their use increases the likelihood that someone will want to get the original, then it's probably not infringing.
Book reviews depend on this-- they excerpt things all the time without needing permission. The whole point of book reviews is to help evaluate whether you want to get a book or not. They also make a profit from republishing snippets, but it's certainly not infringing. Authors don't see a penny of that unless it encourages someone to go out and buy the book.
What Google is doing is providing customized excerpts based on what you're looking for. In all probability it will increase sales for books that it indexes-- there are plenty of times that I haven't bought a book online because I couldn't tell if it included what I wanted. Then I check it out in person at a bookstore if it's available. If it's not, then I just don't buy it.
And as a side note, bookstores already provide a service that allows you to do what people are afraid Google will enable, and possibly for less effort. If you're a reasonably quick reader it's not hard to read an entire book (or at least as much as you want) in a single bookstore visit. There are plenty of books that I've read in the bookstore that I didn't buy.
Making full copies may be irrelevant- libraries routinely make full copies of damaged books that are out of print in order to preserve their content.
Copyright wasn't intended to give authors absolute control over the ability to profit from their content. It strikes a balance (at least it used to) between protecting authors' ability to profit and the benefit to society of having things made publicly available.
I'll spell this out even more clearly. I have written book X recently...[lots of stuff cut out for space]... And there's certain no law that compels me to accept free advertising in recompense for abuse of my copyright.
Ok, so suppose your Book X is on the shelf at the library and I decide to read it there and write a review of it. I pull out a couple of excerpts and include them in the review. I sell this review to some book of reviews, who then publish it with ads. They and I have both profited from your book, and you don't see a penny (except from the single sale to the library). That's free advertising for your book, and there's nothing you can say or do about it- it falls under fair use. I can even say your book sucks in the review, and it's still fair use.
In fact, I just did something like that in this post-- if you read "The Fine Print" at the top of all the replies, all your comments belong to you. I just excerpted your copyrighted material (I could even have fairly included the whole thing, but it would have been annoying for other readers) to write a negative review of it, and Slashdot profits from it.
Yeah, I learned it back then, too.
On the mac if you exited to to debugger and it said it died in "Jackson" it meant that you had so munged the memory that it looked like a Jackson Pollock painting.
But pointers rule. They're hidden in other languages (pass by reference vs. pass by value), but they're incredibly useful and fun, too.
Memory management is just something you have to treat like parentheses.
Perl is great for instant gratification, but the varieties of syntax for saying exactly the same thing can produce more than a little confusion, along with the overloading of symbols in strange ways (e.g. things with different meanings inside square brackets vs. outside them). Python might be good though- I haven't gotten around to looking at it because there's a bunch of stuff in CPAN that does stuff I really want, and it's easier to improve my Perl skills than try to translate or regenerate all that in Python.
First, start with a full-featured interpreted language like Python, Perl, Ruby, or Scheme.... ...Now you should learn Java. The main reason for this is that you will need to learn C, or at least a C-like language
(much deleted)
Yikes! I think I'd recommend doing it in the opposite order. Certainly Perl is not a good first language (though it may be a good last language)--it's not terribly readable, and even when you're fairly experienced it can take more than a little digging through reference books to figure out what some code is maybe supposed to do. I can't really comment on Python, Ruby, or Scheme.
C is a good first language, because there's really not much to the core of it and it's very close to the machine. Java might not be bad either-- I went through some tutorials to learn it but never got around to trying to do anything with it (didn't have anything I wanted to program in it). It deals pretty well with the things that make C++ icky. If you're going to go OO, it's probably a good place to start.
As another poster wrote, once you have a good sense of how to program and have developed a style, Perl is probably a decent way to go. It takes care of all the icky background stuff that you otherwise have to do yourself in something like C, but if you have a decent background in a low level language it's easier to understand what Perl is hiding from you and where you might write a couple lines that take the lifetime of the universe to execute.
My favorite is writing text editors and calculators.
Some things that are more fun to write and play with are cellular automata. Start with implementing John Conway's life, and then try changing the rules or extending it into 3-D.
Another fun one is CoreWars (http://www.corewars.org/). It requires that you write a machine emulator that executes code in a small (~10ish) instruction assembly language with maybe 4 addressing modes and the option to split your program into multiple threads. You also have to write an assembler (or run your emulator as an interpreter), and back in the day you also had to write your own program editor. It's also nice, but not critical, to have a graphical display. I haven't kept up with the current specification, but it wasn't too hard to implement the original spec, and it then gives you a simple assembler playground to mess around in, too.
The OP may or may not have whole house AC. If there is whole house AC unit is probably more efficient and quieter than a small unit for local cooling. The small unit also has to dump its waste heat somewhere, and that's likely into the house. So now the house AC has to cool the waste heat of the little AC, which will be more than the heat of the servers.
Louvers probably won't let that much noise back in, and a little baffling can probably take care of it if it's more than the OP wants. I've been in plenty of recording studios, and spent more than a little time doing acoustic isolation (more for lab purposes with sensitive instruments but some for recording). It probably doesn't have to be studio quality, and I've seen a simple plywood box (with various holes for venting and feedthroughs) do a fine job of reducing noise from a loud pump or compressor without going to a lot of trouble to make baffles at the ports.
And if the OP is actually using them as servers rather than consoles, whenever the door is closed the monitors are off and the light is off. hubs and switches for in the house are going to be in the neighborhood of 20 W each.
If the AC is in the closet it has to be one of the little mobile "local" AC units that has an exhaust vent tube that goes somewhere else to take all the heat it generates away, otherwise it's just a heater. It would be a minor violation of thermodynamics to expect the AC to exhaust the waste heat into the closet and try to remove that heat with the AC. A window type AC unit won't work (unless the back is hanging out the closet door, defeating the purpose because it's going to make more noise than the servers).
1200-1300 W should be no problem for a vent with a fan pulling air out, and another vent for makeup air from the house.
As long as you're going to have a vent hole anyway
Since the OP said that the closet used to house the furnace, there's probably already a chimney anyway. If it was an old furnace it actually lost some heat out the chimney, some of the fanciest new ones have cold exhaust, but either way, there's probably a duct ~5-6" in diameter that goes to the outside. It may have been plugged or blocked when the furnace was removed, but it's probably pretty easy to get to.
Since the servers probably don't produce all that much heat, either just let the heat exhaust itself passively by putting some louvers in the door to supply makeup air, or maybe add a fan to force air out the duct. Either way, it probably doesn't require an air conditioner, which would probably use several times as much power as the servers.
also check out http://battlefront.com/
They started out with a squad level WWII game (Combat Mission) that takes place on the western front and uses a really interesting (and effective) WEGO game play model. The game is everything that Avalon Hill's Squad Leader boardgame wanted to be but couldn't, because of the immense complexity of the Squad Leader rules.
Battlefront/Big Time Software started out distributing only via online purchase (with CD sent in the mail) and were spectacularly successful for an indie game. Despite selling smaller numbers, they seemed quite happy with the financial returns (the principals in the company have long experience as game developers for other companies), and have released a couple of sequels, plus published games for several other developers, and are working on a new, more powerful game engine.
To top it all off, they release for Mac and PC at the same time.
A couple of the things in their formula for development that I think made a big difference:
1) the guys developing it are game players, as well as developers, and developed a game they wanted to play, first and foremost.
2) they developed a great game first, and worried about the eye candy later. Eye candy might help sales up front (wow! you can see where the bolts on that truck were rounded with a wrong sized wrench!) but game play and repeat playability is what keeps the game selling.
3) they developed a community on their message boards and really listened and responded to comments and questions. During the beta days they were very active on the boards. As it got closer to release time they were less active, but when they showed up they gave really good information about what was going on. They've continued like this for subsequent releases.
4) they didn't promise what they didn't intend to deliver. If they weren't going to put something in that people wanted, they generally said so, and often explained why.
5) they had great advance stuff to show off the game. They showed bits from an actual game, with comments by the players, even at the alpha stage. They released a fully functional beta for free, with a couple scenarios, but no editor. The beta had some bugs, and some things that just weren't quite right, but I ordered in advance after I realized that even if all they did was ship the beta plus a scenario editor I was going to enjoy it for a long time. Even with only two scenarios the two player play was good enough that people played them for months against various opponents and never tired of it. They got a ton of good feedback from the beta, and took advantage of all of it to improve the game.
...selling home autoclaves. Convince everyone they have to autoclave their pillows and bedding regularly.
Right now, the people are cheaper and do more than robots.
Huh?
Space is has been pretty well commercialized out to LEO already, and it's almost all robotic. Telecom sats? Robotic. Satellite imagery? Robotic. The satellites that bring you TV (yeah, they fall under telecom)? Robotic.
If corporations needed to put people in space, they would. So far, everything commercial except tourism hasn't required corporations to put people in space. And the people who want to sell tickets on tourist flights? They're either using foreign launchers or developing their own. NASA limits space tourism to their own astronauts and the occasional senator or teacher, but if you just want to buy a ride to space you're not going to get NASA support, and you probably can't get Boeing or Lockheed to sell you a ride either. NASA will even be a little peevish about it when the Russians sell you a ticket.
If you want to go beyond LEO it's *way* cheaper to use robots--it's expensive to lift a lot of mass out of earth orbit, and you can make very capable robots quite small. People don't get very small (except steve martin), and need a lot more than a solar array and a battery to keep them alive for 5 years.
The robotic missions are generally substantially less expensive. The problem with manned missions is the amount of expense that goes into keeping the people alive. The people themselves are massive, and the life support systems still more massive, and the redundant systems still more massive. All that mass adds up to $$$, especially since things have to be higher reliability because it looks *really* bad when people die in space, but when all you've lost is some metal and electronics people don't get as upset.
It's much cheaper to use robots, even if you build two on the expectation that you'll lose one.
Maybe you didn't get the memo but people are getting sick of wars. It's time we move on, I'm not sure what's next
Rollerball.
The Nobel Committee does not want to impugn the integrity of the Prize by doling it out for science that does not pan out
Don't be so sure about that: Egaz Moniz won the prize in 1949 for the invention of the lobotomy.
Except that using the moon blocking the light (as in an eclipse) isn't a good analogy for a nuller. The nulling interferometer doesn't have to put in anything to block the light-- it adjusts the relative phase of light on two different paths so that the on-axis light cancels out, but the off axis light doesn't. There are different instruments that work more like an eclipse, where a stop is used to block the startlight but not the planet light.
The technology required to combine two light beams in a coherent way is wa-a-a-y more expensive than a "cheap" telescope.
On the plus side, if you want a large aperture to get high resolution, an interferometer with a long baseline will be cheaper than a telescope with a similar resolution. There's a lot less glass to grind, and unless you're trying to make a Fizeau interferometer, you need high precision in fewer places.
But you know what? Science and engineering are hard. That's the honest truth.
That reminds me of when I was in TA training-- we actually had a pretty rigorous training for the TAs for the service courses, and the courses themselves were designed much better than the stereotype, and I watched a lot of people learn physics that they might not have otherwise. The prof who was teaching the course was talking about how the students feel much of the time in the class, and said something along the lines of "They get really frustrated, and they don't like it. They're not like us, we go into physics because we enjoy being frustrated."
That's not entirely why we do it, but when it comes down to actually doing science and working at the edge of what's known and not known (and which may or may not be of interest to anyone besides the researcher in question) you spend a lot of time being frustrated. If you aren't, then you're probably not at the edge. But that's where the fun is.
I was on the fence whether to mod or post on this thread, but you just tipped me to post.
You're right that engineering schools in general aren't conducive to learning much liberal arts/history/whatever (though some may do a decent job of it). Science curricula, however do allow for more of the liberal artsy stuff, and will let you go into engineering later if you want, or something liberal artsy (with maybe a technical twist) later. I did physics (and eventually went on to a PhD in it) and managed to study abroad, take history and lit classes, be involved in extracurricular stuff that I'm still glad I did 20 years later. My current job is on the line between science and engineering (tips back and forth), but also occasionally benefits a lot from my liberal education.
A friend did a similar thing, doing chemistry and art history, and uses knowledge of both as a preservationist.
Another friend did biochemistry, also managed to spend a year abroad as an undergrad, is extremely well rounded in science and literature , and went on to a PhD in physics and is now a professor of engineering.
So my advice, if you like science and engineering and technical things, but also like the "soft" stuff, is to go science. Some schools even (in my opinion correctly) put science in the same college as literature and arts, rather than with engineering. Science (the real deal, with calculus and all) is as much a part of a good liberal arts education as art and literature are. If you go with a non-science major, getting into an engineering job or grad school could be hard, and if you go into engineering it could be tough to get into a non-engineering field. If you do a science major with a strong emphasis in a non-science thing, you can probably go either way.
If you want my opinion as to what science will be hot for a long time, it would be neuroscience, but you'll be better at it if you do it in a physics or chemistry (or electrical engineering) department rather than a dedicated neuroscience dept.