This is nothing but another way of locking down open code. Software patents taint open source code and restrict its distribution. This kind of patent-locked open-source is another thing that the GPL version 3 was designed to protect against. If your software is licensed as GPLv3, these jerks can't take it and say "this guy can run it, but this other guy can't unless he pays us". They must grant permission to use the software (and their patents therein) to every user of the software (which can redistribute it freely), or they cannot distribute the software at all.
Modern cars have two or more control networks. The class-1 network controls things vital to the car operation and safety such as the anti-lock brakes, air bags, and steering. The class-2 network(s) are for things such as rolling down your windows, controlling your CD changer, and turning on your headlights. NOTHING is allowed on the class-1 net without rigorous validation. If your satellite radio module goes bad, it won't stop you from being able to safely control your vehicle. And these are just control networks, they are not allowing hundreds of users to bring in their personal computers and an Internet connection.
Reading the story, it seemed like they wanted the airplane's maintenance systems to communicate with ground crews over the Internet, as well the aircraft reporting status to the airline while in flight. Personally, I'm uncomfortable with any part of the aircraft's vital systems being on the Internet.
So that when Windows wants to secretly download an update or send your data back to Microsoft, and you prevent them from doing so at the router level, they'll be able to detect it? No. A black hole router is a router that incorrectly handles MTUs that are bigger than it can pass. That is, instead of fragmenting the packets, it just silently drops them. This makes for some very unreliable connections as only the bigger packets get dropped and smaller ones get through. This is usually a problem at the ISP level and has nothing to do with Windows updates. I now return you to your regularly scheduled tin foil hat.
You make good points, AC. However, it's not the taxes I'm worried about, it's the salaries. Perhaps their cost of living is comparable to mine _for as long as they're in the country_. But their families are in their home countries, where they spend much less. When they return home, they will buy houses at their local rates and they will retire at their country's cost of living level, so they need to save far less than I do (since I will live in the US for the rest of my life). The rates the MS contractors were offering were well below market rates for my experience level. There was audible stammering when I told them what my salary requirements were, they were hoping for something around 60% of that. The fact that they expect to be able to hire someone that matches my experience at that price tells me that there is downward price pressure. I do not fear the H1Bs. I want them to be paid *more* so that companies like MS don't have an excuse to low-ball local software developers.
As for outsourcing, I've been part of it. I agree that it's bad. But it's not terminal. The reality is that anything that is marginalized will transfer offshore, but high quality stuff will stay here. Compare it to the manufacturing industries where all the low paying jobs moved moved to China, but advanced development and design and high precision manufacturing stayed in the U.S. China doesn't come up with anything new, they just copy. India doesn't create new software, they implement specs for something that is well-understood. I come up with something new. I am given a high level description of "make a product that does this," and I think it through. If you're writing marginalized software, reimplementing the same tired stuff or just coding to a spec you had no part in, you should be very worried about outsourcing.
See, now we see what IBM is really all about, using free software to make it easier to throw more people out of work in the states. At least with Microsoft, there's developers actually working in the USA.... Funny you should say that. I've been interviewing for new jobs recently. I've been contacted about a dozen times in the past two weeks for jobs at Microsoft. However, I wasn't contacted my Microsoft. I was contacted by nice ladies and gentlemen with heavy Indian accents who work at companies that place consultants at MS. They were quick to tell me that they sponsor H1Bs. Now, I have nothing against foreign developers and worked with several very talented ones, but it's become apparent to me that MS is turning into an H1B farm, and as an American developer, I don't particularly want to work in that kind of company. But at least you're proud that their products are made in the USA. Now how about we get it made by (mostly) Americans?
That's an easy one: Make it run artificially worse on AMD processors. (See also: Skype) You mean they will kill the port to AMD's ATI-brand GPUs, which are moving into physics simulation and need the Havok engine which runs many games.
I second this. This code is great. It's not for newbies, but the OP's whining is totally unfounded. It is well engineered, well documented, highly readable, and doesn't do stupid things. This is what highly professional code looks like.
I'd say your skepticism is warranted.. I have somewhere in a stack of papers a popular mechanics article from 1986 (yes, 21 years ago), that claims the Moller Skycar will be out in a few years and listing those exact specs. I'm all out of skepticism, all I have left is disbelief.
So they had to switch to Exchange in order to get backups? Sounds like they lost out on that deal. They'll need those backups when Exchanged decides to barf all over the mail database files. On the other hand, LMU is a bastion of hot party girls, so I don't blame them.
Hey Jeremy, I just wanted to thank you and the rest of the Samba developers for moving to GPLv3 and generally doing what it takes to keep free software free.
How much do you want to bet they're not repairing jack shit, just giving you a new phone? My back of the envelope calculation yields the price of manufacturing one iphone to be about $200. That's what it costs Apple to get one out of a factory in China, we're not even counting shipping. The price of replacing a battery in your unit is probably around $10 to $20 per piece (with labor), and it can be done in a warehouse in the U.S. Plus, they don't have to book the unit they took back from you as a return on their books. I'm going to bet that they refurb.
I'm not quite sure what the point would be to take the elevator all the way to the centrifugal counterweight at the far end of the cable. I will be taking the lift to the tension's midpoint, where I'd be weightless and it would require almost no energy to place things in orbit. But to each his own, I guess.
I'm working on a space escalator. Sure, it's not as fast getting up there, but you don't have to wait for the car to come back down from orbit when you press the up button. To get down quickly, there's also a space firehouse pole.
In all seriousness, though, I wish the LiftPort guys luck. I'm not sure how feasible it is, but I'd rather have people investing in creative, sometimes radical technologies than just sitting back and saying "no, that'll never work".
No. Statistics can never predict the outcome, they can only give you a probability of an outcome. That is, of course, unless the probability is 0 or 100%.
The BSDs most certainly do not use the GNU libc. While it is true that you cannot compile the system without gcc, you can definitely have a running BSD system with no GNU tools installed. It would be fairly bare bones (back to csh), but it's possible.
Many posters are commenting here that FAT works across all operating systems and that's why it's being used. If these manufacturers came out with a new file system specification (say, based on BSD UFS), I doubt it would be a big deal for Microsoft, Apple, and the Linux Kernel developers to include it in there.
The reasons we are stuck with FAT is: 1. Simplicity. This is huge for embedded devices (IE, the things that do the writing to all of these cards). A read-only FAT driver can be implemented in a few kilobytes of (compiled) code. It requires trivial amounts of memory to operate (only a few hundred bytes). I've written a bootloader for an embedded product that could load an OS from a FAT partition and it was under 10 kilobytes. A read-write implementation is not much bigger and the memory requirements are similarly trivial. No other major file system out there can claim this. Particularly, modern file systems like NTFS require huge amounts of memory (comparatively) due to the complex structures they need to maintain, and have massive, complex code to read and write.
2. Reliability. I know this seems counterintuitive for such a lousy file system, but FAT is fairly resilient both to power failures (or card yanks), and more subtle corruption such as bad drivers or media defects. Sure, it may corrupt and lose your file, but it very rarely destroys the entire file system and lose the rest of the files on there. This is again because of the simplicity of the structures and the fact that very little needs to change on disk when a modification is made. Remember how many times Windows 95 crashed? How many times of that did you get major FS damage? Compare and contrast with Ext2.
So, yes, FAT is a terrible file system compared to modern ones. But there's a reason everyone uses it.
I guess we can make all future computer science courses one week then, since every problem can be solved by map-reduce!
What do you mean? I just wrote a clone of Half-Life 2 using only Map-Reduce. It works great, but it requires 600 exabytes of pre-computed index values and data sets and about 200 high end machines to reach decent frame rates.
They have more program manager than they have developers. Haha, yes, this is what I noticed, too.
They have NINE managers and SIX developers. They probably sit in meetings all day just like the rest of Microsoft. Actually, I found the whole page hilarious, due to its forced-sounding attempt at being cool. Sounds like it was written by the same people that did the Zune marketing. Notice the contradiction between the two sentences of being right in the center of Microsoft while being a "startup".
The team is a small band of folks with a passion for democratizing development, housed within Microsoft's Developer Division based in Redmond, Washington. Like most startup ventures, the team hustles for resources every day and is innovative, scrappy, and fun. Oh, and we also dream big.
Slashdot Mirror
This is nothing but another way of locking down open code. Software patents taint open source code and restrict its distribution. This kind of patent-locked open-source is another thing that the GPL version 3 was designed to protect against. If your software is licensed as GPLv3, these jerks can't take it and say "this guy can run it, but this other guy can't unless he pays us". They must grant permission to use the software (and their patents therein) to every user of the software (which can redistribute it freely), or they cannot distribute the software at all.
Modern cars have two or more control networks. The class-1 network controls things vital to the car operation and safety such as the anti-lock brakes, air bags, and steering. The class-2 network(s) are for things such as rolling down your windows, controlling your CD changer, and turning on your headlights. NOTHING is allowed on the class-1 net without rigorous validation. If your satellite radio module goes bad, it won't stop you from being able to safely control your vehicle. And these are just control networks, they are not allowing hundreds of users to bring in their personal computers and an Internet connection.
Reading the story, it seemed like they wanted the airplane's maintenance systems to communicate with ground crews over the Internet, as well the aircraft reporting status to the airline while in flight. Personally, I'm uncomfortable with any part of the aircraft's vital systems being on the Internet.
You make good points, AC. However, it's not the taxes I'm worried about, it's the salaries. Perhaps their cost of living is comparable to mine _for as long as they're in the country_. But their families are in their home countries, where they spend much less. When they return home, they will buy houses at their local rates and they will retire at their country's cost of living level, so they need to save far less than I do (since I will live in the US for the rest of my life). The rates the MS contractors were offering were well below market rates for my experience level. There was audible stammering when I told them what my salary requirements were, they were hoping for something around 60% of that. The fact that they expect to be able to hire someone that matches my experience at that price tells me that there is downward price pressure. I do not fear the H1Bs. I want them to be paid *more* so that companies like MS don't have an excuse to low-ball local software developers.
As for outsourcing, I've been part of it. I agree that it's bad. But it's not terminal. The reality is that anything that is marginalized will transfer offshore, but high quality stuff will stay here. Compare it to the manufacturing industries where all the low paying jobs moved moved to China, but advanced development and design and high precision manufacturing stayed in the U.S. China doesn't come up with anything new, they just copy. India doesn't create new software, they implement specs for something that is well-understood. I come up with something new. I am given a high level description of "make a product that does this," and I think it through. If you're writing marginalized software, reimplementing the same tired stuff or just coding to a spec you had no part in, you should be very worried about outsourcing.
Assuming, of course, you can find someone to buy the stock from you
Er, assuming you can find someone to buy from. I think there are more shorts than holders.
Assuming, of course, you can find someone to buy the stock from you. And the stock doesn't go to zero.
I second this. This code is great. It's not for newbies, but the OP's whining is totally unfounded. It is well engineered, well documented, highly readable, and doesn't do stupid things. This is what highly professional code looks like.
So they had to switch to Exchange in order to get backups? Sounds like they lost out on that deal. They'll need those backups when Exchanged decides to barf all over the mail database files. On the other hand, LMU is a bastion of hot party girls, so I don't blame them.
Hey Jeremy, I just wanted to thank you and the rest of the Samba developers for moving to GPLv3 and generally doing what it takes to keep free software free.
It's the Mechanical Turk approach. Amazon is doing it.
I'm not quite sure what the point would be to take the elevator all the way to the centrifugal counterweight at the far end of the cable. I will be taking the lift to the tension's midpoint, where I'd be weightless and it would require almost no energy to place things in orbit. But to each his own, I guess.
I'm working on a space escalator. Sure, it's not as fast getting up there, but you don't have to wait for the car to come back down from orbit when you press the up button. To get down quickly, there's also a space firehouse pole.
In all seriousness, though, I wish the LiftPort guys luck. I'm not sure how feasible it is, but I'd rather have people investing in creative, sometimes radical technologies than just sitting back and saying "no, that'll never work".
Can Statistics Predict the Outcome of a War?
No. Statistics can never predict the outcome, they can only give you a probability of an outcome. That is, of course, unless the probability is 0 or 100%.
The BSD projects still use gcc and GNU C library
The BSDs most certainly do not use the GNU libc. While it is true that you cannot compile the system without gcc, you can definitely have a running BSD system with no GNU tools installed. It would be fairly bare bones (back to csh), but it's possible.
Here's a link to the OpenBSD libc for your browsing pleasure.
Many posters are commenting here that FAT works across all operating systems and that's why it's being used. If these manufacturers came out with a new file system specification (say, based on BSD UFS), I doubt it would be a big deal for Microsoft, Apple, and the Linux Kernel developers to include it in there.
The reasons we are stuck with FAT is:
1. Simplicity. This is huge for embedded devices (IE, the things that do the writing to all of these cards). A read-only FAT driver can be implemented in a few kilobytes of (compiled) code. It requires trivial amounts of memory to operate (only a few hundred bytes). I've written a bootloader for an embedded product that could load an OS from a FAT partition and it was under 10 kilobytes. A read-write implementation is not much bigger and the memory requirements are similarly trivial. No other major file system out there can claim this. Particularly, modern file systems like NTFS require huge amounts of memory (comparatively) due to the complex structures they need to maintain, and have massive, complex code to read and write.
2. Reliability. I know this seems counterintuitive for such a lousy file system, but FAT is fairly resilient both to power failures (or card yanks), and more subtle corruption such as bad drivers or media defects. Sure, it may corrupt and lose your file, but it very rarely destroys the entire file system and lose the rest of the files on there. This is again because of the simplicity of the structures and the fact that very little needs to change on disk when a modification is made. Remember how many times Windows 95 crashed? How many times of that did you get major FS damage? Compare and contrast with Ext2.
So, yes, FAT is a terrible file system compared to modern ones. But there's a reason everyone uses it.
I guess we can make all future computer science courses one week then, since every problem can be solved by map-reduce!
What do you mean? I just wrote a clone of Half-Life 2 using only Map-Reduce. It works great, but it requires 600 exabytes of pre-computed index values and data sets and about 200 high end machines to reach decent frame rates.
Haha, yes, this is what I noticed, too.
They have NINE managers and SIX developers. They probably sit in meetings all day just like the rest of Microsoft. Actually, I found the whole page hilarious, due to its forced-sounding attempt at being cool. Sounds like it was written by the same people that did the Zune marketing. Notice the contradiction between the two sentences of being right in the center of Microsoft while being a "startup".
I hope they don't forget to include a field describing how delicious each life-form is (and perhaps how it is best served).
TJX - commonly known to American consumers as TJ Max and Marshalls retail stores. If you made purchases at these stores, you could be affected.
Pretty much every paging-capable microprocessor in existence uses 4K memory blocks, thus why they're the natural size for a hard disk.
AMD64/x86-64 uses 8KB pages. ARM uses 1KB pages.