The new Mac Mini is actually rated for "14 watts of power when idle", which Apple says is lower than any other computer in the EPA STAR database.
But wait, there's more! With Snow Leopard, Apple introduced a new "Wake on Demand" feature that could allow the Mini to be in sleep mode (~1.5W) most of the time, but still work as a server when needed:
(1) Airport base stations can now provide a "Bonjour Sleep Proxy" service that will keep announcing all the services your computer is hosting on the network, even after the computer goes to sleep. The base station will then wake the server whenever another computer tries to access it. It isn't clear whether this happens only for Bonjour services, or for any IP-based access (which should be possible in principle).
(2) All this magic can happen via the wireless network if your computer is new enough. (Wake-on-LAN was only possible via ethernet before.)
If the server is set to go back to sleep automatically after a few minutes of inactivity, this setup provides a low-power, always-on server arrangement: the computer sleeps most of the time, automatically wakes up whenever someone wants to connect to it, then goes back to sleep whenever it's not needed.
Bladeless designs may be becoming more popular for small, building-scale turbines, but they are not even on the horizon for large, commercial-scale turbines -- the sort that we could use for a large share of our electricity. As far as I know, every single utility-scale wind turbine on the market today uses a horizontal-axis design with 3 (or occasionally 2) blades. Their speed is lower in RPMs, but the tips are still moving around 150 mph.
And birds and bats do indeed fly into these things or get injured by the pressure changes they create. It may not be any more than would fly into cell-phone towers, power lines or other infrastructure, but as a "green" technology, wind power is held to a higher standard. (My personal theory is that high-flying birds and bats are looking for prey in an environment where there never used to be any risk of flying into anything, so they have never developed the ability to avoid objects ahead of them, especially moving objects). At any rate, there is definitely a need for creative ways to keep birds and bats away from wind turbines.
I just downloaded the authors' data and compared the time of the first mention of each meme in blogs versus the mainstream media. Surprisingly, the media make the first mention only 51.3% of the time. The rest of the time, the blogs had the first mention.
The blogs were ahead by 170 hours or more 25% percent of the time, trailed by 4 minutes or less 50% of the time, and trailed by 8 hours or less 75% of the time.
These statistics suggest that these blogs break a story before these mainstream media outlets about half the time. Sometimes the blogs are way ahead, but they are rarely far behind. The blogs have to get their news from somewhere, so it is surprising that they are ahead of (these particular) media outlets so often. Maybe they are pulling news from less-mainstream outlets (not included in this study) and building "buzz" around it, which then gets picked up by the mainstream news outlets?
The point is that a lot of people are claiming the MSM is obsolete and blogs are the way of the future... and this study pretty clearly shows that it isn't true.
I thought that would be the point of the story when I read it, but the story doesn't actually mention this issue at all. The researchers mostly seem to be interested in understanding how stories become popular, and the roles that blogs and traditional media play in that process.
In the original paper
(e.g., Figure 8), they report that there is a 2.5 hour lag between the peak of reporting on a story in the media in general and the peak of discussion in the blogs in general.
They also report the typical time lag for individual news outlets or blogs (Table 1), and show that a few individual blogs (e.g., hotair.com and talkingpointsmemo.com) have tend to report stories before individual media outlets. However, even this doesn't show that news appears in blogs before it appears in the media -- some individual blogs tend to report big stories before individual news outlets, but that may be because (a) they pull stories from many news outlets, so they will inevitably have an earlier average reporting time than any individual news outlet, and (b) the early-mover blogs play a role in determining which stories become popular, even if they aren't the first to report them.
Unfortunately, I didn't see any graph that tracked the earliest appearance of a story in any media outlet, and the earliest appearance of the same story in any blog, and compared the times of those appearances. That would be the way to really answer the question of who is reporting first. And I bet it's the media, by many hours.
There's not as much money in newspaper advertising as there used to be, and this will inevitably lead to a reduction in the amount of news being collected and the number of printed newspapers.
In the old days, your local newspaper(s) had a monopoly or oligopoly on display and classified advertising. This gave them enough money to hire a local reporting staff, and in some cases, to set up remote bureaus. The smaller papers relied on wire services or news agencies for their national or international news, and the bigger papers gathered some of this for themselves. Often there was enough advertising revenue to support two or more newspapers in the same town.
Now, many readers have switched from printed newspapers to the Internet, drying up the display ad revenue (newspapers make money selling readers to advertisers, not selling news to readers). The websites don't get nearly so much revenue per ad-view as the printed papers did. Meanwhile craigslist has grabbed the classified ads. So now each person who reads a story doesn't bring in nearly as much ad revenue as they used to. What's going to happen?
I think a big consolidation is inevitable -- the amount of original news reporting will have to be reduced, so that more people read each story, and the ad revenue per story returns to a high enough level to support the cost of writing it. All the newspapers will lose money for a while, until most of them have failed or radically restructured (e.g., going online-only and closing any remote bureaus). At that point, all national and international news will probably be gathered by a few national wire services, a few national TV networks, and maybe a couple of major national papers. All the other papers, websites and TV stations will rely on these sources for their "content". There will probably also be a big reduction in local reporting, except in the biggest cities. But each original story will be so widely disseminated that the revenue from teeth-whitening ads on Yahoo.com, Applebee's ads on sfgate.com, and Macy's ads in a few dozen local papers will be enough to cover the cost of reporting it.
This is depressing if you care about having a diversity of news sources, but it is probably unavoidable. There just isn't enough ad revenue to support as much news reporting as we have now.
Actually, only one digit is used for the checksum on most credit cards.
For a 16-digit number, there is a 6-digit issuer identification number (including 1-digit major-industry identifier), followed by 9 unique digits for the customer, followed by one check-digit. Some Visa numbers used to be 13 digits, which would have been much less unique, but those seem to have been converted to 16-digit numbers now (all this from Wikipedia).
My Motorola krzr k1m (verizon) used to have great battery life, except when running GPS with turn-by-turn navigation (i.e., continuous update of my GPS position). Then a new, fully-charged battery would run down within an hour, and the phone would get noticeably warm. Eventually I started using the navigator function to plan out the next few steps of my trip, then turned it off for a while, then repeated once I got closer to my destination.
Even with such efforts, GPS seemed to kill batteries quickly -- i.e., the battery physically expanded, and its recharge capacity dropped precipitously after a few months. Lithium-ion batteries don't like to run hot, so maybe that was a factor.
I was hoping that the 3G S had avoided these problems, but maybe they haven't...
I'm in the same boat. For years I've been looking for a file system can hold files larger than 2GB and can be mounted from Windows and Mac OS X (and maybe Linux). That would allow me to store all my work on one partition, and access it from Mac OS X, from Windows via Boot Camp, or from Windows inside Parallels or VMware Fusion. It would also allow me to transfer large files back and forth between my Mac and other Windows computers. I was hoping ZFS would be that file system.
The last time I checked (the middle of 2008), the only way to do this was via NTFS, and the only read-write support for NTFS on OSX was the MacFUSE NTFS driver, which was pretty slow.
I just saw that MacDrive 7.2 now allows Windows Vista x64 (my Boot Camp OS) to read HFS disks, so maybe I'll give that a try. There are also rumors that Snow Leopard's Boot Camp utility will include drivers for Windows to read HFS disks, so maybe that will help too.
You ignore the waste caused storing the energy during windy periods for not windy periods.
I assume that all the wind or nuclear power is used when it is generated (which is a realistic assumption for the foreseeable future). Storage (if it were needed for either of these) wouldn't radically change the land use, but might change the costs.
that is a 14,000acre area of land you are giving up
The wind farms might actually be placed on a larger area of land than that. But only a tiny fraction of this land is actually occupied by the towers and roads. The rest is available for farming, ranching, etc. Have you estimated the land use for uranium mining, tailings piles, waste storage, etc.?
And windfarms are noisy and ugly (so are nuclear plants but they aren't 80m tall.
New wind turbines are inaudible beyond a few dozen meters away. People may differ over whether wind farms are more ugly than uranium mines and nuclear power plants.
And if you want to spread those towers out more so you dont lose a 60km^2 chunk you will end losing more again because of efficiency loss.
Wind farms are highly modular -- yes it helps if they are near existing power lines, but it is also possible to build new power lines. And they certainly don't decrease in efficiency if they are built in multiple locations.
The real issue here is not land use anyway. It's what sort of vision we have for the future of the power system. Big nuclear plants, which require a lot of faith in cost forecasts and our ability to control waste (at the mine, mill, plant and afterward). Or wind and solar, which also require some faith in cost forecasts, may use more land area (but in a different way), and don't have the other risks. Intermittent sources require more ingenuity to integrate into the power system, but this is not insurmountable (people on slashdot should recognize that, if anyone). And don't overlook the fact that nuclear plants are as inflexible as wind or solar power -- they have to run around the clock, so if we want to use them for a large share of our power, we will have to find a way to use their night-time output (if you throw it away, that will kill the economics of the plant).
The 40 billion Yuan cost is not for one reactor; it is for two of the same kind.
OK, looks like I got that wrong, at least for this press release. But those costs do not seem to be in line with what U.S. nuclear plant developers expect to pay.
For example, this Wall Street Journal article (follow first link for full text) indicates that FPL Groups expects new AP1000's at its Turkey Point plant in Florida to cost $6-$9 billion each, and Georgia Power Co. expects a 45.7% stake in two similar reactors (i.e., 90% of one reactor) to cost $6.4 billion. (The first two reactors at Georgia Power's plant cost nearly $4.5 billion each 20 years ago, over 10 times the $330 million per plant originally estimated.)
So I'd be pretty skeptical of anyone who claims they can deliver nuclear power for less than $0.06-$0.10 per kWh. On the other hand, lots of wind power is being delivered at those costs or lower.
Texas passed a law in 1999 requiring 2 GW of wind power by 2009, but they keep exceeding their goal. By 2007 they had around 2.7 GW of wind capacity. Then they added another 1.6 GW in 2007, and another 2.7 GW in 2008, bringing the total to around 7 GW. This is driven hardly at all by the Texas requirements, a little bit by the federal tax incentives (around $0.01/kWh), and mostly by the fact that Texas has great wind sites and wind power is now competitive with other sources of electricity.
The industry has been growing by 30-40% per year worldwide, and that kind of growth has a way of sneaking up on you.
At a cost of $5.85 billion, and assuming a lifetime of 40 years, an interest rate of 6%, this nuclear plant will have an annual mortgage of $389 million. With a nameplate rating of 1100 MW, if it runs 92% of the time, it will produce 8.9 billion kWh per year, so the capital repayments will amount to $0.044/kWh, assuming it doesn't go over budget. Assuming an optimistic cost for fuel around $0.005/kwh, this gives a total cost of $0.049/kWh, neglecting the cost of maintenance, waste disposal, and any risk of contamination or weapons proliferation.
So, which would you rather spend $0.049/kWh on -- a nuclear plant that might go over budget, might leak radiation at some point during its life, whose waste will need to be carefully controlled and permanently stored somewhere that hasn't yet been identified; or a wind farm whose costs are much more certain and which comes without all those ancillary risks?
Yes, any individual wind farm will not provide a firm supply of power. But if a lot of wind farms are used, and they are combined with solar, geothermal and other renewable resources, they will provide a fairly stable power supply. There is also a lot of potential for reshaping electricity loads to match the supply of power (e.g., recharge electric vehicles when the wind is blowing or the sun is shining). And finally, if you must have a firm supply of power, you can convert a wind farm into a completely firm supply (at 35% of its nameplate rating) by spending about 10% extra and building rarely-used natural gas peaker plants ($634/kW * 35% = $222kW).
I don't normally go in for grammar corrections, but I can't resist here:
1) "pseudo"
2) you are -> you're
3) owned by it -> its (it is -> it's)
4) The quote you attribute to Thomas Jefferson was actually made by Gerald Ford. Nobody imagined "a government big enough to give you everything you want" in Jefferson's day.
Really, tinyurl has been around the entire 11+ years I've been on the internet, and somehow the internet's survived just fine.
The point is not what happens while tinurl.com is around, but what happens when it is not? Or when they decide to start charging for their services? "Free link redicrector forever" is not a sustainable business model.
I completely agree with you. I read GEB when I was 15 or 16, and it has changed how I see everything since then. It's an amazing introduction to the idea of self-referential systems, and once you get it, you see self-reference everywhere (makes a great introduction to chaos theory or self-modifying code as well).
I could see how some people would dislike the dialogues, but I loved them, especially the way the structure of the stories reflected the concepts the characters were discussing.
GEB also taught me to appreciate Escher and Bach, so maybe you'll get some thanks from the art and music departments too!
I'm kind of pissed that "a PC" means Windows... it means personal computer, does it not? Isn't my Linux machine a personal computer?
I think the "PC" thing dates back to the 1980s, when the IBM PC first came out. Other manufacturers figured out how to clone the IBM PC's BIOS (against IBM's will), and then started making "IBM PC-compatible" machines that could run the same software (i.e., MS-DOS and everything that ran under it). After a while "IBM PC-compatible" computers became "PC-compatible" and then just "PCs," and "PC" also took on the sense that the computer was running MS-DOS, or later, MS Windows. I guess the meaning of "PC" has now narrowed to usually mean only computers running MS Windows, but sometimes it still refers to the hardware, e.g., "Macintosh OS X runs on standard PC hardware". Macs, Linux and Windows machines are all PCs now in either the first sense ("personal computer", generically) or the second sense (able to run IBM PC binaries and their successors), but not the third sense (running Windows at any given moment).
I'm kind of pissed that "a PC" means Windows... it means personal computer, does it not? Isn't my Linux machine a personal computer?
I think the "PC" thing dates back to the 1980s, when the IBM PC first came out. Other manufacturers figured out how to clone the IBM PC's BIOS, and then started making "IBM PC-compatible" machines that could run the same software (against IBM's will). After a while "IBM PC-compatible" computers became "PC-compatible" and then just "PCs". Then when the field dwindled to Macs and "PCs", I guess "PC" came to mean Windows machines. But in either the first sense ("personal computer", generically) or the second sense (able to run IBM PC binaries and their successors) newer Macs, Linux and Windows machines are all PCs now.
Actually, this technique needs less corporate cooperation than you suggest.
1) The specification allows domain owners to report their signing policy in their domain DNS record. So Paypal can announce publicly that ALL mail from paypal.com must be signed. i.e., the senders can add themselves to the whitelist. If your e-mail client gets mail claiming to be from paypal.com, but without the proper certification, it knows with 100% certainty that it is junk.
2) It wouldn't take much for the big retail e-mail providers to start certifying all mail coming from their domains. Of course, they could only do this if they required their users to send outgoing e-mail using a secure connection to their own SMTP server. This would be annoying for people who like to send yahoo.com e-mail from their home SMTP server, but in the long run it's a better way to do it. This would move toward better symmetry between POP/IMAP and SMTP protocols (i.e., you use a secured logon to your e-mail provider for both, which is probably how it always should have been. It used to drive me crazy having to change my SMTP server for each network I used.)
3) It probably won't be long before every major e-mail interface shows a prominent icon indicating whether a message is certified or not. Then, banks can say, "Don't reply to any message from us, unless your client shows that it is certified to come from bankofamerica.com." Now that I think of it, the e-mail clients should not just say that the mail is certified -- they should also say from where, and preferably in plain text ("Bank of America Corporation"). There should also be a system to ensure that phishers cannot sign up for similar names to real companies. Maybe this could piggyback on the system for SSL certificates.
I think this is straying into an area where "reasonable people can differ." I haven't seen a mountain of evidence that increasing efficiency (i.e., expanding the economy) is a better way to make the poor better off than redistributing the wealth that already exists. Economists like to say that the only way to make anyone better off without making someone else worse off is to expand the economy. And with that, they neatly wash their hands of the question of fairness.
But by neglecting the fairness question, we virtually guarantee that the poor will not be made better off, even if the economy expands, because they have no power to grab a bigger piece of the growing pie. They're likely to keep going into wage negotiations and settling for just enough to live on. So then, we see entrenched poverty, and rising disparity in incomes and wealth. Maybe, practically speaking, we are more likely to end poverty via an explicitly redistributive system (i.e., progressive taxes), rather than just trying to grow our way out of it and hope for the best. Certainly, an economic argument could be made that the total utility realized by everyone in the economy would be increased by redistributing wealth in a fairer way -- a dollar is worth much more to a minimum wage worker than to Donald Trump, so if we taxed Donald Trump an extra $1 million, and redistributed it to his employees, the total supply of happiness would increase. Which is what economists want, isn't it? (Or to be less simplistic, that $1 million could do an awful lot of good in inner city schools, job training programs, libraries, etc.)
It may be equally likely that alleviating poverty will increase the size of the economy, rather than the other way around. Think how much harder it is for a poor person to work their way up to a job that really uses their talents, compared to a middle class person. Think how many never do, and how much wasted economic potential that represents.
Well, if you're going to throw in elasticities, then it almost doesn't matter where the tax is placed. It doesn't matter whether we tax wages on the employer or employee side, because the final wage will still be at the intersection of the employer's "demand curve" for labor and the worker's "supply curve," with the tax stuck in between. If we tax dividends or capital gains instead of wages, employers will push down the salary they're willing to pay workers, in order to show more profits, in order to attract shareholders despite the tax on profits. If we tax pollution instead of income, employees may be willing to take lower wages without the tax burden, so that's what they'll get. Meanwhile, employers could use those concessions to keep costs down and profits up, despite additional compliance costs.
In the end, maybe it comes down to this: Everyone in the country collectively contributes to a large economic pie. The government takes its share, by whatever simple or arcane tax code. And everyone else is left to fight for their share of the rest. The share you get depends less on the intricacies of the tax code than on your power in society: Do you have valuable skills? Do you believe your labor is worth a lot? Are you willing to hold out for a higher wage, or a higher return on your capital? Can you afford to? Those that have money and skills have power, and will do OK. Those that don't are going to have to work hard or make do with less.
Your cursor keeps moving? If I flick my trackpad, it generally goes a couple hundred pixels then stops. I should have been more precise about this. If I want to move my cursor up to the menu, I make one quick stroke up to the top of the trackpad (I called it a flick, but then as I was writing I realized that my finger never leaves the trackpad). Usually my mouse reaches the top of the screen before my finger goes off the top of the trackpad, in which case, I can switch immediately into a slower stroke to find the specific menu title and run down to the specific menu item. Sometimes my finger goes off the trackpad before I reach the menubar (e.g., if the cursor starts out low on the screen and/or my finger starts out high on the trackpad). But even then, the first stroke gets near the top, and I can usually complete the whole menu selection with the next stroke (up to the menu title and down to the menu item).
I can reliably get about 700 pixels of vertical motion from that first stroke, which seems like plenty for this. (I hadn't noticed before, but I can do the whole thing in one stroke more reliably if I'm editing text in the middle of the screen instead of along the very bottom.) The key seems to be to start low on the trackpad, and move decisively to the top (i.e., quickly, but without removing your finger). This works for side-to-side movement too -- I can generally hit any edge of the screen in one quick stroke. For what it's worth, the "tracking speed" on my MacBook Pro is set slightly below the center of the scale (tickmark number 5 out of 10)
Actually, since you don't really need to aim, you can get to the top with a quick twitch of the finger -- trackpad acceleration takes care of the rest. On my MacBook Pro, I consistently use these finger movements to hit menu items (e.g., for Save As...):
1) One quick flick on the trackpad to get the pointer near or into the menubar 2) Move at medium speed to the menu I want (with a single diagonal upward movement, moving the pointer along the top edge of the screen) 3) Click and hold on the menu title (File) 4) Sweep finger quickly down until it is close to my menu item (usually one stroke) 5) Move the finger more slowly until I am right on the menu item (Save As...) 6) Release the mouse button
Because they are U-shaped, all of these movements can often be completed together without lifting my finger from the trackpad, or with only one lift. All of this takes no more than a second or two, because the only close tracking I have to do is moving horizontally to the File menu after I reach the top of the screen, and fine tuning to hit "Save As..." once I'm nearly there. I imagine this is similar to the effort needed to navigate a context menu, and slightly less than a window-top menu in Windows. Empirically, all of these must be fairly close; otherwise people who have tried more than one approach would have a universally strong preference for one or the other.
Speaking as a frequent Mac user, I can see the concerns people have been getting at in this thread, but I find most of them are manageable. In exchange, I get a system that I don't have to become an OS expert to setup and manage (I don't want to spend my time that way anymore), and one with a richer interface (more 3D and "solid") and better search capability than XP (e.g., instant search for files, or for any e-mail message I've ever received or sent). System and application settings are also much more transparent than in XP (the Library file system is much easier to manage than the registry).
Going into some of the points thrown around here: I'm surprised that setting preferences would be among the most common menu actions for so many people. Sounds like a lot of tinkering. But anyway, for most common actions on the Mac (including preferences), there are standard keyboard shortcuts. This makes the position of the menubar less important. When I want to edit or save a particular document on the Mac, I just use Cmd-Tab to get there, then start typing or use Cmd-S to save. Alternatively, I could click on the relevant window with the mouse, then use one of these keyboard shortcuts. I don't see how these could be any faster in Linux. If I need to use the menu for a particular document, I'm in the habit of Cmd-Tabbing to it (without moving the mouse), then throwing the mouse pointer up to the menu bar. This may be slower than just hitting a menubar within the relevant window, but it isn't too painful. (One thing I would like is an intuitive keyboard-based window-switcher instead of application-switcher; Witch gets close, but takes a while to popup the window list at first, sorts Carbon windows unreliably, and requires System Preferences that are incompatible with Parallels.)
Also, the dock need not take up vertical space on the screen. I keep mine along the left edge. All the windows know it's there, so they usually end up filling the screen vertically without any trouble.
I can see the case for a built-in right-click function, but I mostly notice this when I am working with Windows programs (e.g., pasting at a Command Prompt, using the Send To -> Notepad option in Explorer, or creating a new folder in Explorer). Arguably, the Mac alternatives to these actions are better than right-click anyway (Cmd-V in Terminal, or drag a document to the TextEdit icon in the dock, or Shift-Cmd-N to make a new folder in Finder [I can't stand waiting for Explorer to dig up icons for every possible new document template when I just want a new folder!]). The point being, you need right-click much more if the interface is designed to use it, and Macs work fine most of the time without it (and that may be why Apple doesn't build in a right-click button -- so developers won't force people to use it). I find myself getting a lot of phantom right-clicks if I use the two-fingered tap option, so I disable that and just ctrl-click when I need it; a real right-click built in to the laptop would be easier. Maybe someday Apple will make the trackpad button work like the Mighty Mouse -- power users could set the left and right sides of the (single) button to work differently.
And I agree, something could be done to improve application finding and launching. My Applications folder winds up full of a mixture of individual applications, and folders containing applications and support files. By default, if I want anything other than what I've put in the Dock, I have to open up Finder and dig into this mess. I could create custom categorized folders full of aliases, but that would take some time and upkeep. I have put my Applications and Utilities folders in the Dock, but I have to either ctrl-click or click-and-hold to get them to popup, and they take a few seconds extra if they haven't been used for a while. All these little delays make that approach unappealing. In the end, I now just use Spotlight most of the time to get to applications that aren't in the Dock. Cmd-Spacebar, "disk in", down-arrow, return, will launch "Disk Inventory X", wherever it is.
I just got a MacBook Pro with the 15" matte screen. The resolution is very clear and sharp (hard to say what I mean by that exactly, but think of the opposite of fuzzy pixels on a CRT display). It is bright enough that I keep it set a few notches below top brightness for normal use. That may be because I'm used to using an iBook G4, which may have been dimming as it got older. I haven't thought much about contrast or response time, but I've seen no sign of trouble.
However, the color accuracy does suffer more than I would expect when I move off the perpendicular viewing axis. Moving left or right is not too bad: the white balance and contrast seem to change a little, but the picture remains basically correct and viewable out to at least about 60 degrees off-axis (i.e., a 120 degree viewing angle). When I move off-axis vertically, it can be worse. Looking from a few degrees below, the contrast is exaggerated, but not too badly. However, if I tilt the top of the screen towards me by even a couple of inches, the image can become reversed in light areas, and highlights in photos become completely washed out. It's rare to move my head up or down enough to notice this (e.g., if I'm sitting in a chair and the computer is on a desk). But if I have the computer propped on my knees while lying down, I may end up fiddling with the screen angle more than I'd expect when I reposition the computer, to get the image right again. This sort of thing is probably what the TH editor was referring to.
Slashdot Mirror
The new Mac Mini is actually rated for "14 watts of power when idle", which Apple says is lower than any other computer in the EPA STAR database.
But wait, there's more! With Snow Leopard, Apple introduced a new "Wake on Demand" feature that could allow the Mini to be in sleep mode (~1.5W) most of the time, but still work as a server when needed:
(1) Airport base stations can now provide a "Bonjour Sleep Proxy" service that will keep announcing all the services your computer is hosting on the network, even after the computer goes to sleep. The base station will then wake the server whenever another computer tries to access it. It isn't clear whether this happens only for Bonjour services, or for any IP-based access (which should be possible in principle).
(2) All this magic can happen via the wireless network if your computer is new enough. (Wake-on-LAN was only possible via ethernet before.)
If the server is set to go back to sleep automatically after a few minutes of inactivity, this setup provides a low-power, always-on server arrangement: the computer sleeps most of the time, automatically wakes up whenever someone wants to connect to it, then goes back to sleep whenever it's not needed.
Bladeless designs may be becoming more popular for small, building-scale turbines, but they are not even on the horizon for large, commercial-scale turbines -- the sort that we could use for a large share of our electricity. As far as I know, every single utility-scale wind turbine on the market today uses a horizontal-axis design with 3 (or occasionally 2) blades. Their speed is lower in RPMs, but the tips are still moving around 150 mph.
And birds and bats do indeed fly into these things or get injured by the pressure changes they create. It may not be any more than would fly into cell-phone towers, power lines or other infrastructure, but as a "green" technology, wind power is held to a higher standard. (My personal theory is that high-flying birds and bats are looking for prey in an environment where there never used to be any risk of flying into anything, so they have never developed the ability to avoid objects ahead of them, especially moving objects). At any rate, there is definitely a need for creative ways to keep birds and bats away from wind turbines.
I just downloaded the authors' data and compared the time of the first mention of each meme in blogs versus the mainstream media. Surprisingly, the media make the first mention only 51.3% of the time. The rest of the time, the blogs had the first mention.
The blogs were ahead by 170 hours or more 25% percent of the time, trailed by 4 minutes or less 50% of the time, and trailed by 8 hours or less 75% of the time.
These statistics suggest that these blogs break a story before these mainstream media outlets about half the time. Sometimes the blogs are way ahead, but they are rarely far behind. The blogs have to get their news from somewhere, so it is surprising that they are ahead of (these particular) media outlets so often. Maybe they are pulling news from less-mainstream outlets (not included in this study) and building "buzz" around it, which then gets picked up by the mainstream news outlets?
The point is that a lot of people are claiming the MSM is obsolete and blogs are the way of the future ... and this study pretty clearly shows that it isn't true.
I thought that would be the point of the story when I read it, but the story doesn't actually mention this issue at all. The researchers mostly seem to be interested in understanding how stories become popular, and the roles that blogs and traditional media play in that process.
In the original paper (e.g., Figure 8), they report that there is a 2.5 hour lag between the peak of reporting on a story in the media in general and the peak of discussion in the blogs in general.
They also report the typical time lag for individual news outlets or blogs (Table 1), and show that a few individual blogs (e.g., hotair.com and talkingpointsmemo.com) have tend to report stories before individual media outlets. However, even this doesn't show that news appears in blogs before it appears in the media -- some individual blogs tend to report big stories before individual news outlets, but that may be because (a) they pull stories from many news outlets, so they will inevitably have an earlier average reporting time than any individual news outlet, and (b) the early-mover blogs play a role in determining which stories become popular, even if they aren't the first to report them.
Unfortunately, I didn't see any graph that tracked the earliest appearance of a story in any media outlet, and the earliest appearance of the same story in any blog, and compared the times of those appearances. That would be the way to really answer the question of who is reporting first. And I bet it's the media, by many hours.
There's not as much money in newspaper advertising as there used to be, and this will inevitably lead to a reduction in the amount of news being collected and the number of printed newspapers.
In the old days, your local newspaper(s) had a monopoly or oligopoly on display and classified advertising. This gave them enough money to hire a local reporting staff, and in some cases, to set up remote bureaus. The smaller papers relied on wire services or news agencies for their national or international news, and the bigger papers gathered some of this for themselves. Often there was enough advertising revenue to support two or more newspapers in the same town.
Now, many readers have switched from printed newspapers to the Internet, drying up the display ad revenue (newspapers make money selling readers to advertisers, not selling news to readers). The websites don't get nearly so much revenue per ad-view as the printed papers did. Meanwhile craigslist has grabbed the classified ads. So now each person who reads a story doesn't bring in nearly as much ad revenue as they used to. What's going to happen?
I think a big consolidation is inevitable -- the amount of original news reporting will have to be reduced, so that more people read each story, and the ad revenue per story returns to a high enough level to support the cost of writing it. All the newspapers will lose money for a while, until most of them have failed or radically restructured (e.g., going online-only and closing any remote bureaus). At that point, all national and international news will probably be gathered by a few national wire services, a few national TV networks, and maybe a couple of major national papers. All the other papers, websites and TV stations will rely on these sources for their "content". There will probably also be a big reduction in local reporting, except in the biggest cities. But each original story will be so widely disseminated that the revenue from teeth-whitening ads on Yahoo.com, Applebee's ads on sfgate.com, and Macy's ads in a few dozen local papers will be enough to cover the cost of reporting it.
This is depressing if you care about having a diversity of news sources, but it is probably unavoidable. There just isn't enough ad revenue to support as much news reporting as we have now.
Actually, only one digit is used for the checksum on most credit cards.
For a 16-digit number, there is a 6-digit issuer identification number (including 1-digit major-industry identifier), followed by 9 unique digits for the customer, followed by one check-digit. Some Visa numbers used to be 13 digits, which would have been much less unique, but those seem to have been converted to 16-digit numbers now (all this from Wikipedia).
My Motorola krzr k1m (verizon) used to have great battery life, except when running GPS with turn-by-turn navigation (i.e., continuous update of my GPS position). Then a new, fully-charged battery would run down within an hour, and the phone would get noticeably warm. Eventually I started using the navigator function to plan out the next few steps of my trip, then turned it off for a while, then repeated once I got closer to my destination.
Even with such efforts, GPS seemed to kill batteries quickly -- i.e., the battery physically expanded, and its recharge capacity dropped precipitously after a few months. Lithium-ion batteries don't like to run hot, so maybe that was a factor.
I was hoping that the 3G S had avoided these problems, but maybe they haven't...
I'm in the same boat. For years I've been looking for a file system can hold files larger than 2GB and can be mounted from Windows and Mac OS X (and maybe Linux). That would allow me to store all my work on one partition, and access it from Mac OS X, from Windows via Boot Camp, or from Windows inside Parallels or VMware Fusion. It would also allow me to transfer large files back and forth between my Mac and other Windows computers. I was hoping ZFS would be that file system.
The last time I checked (the middle of 2008), the only way to do this was via NTFS, and the only read-write support for NTFS on OSX was the MacFUSE NTFS driver, which was pretty slow.
I just saw that MacDrive 7.2 now allows Windows Vista x64 (my Boot Camp OS) to read HFS disks, so maybe I'll give that a try. There are also rumors that Snow Leopard's Boot Camp utility will include drivers for Windows to read HFS disks, so maybe that will help too.
You ignore the waste caused storing the energy during windy periods for not windy periods.
I assume that all the wind or nuclear power is used when it is generated (which is a realistic assumption for the foreseeable future). Storage (if it were needed for either of these) wouldn't radically change the land use, but might change the costs.
that is a 14,000acre area of land you are giving up
The wind farms might actually be placed on a larger area of land than that. But only a tiny fraction of this land is actually occupied by the towers and roads. The rest is available for farming, ranching, etc. Have you estimated the land use for uranium mining, tailings piles, waste storage, etc.?
And windfarms are noisy and ugly (so are nuclear plants but they aren't 80m tall.
New wind turbines are inaudible beyond a few dozen meters away. People may differ over whether wind farms are more ugly than uranium mines and nuclear power plants.
And if you want to spread those towers out more so you dont lose a 60km^2 chunk you will end losing more again because of efficiency loss.
Wind farms are highly modular -- yes it helps if they are near existing power lines, but it is also possible to build new power lines. And they certainly don't decrease in efficiency if they are built in multiple locations.
The real issue here is not land use anyway. It's what sort of vision we have for the future of the power system. Big nuclear plants, which require a lot of faith in cost forecasts and our ability to control waste (at the mine, mill, plant and afterward). Or wind and solar, which also require some faith in cost forecasts, may use more land area (but in a different way), and don't have the other risks. Intermittent sources require more ingenuity to integrate into the power system, but this is not insurmountable (people on slashdot should recognize that, if anyone). And don't overlook the fact that nuclear plants are as inflexible as wind or solar power -- they have to run around the clock, so if we want to use them for a large share of our power, we will have to find a way to use their night-time output (if you throw it away, that will kill the economics of the plant).
OK, looks like I got that wrong, at least for this press release. But those costs do not seem to be in line with what U.S. nuclear plant developers expect to pay.
For example, this Wall Street Journal article (follow first link for full text) indicates that FPL Groups expects new AP1000's at its Turkey Point plant in Florida to cost $6-$9 billion each, and Georgia Power Co. expects a 45.7% stake in two similar reactors (i.e., 90% of one reactor) to cost $6.4 billion. (The first two reactors at Georgia Power's plant cost nearly $4.5 billion each 20 years ago, over 10 times the $330 million per plant originally estimated.)
So I'd be pretty skeptical of anyone who claims they can deliver nuclear power for less than $0.06-$0.10 per kWh. On the other hand, lots of wind power is being delivered at those costs or lower.
Texas passed a law in 1999 requiring 2 GW of wind power by 2009, but they keep exceeding their goal. By 2007 they had around 2.7 GW of wind capacity. Then they added another 1.6 GW in 2007, and another 2.7 GW in 2008, bringing the total to around 7 GW. This is driven hardly at all by the Texas requirements, a little bit by the federal tax incentives (around $0.01/kWh), and mostly by the fact that Texas has great wind sites and wind power is now competitive with other sources of electricity.
The industry has been growing by 30-40% per year worldwide, and that kind of growth has a way of sneaking up on you.
At a cost of $5.85 billion, and assuming a lifetime of 40 years, an interest rate of 6%, this nuclear plant will have an annual mortgage of $389 million. With a nameplate rating of 1100 MW, if it runs 92% of the time, it will produce 8.9 billion kWh per year, so the capital repayments will amount to $0.044/kWh, assuming it doesn't go over budget. Assuming an optimistic cost for fuel around $0.005/kwh, this gives a total cost of $0.049/kWh, neglecting the cost of maintenance, waste disposal, and any risk of contamination or weapons proliferation.
Now let's look at a new wind farm. A 50 MW wind farm would cost around $96 million (at $1923/kW), which yields an annual capital repayment of $7.5 million (assuming a lifetime of 25 years). If the plant runs at a 35% capacity factor, it will produce 153 million kWh per year. So the total cost will be $0.049/kWh.
So, which would you rather spend $0.049/kWh on -- a nuclear plant that might go over budget, might leak radiation at some point during its life, whose waste will need to be carefully controlled and permanently stored somewhere that hasn't yet been identified; or a wind farm whose costs are much more certain and which comes without all those ancillary risks?
Yes, any individual wind farm will not provide a firm supply of power. But if a lot of wind farms are used, and they are combined with solar, geothermal and other renewable resources, they will provide a fairly stable power supply. There is also a lot of potential for reshaping electricity loads to match the supply of power (e.g., recharge electric vehicles when the wind is blowing or the sun is shining). And finally, if you must have a firm supply of power, you can convert a wind farm into a completely firm supply (at 35% of its nameplate rating) by spending about 10% extra and building rarely-used natural gas peaker plants ($634/kW * 35% = $222kW).
I don't normally go in for grammar corrections, but I can't resist here:
1) "pseudo"
2) you are -> you're
3) owned by it -> its (it is -> it's)
4) The quote you attribute to Thomas Jefferson was actually made by Gerald Ford. Nobody imagined "a government big enough to give you everything you want" in Jefferson's day.
The point is not what happens while tinurl.com is around, but what happens when it is not? Or when they decide to start charging for their services? "Free link redicrector forever" is not a sustainable business model.
I'd highly recommend Tesla: Man out of Time, by Margaret Cheney. It is very approachable and engaging, and will give them the following:
a) appreciation of an underappreciated scientific genius
b) understanding and awe of the power of resonance
I completely agree with you. I read GEB when I was 15 or 16, and it has changed how I see everything since then. It's an amazing introduction to the idea of self-referential systems, and once you get it, you see self-reference everywhere (makes a great introduction to chaos theory or self-modifying code as well).
I could see how some people would dislike the dialogues, but I loved them, especially the way the structure of the stories reflected the concepts the characters were discussing.
GEB also taught me to appreciate Escher and Bach, so maybe you'll get some thanks from the art and music departments too!
I'm kind of pissed that "a PC" means Windows ... it means personal computer, does it not? Isn't my Linux machine a personal computer?
I think the "PC" thing dates back to the 1980s, when the IBM PC first came out. Other manufacturers figured out how to clone the IBM PC's BIOS (against IBM's will), and then started making "IBM PC-compatible" machines that could run the same software (i.e., MS-DOS and everything that ran under it). After a while "IBM PC-compatible" computers became "PC-compatible" and then just "PCs," and "PC" also took on the sense that the computer was running MS-DOS, or later, MS Windows. I guess the meaning of "PC" has now narrowed to usually mean only computers running MS Windows, but sometimes it still refers to the hardware, e.g., "Macintosh OS X runs on standard PC hardware". Macs, Linux and Windows machines are all PCs now in either the first sense ("personal computer", generically) or the second sense (able to run IBM PC binaries and their successors), but not the third sense (running Windows at any given moment).
I'm kind of pissed that "a PC" means Windows ... it means personal computer, does it not? Isn't my Linux machine a personal computer?
I think the "PC" thing dates back to the 1980s, when the IBM PC first came out. Other manufacturers figured out how to clone the IBM PC's BIOS, and then started making "IBM PC-compatible" machines that could run the same software (against IBM's will). After a while "IBM PC-compatible" computers became "PC-compatible" and then just "PCs". Then when the field dwindled to Macs and "PCs", I guess "PC" came to mean Windows machines. But in either the first sense ("personal computer", generically) or the second sense (able to run IBM PC binaries and their successors) newer Macs, Linux and Windows machines are all PCs now.
Actually, this technique needs less corporate cooperation than you suggest.
1) The specification allows domain owners to report their signing policy in their domain DNS record. So Paypal can announce publicly that ALL mail from paypal.com must be signed. i.e., the senders can add themselves to the whitelist. If your e-mail client gets mail claiming to be from paypal.com, but without the proper certification, it knows with 100% certainty that it is junk.
2) It wouldn't take much for the big retail e-mail providers to start certifying all mail coming from their domains. Of course, they could only do this if they required their users to send outgoing e-mail using a secure connection to their own SMTP server. This would be annoying for people who like to send yahoo.com e-mail from their home SMTP server, but in the long run it's a better way to do it. This would move toward better symmetry between POP/IMAP and SMTP protocols (i.e., you use a secured logon to your e-mail provider for both, which is probably how it always should have been. It used to drive me crazy having to change my SMTP server for each network I used.)
3) It probably won't be long before every major e-mail interface shows a prominent icon indicating whether a message is certified or not. Then, banks can say, "Don't reply to any message from us, unless your client shows that it is certified to come from bankofamerica.com." Now that I think of it, the e-mail clients should not just say that the mail is certified -- they should also say from where, and preferably in plain text ("Bank of America Corporation"). There should also be a system to ensure that phishers cannot sign up for similar names to real companies. Maybe this could piggyback on the system for SSL certificates.
I think this is straying into an area where "reasonable people can differ." I haven't seen a mountain of evidence that increasing efficiency (i.e., expanding the economy) is a better way to make the poor better off than redistributing the wealth that already exists. Economists like to say that the only way to make anyone better off without making someone else worse off is to expand the economy. And with that, they neatly wash their hands of the question of fairness. But by neglecting the fairness question, we virtually guarantee that the poor will not be made better off, even if the economy expands, because they have no power to grab a bigger piece of the growing pie. They're likely to keep going into wage negotiations and settling for just enough to live on. So then, we see entrenched poverty, and rising disparity in incomes and wealth. Maybe, practically speaking, we are more likely to end poverty via an explicitly redistributive system (i.e., progressive taxes), rather than just trying to grow our way out of it and hope for the best. Certainly, an economic argument could be made that the total utility realized by everyone in the economy would be increased by redistributing wealth in a fairer way -- a dollar is worth much more to a minimum wage worker than to Donald Trump, so if we taxed Donald Trump an extra $1 million, and redistributed it to his employees, the total supply of happiness would increase. Which is what economists want, isn't it? (Or to be less simplistic, that $1 million could do an awful lot of good in inner city schools, job training programs, libraries, etc.) It may be equally likely that alleviating poverty will increase the size of the economy, rather than the other way around. Think how much harder it is for a poor person to work their way up to a job that really uses their talents, compared to a middle class person. Think how many never do, and how much wasted economic potential that represents.
Well, if you're going to throw in elasticities, then it almost doesn't matter where the tax is placed. It doesn't matter whether we tax wages on the employer or employee side, because the final wage will still be at the intersection of the employer's "demand curve" for labor and the worker's "supply curve," with the tax stuck in between. If we tax dividends or capital gains instead of wages, employers will push down the salary they're willing to pay workers, in order to show more profits, in order to attract shareholders despite the tax on profits. If we tax pollution instead of income, employees may be willing to take lower wages without the tax burden, so that's what they'll get. Meanwhile, employers could use those concessions to keep costs down and profits up, despite additional compliance costs.
In the end, maybe it comes down to this: Everyone in the country collectively contributes to a large economic pie. The government takes its share, by whatever simple or arcane tax code. And everyone else is left to fight for their share of the rest. The share you get depends less on the intricacies of the tax code than on your power in society: Do you have valuable skills? Do you believe your labor is worth a lot? Are you willing to hold out for a higher wage, or a higher return on your capital? Can you afford to? Those that have money and skills have power, and will do OK. Those that don't are going to have to work hard or make do with less.
Actually, this can be done with a little fiddling. You can add an applescript to the toolbar:
trytell application "Finder" to update items of front window
end try
This works well for local files, but may not update views of samba shares.
Or you can use a context menu extension called nudge, which may work better for network shares.
There's also an interesting discussion of the issue here.I can reliably get about 700 pixels of vertical motion from that first stroke, which seems like plenty for this. (I hadn't noticed before, but I can do the whole thing in one stroke more reliably if I'm editing text in the middle of the screen instead of along the very bottom.) The key seems to be to start low on the trackpad, and move decisively to the top (i.e., quickly, but without removing your finger). This works for side-to-side movement too -- I can generally hit any edge of the screen in one quick stroke. For what it's worth, the "tracking speed" on my MacBook Pro is set slightly below the center of the scale (tickmark number 5 out of 10)
Actually, since you don't really need to aim, you can get to the top with a quick twitch of the finger -- trackpad acceleration takes care of the rest. On my MacBook Pro, I consistently use these finger movements to hit menu items (e.g., for Save As...):
1) One quick flick on the trackpad to get the pointer near or into the menubar
2) Move at medium speed to the menu I want (with a single diagonal upward movement, moving the pointer along the top edge of the screen)
3) Click and hold on the menu title (File)
4) Sweep finger quickly down until it is close to my menu item (usually one stroke)
5) Move the finger more slowly until I am right on the menu item (Save As...)
6) Release the mouse button
Because they are U-shaped, all of these movements can often be completed together without lifting my finger from the trackpad, or with only one lift. All of this takes no more than a second or two, because the only close tracking I have to do is moving horizontally to the File menu after I reach the top of the screen, and fine tuning to hit "Save As..." once I'm nearly there. I imagine this is similar to the effort needed to navigate a context menu, and slightly less than a window-top menu in Windows. Empirically, all of these must be fairly close; otherwise people who have tried more than one approach would have a universally strong preference for one or the other.
Speaking as a frequent Mac user, I can see the concerns people have been getting at in this thread, but I find most of them are manageable. In exchange, I get a system that I don't have to become an OS expert to setup and manage (I don't want to spend my time that way anymore), and one with a richer interface (more 3D and "solid") and better search capability than XP (e.g., instant search for files, or for any e-mail message I've ever received or sent). System and application settings are also much more transparent than in XP (the Library file system is much easier to manage than the registry).
Going into some of the points thrown around here: I'm surprised that setting preferences would be among the most common menu actions for so many people. Sounds like a lot of tinkering. But anyway, for most common actions on the Mac (including preferences), there are standard keyboard shortcuts. This makes the position of the menubar less important. When I want to edit or save a particular document on the Mac, I just use Cmd-Tab to get there, then start typing or use Cmd-S to save. Alternatively, I could click on the relevant window with the mouse, then use one of these keyboard shortcuts. I don't see how these could be any faster in Linux. If I need to use the menu for a particular document, I'm in the habit of Cmd-Tabbing to it (without moving the mouse), then throwing the mouse pointer up to the menu bar. This may be slower than just hitting a menubar within the relevant window, but it isn't too painful. (One thing I would like is an intuitive keyboard-based window-switcher instead of application-switcher; Witch gets close, but takes a while to popup the window list at first, sorts Carbon windows unreliably, and requires System Preferences that are incompatible with Parallels.)
Also, the dock need not take up vertical space on the screen. I keep mine along the left edge. All the windows know it's there, so they usually end up filling the screen vertically without any trouble.
I can see the case for a built-in right-click function, but I mostly notice this when I am working with Windows programs (e.g., pasting at a Command Prompt, using the Send To -> Notepad option in Explorer, or creating a new folder in Explorer). Arguably, the Mac alternatives to these actions are better than right-click anyway (Cmd-V in Terminal, or drag a document to the TextEdit icon in the dock, or Shift-Cmd-N to make a new folder in Finder [I can't stand waiting for Explorer to dig up icons for every possible new document template when I just want a new folder!]). The point being, you need right-click much more if the interface is designed to use it, and Macs work fine most of the time without it (and that may be why Apple doesn't build in a right-click button -- so developers won't force people to use it). I find myself getting a lot of phantom right-clicks if I use the two-fingered tap option, so I disable that and just ctrl-click when I need it; a real right-click built in to the laptop would be easier. Maybe someday Apple will make the trackpad button work like the Mighty Mouse -- power users could set the left and right sides of the (single) button to work differently.
And I agree, something could be done to improve application finding and launching. My Applications folder winds up full of a mixture of individual applications, and folders containing applications and support files. By default, if I want anything other than what I've put in the Dock, I have to open up Finder and dig into this mess. I could create custom categorized folders full of aliases, but that would take some time and upkeep. I have put my Applications and Utilities folders in the Dock, but I have to either ctrl-click or click-and-hold to get them to popup, and they take a few seconds extra if they haven't been used for a while. All these little delays make that approach unappealing. In the end, I now just use Spotlight most of the time to get to applications that aren't in the Dock. Cmd-Spacebar, "disk in", down-arrow, return, will launch "Disk Inventory X", wherever it is.
I just got a MacBook Pro with the 15" matte screen. The resolution is very clear and sharp (hard to say what I mean by that exactly, but think of the opposite of fuzzy pixels on a CRT display). It is bright enough that I keep it set a few notches below top brightness for normal use. That may be because I'm used to using an iBook G4, which may have been dimming as it got older. I haven't thought much about contrast or response time, but I've seen no sign of trouble.
However, the color accuracy does suffer more than I would expect when I move off the perpendicular viewing axis. Moving left or right is not too bad: the white balance and contrast seem to change a little, but the picture remains basically correct and viewable out to at least about 60 degrees off-axis (i.e., a 120 degree viewing angle). When I move off-axis vertically, it can be worse. Looking from a few degrees below, the contrast is exaggerated, but not too badly. However, if I tilt the top of the screen towards me by even a couple of inches, the image can become reversed in light areas, and highlights in photos become completely washed out. It's rare to move my head up or down enough to notice this (e.g., if I'm sitting in a chair and the computer is on a desk). But if I have the computer propped on my knees while lying down, I may end up fiddling with the screen angle more than I'd expect when I reposition the computer, to get the image right again. This sort of thing is probably what the TH editor was referring to.