Having actually used a SiCortex machine, I can tell you that the problem wasn't the VC, or the compilers, or even really the hardware.
The problem was the market.
There are two types of x86-based small clusters (the market that SiCortex was aiming for): clusters with Gigabit Ethernet and clusters with expensive interconnects (Mirinet, InfiniBand, or 10G Ethernet).
Gigabit Ethernet clusters do a good job with problems that are embarrassingly parallel (or at least have minimal communication demands). $150k gets you 300 Nehalem cores and a lot of memory. SiCortex fails here because their competition (the SC1458) is much more expensive and much slower. The fact that the SC1458 uses less power (around 5kW instead of 10kW) is impressive, but unless you're very power or cooling constrained, it's simply more cost effective to deal with the extra power and cooling cost.
SiCortex hardware was more cost effective against clusters with expensive interconnects. The problem is, the people who buy clusters with expensive interconnects do so because their problem is interconnect heavy. Unfortunately, despite all of the cool CS behind SiCortex's interconnect, the fact is that it just didn't do that well against InfiniBand. That's partly because the SiCortex system has more nodes, which means that more messages have to use the interconnect. It's partly because for very small clusters, it's possible to use a single IB switch that connects every node to every other node. And it's partly because SiCortex didn't have the kind of mature hardware/software stack that someone like Mellanox has.
So, there you have it. For the problems that ran well on SiCortex hardware, you could get the same performance at dramatically lower cost using Gigabit Ethernet. For the problems that require an expensive interconnect, the SiCortex approach of "more, smaller nodes" results in dramatically more overhead compared with the "fewer, faster nodes" strategy.
If that's true for N, I smell a lawsuit. Neutral had always meant "physically disconnect engine from wheels". Are you absolutely sure you're correct? I wouldn't ever drive a car in which this isn't true.
This is impossible in a Toyota hybrid, because the gears are never 'shifted' and always remain in the same mechanical configuration.
HOWEVER, it takes active computer control to actually deliver power. With both motors disconnected, the gears are free to spin and the engine cannot deliver power even if it were running.
I have owned lots of latops, have 2 brand new ones from work, a hp and a dell. Not the cheapest of the lot, not the most expensive. I would put my mac up against those machines, any day and do. They often dont' sleep properly, and they take ONE minute to sleep when they do, whereas my mac sleeps in less than a second, wakes up in a second to.
I own a $400 Acer 1410 laptop and two self-built desktops, one AMD and one Intel. I also recently sold a ThinkPad T61 and an Acer EEE PC 900HA.
ALL of them sleep and wake up in less than 3 seconds. My Acer wakes up in about a second (it has an Intel SSD that I installed); the others take a couple of seconds to spin up the hard drive.
The "Macs sleep and wake and PCs don't" argument is total bullshit. Find me one machine that's sold in Best Buy that doesn't sleep and wake properly.
I have owned a TON of hardware over the last 5 years, including the following: - Acer EEE PC 900HA - HP Compaq 6910p - ThinkPad T61 - MSI Wind Nettop 100 - Custom-built AMD dekstop (GeForce 6100 chipset, Gigabyte motherboard) - Custom-built Intel desktop (Intel G31 chipset, Gigabyte motherboard) - Acer Aspire 1410 - Toshiba Portege M200 tablet PC - Compal EFL30 generic notebook
All of this hardware properly sleeps and wakes properly.
My Acer 1410 was $400. It's 3lbs, has an 11.6" 1366x768 display, 4GB of memory, runs for 6-8 hours on a charge, has an HDMI port (including audio support), and a full size keyboard.
My Acer is made of plastic. It doesn't have a unibody case. It doesn't have a glowing Apple logo. It doesn't run Mac OS X.
But, you know what? I would take it over any Mac at any price (assuming that I couldn't sell the Mac). I have a desktop when I need computing power. My laptop needs to be light, small, have a usable keyboard and screen, and get good battery life.
On one hand, I think that understanding the hardware is still an important part of producing code that runs well. On modern high-end CPUs, it's more about things like memory accesses than it is about optimizing out every last instruction, but it's still critical.
On the other hand, it's unreasonable to expect that someone with an undergraduate degree would understand how a modern out-of-order CPU works. At best, we might hope that students get some understanding of data dependencies and simple out-of-order concepts like Tomasulo's algorithm and register renaming. Of course, every student should understand how caches work, why having unpredictable branches is bad, and just how badly you're screwed when you start hitting main memory frequently.
The thing is, compilers already do a pretty good job with these kinds of optimizations. You can do brain-dead things like access multidimensional arrays in the wrong order, and the compiler will often (but not always) save your butt. You can put function calls all over the place, and many of them will get inlined. You can use constant variables instead of hardcoding values, and it won't matter.
So, yes, if you're running on a system that's esoteric, or doing something that's unusual, you can sometimes dramatically improve performance by understanding the machine code and the hardware. But when you're running on a platform that's well known (x86 or ARM, for example), using a good compiler (GCC counts), you can be surprisingly dumb and still get good performance.
If everyone were a systems expert, we would probably all be using VLIW systems right now.
ShipIt was great. It's not that I ever used the CDs myself, but they were great to give away at LUG meetings, installfests, and other events. There's something to be said for a pressed CD in nice packaging (compared with a burned CD).
That said, when you've got 30Mbps at home, CDs in the mail seem a bit silly.
We're the only developed country that has this problem
That's just false. There are significant anti-vaccine campaigns at least in the UK and Australia, and likely many other places that I don't know about.
This is a serious global problem. As you point out, not getting a vaccine puts everyone else at increased risk. If the anti-vaccine crowd gets their way, we could very well see resurgences in diseases such as Whooping Cough.
The reason it happens is that people are bad at understanding risk. The risk of dying to Influenza in a typical Flu season? About 1 in 10,000 (less if you're a healthy adult, more if you're in a high-risk group). That's about the same as your risk of being killed in a car accident.
It actually doesn't require a second drive... you can have it back to up another partition on the same drive. It will warn you that you're about to do something stupid, but it will let you do it if you really want to.
I trust that it's clear why backing up your data to another partition on the same drive is generally a dumb thing to do.
No, it's not. The reality is that while hard drive failures get most of the attention, disks are actually pretty reliable and rarely fail catastrophically. In my case, that means never - and I've owned and used dozens of drives from multiple manufacturers.
In contrast, human mistakes are far more common. Deleting a file accidentally, overwriting it with a bogus version, or just wanting to revert to an earlier revision all happen more often than the relatively rare disk failures.
That's not to say that a disk can't fail, that your laptop couldn't be lost or stolen, or that external backups aren't superior. They are.
However, Windows 7 (and Vista Business/Enterprise/Ultimate) have a feature that's ultimately more useful than Time Capsule: previous versions. Previous versions doesn't protect you from hardware failures, but it's transparent, backs up to the same disk, and most importantly is on by default. The fact that Time Capsule requires a separate partition (or an external drive) to activate means that it's not going to be used by a substantial fraction of Mac users. Previous versions, in contrast, protects everyone who hasn't explicitly disabled it.
The only problem with the Mac Mini is that it's pricey. You can get a similarly-small ION-based dual-Atom box for $330 (including 2GB of memory, disk, and DVD drive), just over half the price of the Mac Mini.
Or, if you're even cheaper, you can build a full-size Pentium Dual-Core box for around $250, which has the added advantage of multiple SATA ports and plenty of room for extra disks (presumably you want your media box to be able to store media). Of course, it won't fit nicely under your TV.
What would stop someone carrying out a man in the middle attack on a web browser or distribution download that provided a different Firefox that contains different CA keys.
On Windows, the Firefox installer is signed using Authenticode using a certificate signed by a "trusted" root CA (e.g. Verisign). When you see "Mozilla Corporation" come up as the "Verified Publisher" in the UAC dialog (in Vista or 7), that string is not simply a field in the executable, it's part of the certificate that Mozilla uses to sign Firefox installers.
Now, as to whether or not you should trust the root CAs, that's a different question. But if you can't trust the root CAs, then it doesn't really matter whether your Firefox build has a modified root CA list.
On Linux, your Firefox packages will be signed with your distro's private key(s).
So 57 years ago France was already broadcasting 441 lines. I was under the impression, that in the USA, today, that 480 lines were being broadcast and sold as the low end of HD.
You would be wrong. NTSC (US analog) standard definition is 525 vertical lines, of which 486 are visible (plus or minus several depending on overscan). This is usually referred to as "480i60", as in, 480 vertical lines, interlaced, 60 fields per second.
The ATSC over-the-air television standard used in the US can carry a variety of formats. It can carry "standard definition" 480i60 (mostly used for lower cost stations, older programming, digital subchannels, and the like). But ATSC can also carry 480p60 "enhanced definition" signals, though this is seldom used.
HD, in the context of television in the US, universally refers to one of two ATSC formats: 720p60 or 1080i60.
480i60 has never been considered "HD" by anyone who has a clue. 480p60 has been marketed as "enhanced definition", but it's frequently just considered SD today.
Personally, I prefer 720p60 for content originally authored for TV (LCD TVs are progressive, so 1080i60 requires deinterlacing which creates motion artifacts). For content that comes from a film source, 1080i60 performs identically to 1080p24, assuming that the telecine is properly done and the deinterlacer in the TV (or set-top-box) can detect film content properly.
Technically, ATSC supports 1080p24 (also 1080p23.976), along with a bunch of other formats. But these formats aren't used in practice, and it's unclear whether hardware even supports them.
The trick to making strong passwords is to not use them at all.
Random passwords don't work. People don't remember them, or they write them down, or they use the same one everywhere. Any of these options compromises the security of a 'bulletproof' random password.
SSH private keys can't be guessed, they aren't compromised if you use them on more than one system (even untrusted systems), and you can revoke them if the machine they are on is compromised.
Better yet, smart cards are even harder to clone, especially if you don't have physical access to the card.
I live in Japan, and it only cost $60-80 USD a month to have a 100MB up & down fiber optic connection in every room of my house.
Cablevision offers 100Mbps for $99/mo in the US. Comcast offers 50Mbps for $99/mo as well.
Comcast has more customers than there are people in Korea. They will achieve 80% DOCSIS 3 coverage by the end of 2009. Delivering "100Mbps" is as simple as updating a configuration file. The problem is that the contention ratio would be horrible.
Guess what, though? The contention ratio is horrible with ANY 100Mbps service. Fiber doesn't change that because you still need to backhaul the data through something. 5000 100Mbps subscribers with a contention ratio of 20:1 is 25Gbps, and I can bet you that they aren't using an OC-768 / STM-256x for every 5000 subscribers.
My university offers "1Gbps" broadband (the LAN is 1Gbps and 10Gbps, and the Internet connection is 2Gbps). But if more than 10 people try to use 1Gbps, the network can't deliver.
They don't spend nearly as much on games as other parts of the world.. and they don't have the greatest broadband. Other than the fact that you're white, is there a reason why you're not rolling out in Korea first?
Because once you look at the contention ratio on Korean superbroadband you realize how utterly screwed you are if you attempt to push 50Mbps to a large number of subscribers.
Here's a hint: A 10Gbps connection (OC-192 or 10G Ethernet) can only feed 200 subscribers if each of them is using 50Mbps. The last mile may be faster in Korea, but the backbone isn't substantially different.
Mosey on down to yer local home depot, or other fine retailer of electrician supplies, and ask for a piece of electrical cable capable of passing 200 amps or so, depending on the clerk's competence and local electrical codes, they'll probably suggest 2/0 gauge copper per NEC standards, which vaguely resembles a copper wire rope the diameter of yer thumb
You don't connect the phone to the wall using the battery voltage. 330W at 48V is a very reasonable 6.9A. You step down the voltage to the correct charge voltage using a DC-DC converter. Models on PC motherboards already handle 150A or more.
Now, how you keep the battery from catching fire while you do that is a completely different question.
No, most of the rest are normal students who happen to play baseball or soccer after school.
At my high school, enrollment in AP courses correlated highly with participation in athletics. The football team was pretty much filled with top students.
It's all in how you set up your program. Academic requirements to participate in athletics go a long way.
The Tesla Roadster is assembled in California using parts from the US and other countries. The frame comes from Lotus in the UK, but it is most definitely not a 'British' vehicle.
Tesla is the only company producing street-legal full-size BEVs today (not NEVs). They are developing a sedan as well. They target the luxury market because it's the only way to get people to pay the $50k+ that's necessary to build a useful BEV today.
Well, speaking as someone who has actually watched usability testing at Microsoft and as someone who has actually seen significant changes made to a product because of it, I can tell you that you're full of crap.
Bitching that the clipboard preserves formatting is a little silly. It's normal, expected behavior, and you can use the 'smart tag' (since Office XP) if you don't want to preserve formatting. The fact that you don't like the default behavior doesn't make it wrong.
I rarely notice auto-bulleting in Word because I don't start my lines with dashes. If I wanted to type plain text, I would use a text editor.
If your point in all of this is that MS products have usability flaws, well, I'd agree. But then so do Apple's. I learned that the first time I wiped my iPod by clicking on the wrong button when I connected it to another PC.
The usability studies do matter, and they do improve the product. Perhaps not as much as we might want, but to say that they are a show is simply silly.
Zune is basically what Apple is pushing (limited storage space, lots of superfluous extras), without the benefit of iTMS and iTunes (which is sad), but hardware-wise about the same
The Zune desktop software in my opinion is vastly superior to iTunes. It's prettier, has a bunch of 'social' features (which sound stupid, but it's actually cool to see what your friends are listening to), and, in general, faster. I've heard that iTunes isn't such a bloated pile of crap on Mac OS, but I can't verify that.
Where the Zune HD falls short is on the device experience. It's every bit as capable as the iPod touch when it comes to playing music or movies (and perhaps even better), but the iPod touch has 70k+ apps. The iPod touch has real games from real developers (e.g SimCity); the Zune has a couple of games from Microsoft. The iPod touch has SSH clients; the Zune doesn't. The iPod touch has an email app, calendar, and other nice utilities; the Zune doesn't.
That's the problem with the Zune. The Zune HD is a superior device to anything that Apple has ever made, excluding the iPod Touch. But that's the problem - they aren't competing with the Nano, they're competing with the Touch. And they fall short.
The problem is that no one in North America is implementing a European-style cell billing system.
No, the problem is that the US doesn't distinguish between landline numbers and mobile numbers. We pay to receive instead of having the caller pay out the rear for the privilege of calling a mobile phone.
Go actually compare rates in North America and Europe. You'll find that European providers offer lower-priced options, and that there are more prepaid options as well. But in the price categories that most people in the US pay, US carriers are actually quite competitive.
T-Mobile UK offers 1GB of mobile data, 1000 minutes, and unlimited SMS for £40.50/mo (about $66/mo).
Sprint offers "unlimited" (realistically, 5GB) of mobile data, 450 minutes, and unlimited SMS/MMS for $70/mo. The 450 minutes looks really bad until you consider the fact that Sprint doens't count calls made after 7pm, on weekends, or to any mobile phone in the US against that total. Only calls made to landlines on weekdays before 7pm count.
Which is better? If you call landlines a lot, the T-Mobile UK plan is better. If you use lots of mobile data or mostly call mobile phones (or at night or on weekends), the Sprint plan might be better.
The point is that it isn't a huge gulf like people seem to believe.
The TI-89 runs for weeks/months on a set of AAAs, boots quickly, and has a bunch of nice hardware keys that make it perfect for entering math problems.
Yes, it's too expensive. But it's a darn useful device.
Slashdot Mirror
These devices are usually wired for 220V in the US:
- Electric clothes dryer
- Electric ovens
- Electric stoves
Dishwashers run fine on a 15A 110V circuit. So do washing machines.
Having actually used a SiCortex machine, I can tell you that the problem wasn't the VC, or the compilers, or even really the hardware.
The problem was the market.
There are two types of x86-based small clusters (the market that SiCortex was aiming for): clusters with Gigabit Ethernet and clusters with expensive interconnects (Mirinet, InfiniBand, or 10G Ethernet).
Gigabit Ethernet clusters do a good job with problems that are embarrassingly parallel (or at least have minimal communication demands). $150k gets you 300 Nehalem cores and a lot of memory. SiCortex fails here because their competition (the SC1458) is much more expensive and much slower. The fact that the SC1458 uses less power (around 5kW instead of 10kW) is impressive, but unless you're very power or cooling constrained, it's simply more cost effective to deal with the extra power and cooling cost.
SiCortex hardware was more cost effective against clusters with expensive interconnects. The problem is, the people who buy clusters with expensive interconnects do so because their problem is interconnect heavy. Unfortunately, despite all of the cool CS behind SiCortex's interconnect, the fact is that it just didn't do that well against InfiniBand. That's partly because the SiCortex system has more nodes, which means that more messages have to use the interconnect. It's partly because for very small clusters, it's possible to use a single IB switch that connects every node to every other node. And it's partly because SiCortex didn't have the kind of mature hardware/software stack that someone like Mellanox has.
So, there you have it. For the problems that ran well on SiCortex hardware, you could get the same performance at dramatically lower cost using Gigabit Ethernet. For the problems that require an expensive interconnect, the SiCortex approach of "more, smaller nodes" results in dramatically more overhead compared with the "fewer, faster nodes" strategy.
This is impossible in a Toyota hybrid, because the gears are never 'shifted' and always remain in the same mechanical configuration.
HOWEVER, it takes active computer control to actually deliver power. With both motors disconnected, the gears are free to spin and the engine cannot deliver power even if it were running.
I own a $400 Acer 1410 laptop and two self-built desktops, one AMD and one Intel. I also recently sold a ThinkPad T61 and an Acer EEE PC 900HA.
ALL of them sleep and wake up in less than 3 seconds. My Acer wakes up in about a second (it has an Intel SSD that I installed); the others take a couple of seconds to spin up the hard drive.
The "Macs sleep and wake and PCs don't" argument is total bullshit. Find me one machine that's sold in Best Buy that doesn't sleep and wake properly.
I have owned a TON of hardware over the last 5 years, including the following:
- Acer EEE PC 900HA
- HP Compaq 6910p
- ThinkPad T61
- MSI Wind Nettop 100
- Custom-built AMD dekstop (GeForce 6100 chipset, Gigabyte motherboard)
- Custom-built Intel desktop (Intel G31 chipset, Gigabyte motherboard)
- Acer Aspire 1410
- Toshiba Portege M200 tablet PC
- Compal EFL30 generic notebook
All of this hardware properly sleeps and wakes properly.
My Acer 1410 was $400. It's 3lbs, has an 11.6" 1366x768 display, 4GB of memory, runs for 6-8 hours on a charge, has an HDMI port (including audio support), and a full size keyboard.
My Acer is made of plastic. It doesn't have a unibody case. It doesn't have a glowing Apple logo. It doesn't run Mac OS X.
But, you know what? I would take it over any Mac at any price (assuming that I couldn't sell the Mac). I have a desktop when I need computing power. My laptop needs to be light, small, have a usable keyboard and screen, and get good battery life.
You're quoting RoughlyDrafted on Steve Jobs? That's like quoting The Enquirer about celebrity news.
Seriously, go read some of the shit that's on RoughlyDrafted. Daniel Dilger makes Walt Mossberg look like a Microsoft fanboy.
Well, to be honest, I'm mixed about your comment.
On one hand, I think that understanding the hardware is still an important part of producing code that runs well. On modern high-end CPUs, it's more about things like memory accesses than it is about optimizing out every last instruction, but it's still critical.
On the other hand, it's unreasonable to expect that someone with an undergraduate degree would understand how a modern out-of-order CPU works. At best, we might hope that students get some understanding of data dependencies and simple out-of-order concepts like Tomasulo's algorithm and register renaming. Of course, every student should understand how caches work, why having unpredictable branches is bad, and just how badly you're screwed when you start hitting main memory frequently.
The thing is, compilers already do a pretty good job with these kinds of optimizations. You can do brain-dead things like access multidimensional arrays in the wrong order, and the compiler will often (but not always) save your butt. You can put function calls all over the place, and many of them will get inlined. You can use constant variables instead of hardcoding values, and it won't matter.
So, yes, if you're running on a system that's esoteric, or doing something that's unusual, you can sometimes dramatically improve performance by understanding the machine code and the hardware. But when you're running on a platform that's well known (x86 or ARM, for example), using a good compiler (GCC counts), you can be surprisingly dumb and still get good performance.
If everyone were a systems expert, we would probably all be using VLIW systems right now.
ShipIt was great. It's not that I ever used the CDs myself, but they were great to give away at LUG meetings, installfests, and other events. There's something to be said for a pressed CD in nice packaging (compared with a burned CD).
That said, when you've got 30Mbps at home, CDs in the mail seem a bit silly.
That's just false. There are significant anti-vaccine campaigns at least in the UK and Australia, and likely many other places that I don't know about.
This is a serious global problem. As you point out, not getting a vaccine puts everyone else at increased risk. If the anti-vaccine crowd gets their way, we could very well see resurgences in diseases such as Whooping Cough.
The reason it happens is that people are bad at understanding risk. The risk of dying to Influenza in a typical Flu season? About 1 in 10,000 (less if you're a healthy adult, more if you're in a high-risk group). That's about the same as your risk of being killed in a car accident.
Yes, you can. Intel, AMD (ATI), and NVIDIA both release drivers for Vista/7 64-bit and Vista/7 32-bit at the same time.
You can switch on the ThinkPad T400 without restarting or logging out, as long as you're running Windows Vista or Windows 7.
No, it's not. The reality is that while hard drive failures get most of the attention, disks are actually pretty reliable and rarely fail catastrophically. In my case, that means never - and I've owned and used dozens of drives from multiple manufacturers.
In contrast, human mistakes are far more common. Deleting a file accidentally, overwriting it with a bogus version, or just wanting to revert to an earlier revision all happen more often than the relatively rare disk failures.
That's not to say that a disk can't fail, that your laptop couldn't be lost or stolen, or that external backups aren't superior. They are.
However, Windows 7 (and Vista Business/Enterprise/Ultimate) have a feature that's ultimately more useful than Time Capsule: previous versions. Previous versions doesn't protect you from hardware failures, but it's transparent, backs up to the same disk, and most importantly is on by default. The fact that Time Capsule requires a separate partition (or an external drive) to activate means that it's not going to be used by a substantial fraction of Mac users. Previous versions, in contrast, protects everyone who hasn't explicitly disabled it.
The only problem with the Mac Mini is that it's pricey. You can get a similarly-small ION-based dual-Atom box for $330 (including 2GB of memory, disk, and DVD drive), just over half the price of the Mac Mini.
Or, if you're even cheaper, you can build a full-size Pentium Dual-Core box for around $250, which has the added advantage of multiple SATA ports and plenty of room for extra disks (presumably you want your media box to be able to store media). Of course, it won't fit nicely under your TV.
On Windows, the Firefox installer is signed using Authenticode using a certificate signed by a "trusted" root CA (e.g. Verisign). When you see "Mozilla Corporation" come up as the "Verified Publisher" in the UAC dialog (in Vista or 7), that string is not simply a field in the executable, it's part of the certificate that Mozilla uses to sign Firefox installers.
Now, as to whether or not you should trust the root CAs, that's a different question. But if you can't trust the root CAs, then it doesn't really matter whether your Firefox build has a modified root CA list.
On Linux, your Firefox packages will be signed with your distro's private key(s).
So 57 years ago France was already broadcasting 441 lines. I was under the impression, that in the USA, today, that 480 lines were being broadcast and sold as the low end of HD.
You would be wrong. NTSC (US analog) standard definition is 525 vertical lines, of which 486 are visible (plus or minus several depending on overscan). This is usually referred to as "480i60", as in, 480 vertical lines, interlaced, 60 fields per second.
The ATSC over-the-air television standard used in the US can carry a variety of formats. It can carry "standard definition" 480i60 (mostly used for lower cost stations, older programming, digital subchannels, and the like). But ATSC can also carry 480p60 "enhanced definition" signals, though this is seldom used.
HD, in the context of television in the US, universally refers to one of two ATSC formats: 720p60 or 1080i60.
480i60 has never been considered "HD" by anyone who has a clue. 480p60 has been marketed as "enhanced definition", but it's frequently just considered SD today.
Personally, I prefer 720p60 for content originally authored for TV (LCD TVs are progressive, so 1080i60 requires deinterlacing which creates motion artifacts). For content that comes from a film source, 1080i60 performs identically to 1080p24, assuming that the telecine is properly done and the deinterlacer in the TV (or set-top-box) can detect film content properly.
Technically, ATSC supports 1080p24 (also 1080p23.976), along with a bunch of other formats. But these formats aren't used in practice, and it's unclear whether hardware even supports them.
The trick to making strong passwords is to not use them at all.
Random passwords don't work. People don't remember them, or they write them down, or they use the same one everywhere. Any of these options compromises the security of a 'bulletproof' random password.
SSH private keys can't be guessed, they aren't compromised if you use them on more than one system (even untrusted systems), and you can revoke them if the machine they are on is compromised.
Better yet, smart cards are even harder to clone, especially if you don't have physical access to the card.
I have a pair of 413s, but they are too big to fit me.
Cablevision offers 100Mbps for $99/mo in the US. Comcast offers 50Mbps for $99/mo as well.
Comcast has more customers than there are people in Korea. They will achieve 80% DOCSIS 3 coverage by the end of 2009. Delivering "100Mbps" is as simple as updating a configuration file. The problem is that the contention ratio would be horrible.
Guess what, though? The contention ratio is horrible with ANY 100Mbps service. Fiber doesn't change that because you still need to backhaul the data through something. 5000 100Mbps subscribers with a contention ratio of 20:1 is 25Gbps, and I can bet you that they aren't using an OC-768 / STM-256x for every 5000 subscribers.
My university offers "1Gbps" broadband (the LAN is 1Gbps and 10Gbps, and the Internet connection is 2Gbps). But if more than 10 people try to use 1Gbps, the network can't deliver.
You don't connect the phone to the wall using the battery voltage. 330W at 48V is a very reasonable 6.9A. You step down the voltage to the correct charge voltage using a DC-DC converter. Models on PC motherboards already handle 150A or more.
Now, how you keep the battery from catching fire while you do that is a completely different question.
No, most of the rest are normal students who happen to play baseball or soccer after school.
At my high school, enrollment in AP courses correlated highly with participation in athletics. The football team was pretty much filled with top students.
It's all in how you set up your program. Academic requirements to participate in athletics go a long way.
The Tesla Roadster is assembled in California using parts from the US and other countries. The frame comes from Lotus in the UK, but it is most definitely not a 'British' vehicle.
Tesla is the only company producing street-legal full-size BEVs today (not NEVs). They are developing a sedan as well. They target the luxury market because it's the only way to get people to pay the $50k+ that's necessary to build a useful BEV today.
Well, speaking as someone who has actually watched usability testing at Microsoft and as someone who has actually seen significant changes made to a product because of it, I can tell you that you're full of crap.
Bitching that the clipboard preserves formatting is a little silly. It's normal, expected behavior, and you can use the 'smart tag' (since Office XP) if you don't want to preserve formatting. The fact that you don't like the default behavior doesn't make it wrong.
I rarely notice auto-bulleting in Word because I don't start my lines with dashes. If I wanted to type plain text, I would use a text editor.
If your point in all of this is that MS products have usability flaws, well, I'd agree. But then so do Apple's. I learned that the first time I wiped my iPod by clicking on the wrong button when I connected it to another PC.
The usability studies do matter, and they do improve the product. Perhaps not as much as we might want, but to say that they are a show is simply silly.
Good post but I take exception with one thing:
The Zune desktop software in my opinion is vastly superior to iTunes. It's prettier, has a bunch of 'social' features (which sound stupid, but it's actually cool to see what your friends are listening to), and, in general, faster. I've heard that iTunes isn't such a bloated pile of crap on Mac OS, but I can't verify that.
Where the Zune HD falls short is on the device experience. It's every bit as capable as the iPod touch when it comes to playing music or movies (and perhaps even better), but the iPod touch has 70k+ apps. The iPod touch has real games from real developers (e.g SimCity); the Zune has a couple of games from Microsoft. The iPod touch has SSH clients; the Zune doesn't. The iPod touch has an email app, calendar, and other nice utilities; the Zune doesn't.
That's the problem with the Zune. The Zune HD is a superior device to anything that Apple has ever made, excluding the iPod Touch. But that's the problem - they aren't competing with the Nano, they're competing with the Touch. And they fall short.
No, the problem is that the US doesn't distinguish between landline numbers and mobile numbers. We pay to receive instead of having the caller pay out the rear for the privilege of calling a mobile phone.
Go actually compare rates in North America and Europe. You'll find that European providers offer lower-priced options, and that there are more prepaid options as well. But in the price categories that most people in the US pay, US carriers are actually quite competitive.
T-Mobile UK offers 1GB of mobile data, 1000 minutes, and unlimited SMS for £40.50/mo (about $66/mo).
Sprint offers "unlimited" (realistically, 5GB) of mobile data, 450 minutes, and unlimited SMS/MMS for $70/mo. The 450 minutes looks really bad until you consider the fact that Sprint doens't count calls made after 7pm, on weekends, or to any mobile phone in the US against that total. Only calls made to landlines on weekdays before 7pm count.
Which is better? If you call landlines a lot, the T-Mobile UK plan is better. If you use lots of mobile data or mostly call mobile phones (or at night or on weekends), the Sprint plan might be better.
The point is that it isn't a huge gulf like people seem to believe.
The TI-89 runs for weeks/months on a set of AAAs, boots quickly, and has a bunch of nice hardware keys that make it perfect for entering math problems.
Yes, it's too expensive. But it's a darn useful device.