And going back to 2 gig memory limit and 32 bits is going to be really fun.
... the transition is scheduled over the next two years, and Intel is already producing 64 bit processors. For 99% of people, this really is an insignificant change -- the people with the roughest job are in Apple's marketing department.
Problem is, some of those images took over 10 hours to render using multiple computers. You're still looking at a serious investment in horsepower if you want to produce a 130k frame film (that's about 90 minutes)...:)
The whole point of the GPL is that users can make modify the code.
This is true.
If the deriviative code they have released cannot be loaded without rendering the unit unusable, then they have clearly violated the spirit of the GPL. Maybe they've found some kind of loophole, I don't know.
This is not true. The GPL's No Warranty section explicitly states:
THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
In short, the GPL makes no guarantee that the code you receive is functional, only that you must release it upon request. If it doesn't run on the platform you have available, that's your problem.. but I'm sure the community would encourage you find a work around!
Regardless, it is perfectly legal for a proprietary system to checksum a GPL'd binary to ensure its legitimacy. Commercial security tools, such as Tripwire, do this already. Antivirus BIOSes will checksum your boot sectors. Heck, I'm sure there are LSMs (Linux Security Modules) out there that md5 binaries to ensure legitimacy.
Granted, you said "spirit of"... but I think the GPL is pretty explicit in this particular case.
Hehe, well, I'm an American, I just happen to travel a lot.;)
My point about arrogance is that technically it's not a difficult problem, and the majority of people who are online these days regularly use characters outside the ASCII chart. Your point about inventor's control is fair, but it doesn't really apply here since the original inventors gave up control of the system some decades ago, and the various committees in charge have all embraced internationalization.
So, I'm not irritated with the inventors -- far from it! I'm quite pleased with DNS, as my memory for numbers is horrid. Heh. However, I am irritated with people who uphold the idea that alternate character sets have no place in DNS (or any other computing system, really).
This seems like a GOOD THING! Honestly, restricting the DNS universe to a small subset of ASCII is simply arrogance. I want my umlauts and graves, durnit.
Besides, if they already have ways of restricting access to Internet sites -- adding a character set to a level 7 protocol isn't a practical way to censor anything.
What, exactly, is corrupt about GIS? Maybe I missed something, but that looks like a demographics tool. I don't think I saw a HEATHENS chart or a SMITE button.
... not to nit pick, but this is a false statement:
The fast food industry are the main reason that things like Ecoli is now so common in store bought foods.
The rate of food borne pathogens has gone down dramatically in the last 100 years, and the trend continues through the advent of fast food in the last 50 years. In particular, E. coli contamination in particular has been reduced by over 15% since 1999, when substantial advances were made in the quantity and sensitivity of testing.
So, Ruby has now reached the point where it is a serious contender. Expect it to start replacing J2EE somewhere between 2008-2010.
Java is big for two big reasons:
One, it's a pretty good language. Not great, not fabulous, but as an alternative to C++ it was major step in the right direction at the time it arrived on the scene. It came at the right time, from the right place. On the other hand, Ruby is a pretty darned cool language, but it's not as special as Java was at the time -- it's one of a half dozen popular high level scripting languages in a relatively lean marketplace.
Two, Java started with a major marketing budget, has had 10 years of Big Corporate Pimping behind it, successful high visibility lawsuits, and the enormous growth of the dotcom era. Today, there is no political leverage to be gained for businesses who want to adopt or push Ruby in the marketplace.
Ruby isn't a revolutionary language/environment, like Java was, and I don't expect it will be replacing J2EE in the business world any time in the next 25 years. Heh. Our children will probably be cursing Java, just as we curse COBOL and Fortran.;)
I like Ruby, but I really don't think it'll be the next Java unless a major business steps up to the plate and makes it the cause du jour. However, it wouldn't surprise me if it followed the same path as Perl as a phenomenally successful tool, rather than a phenomenon in and of itself.
Bah, making publishers pay is for politicians, and we all know how that ends up. End users should pay to play, but they need a much more flexible payment system that scales down to individual students and engineers.
What if I could buy a single copy of a current paper for $10? Maybe something from last year for $5? Perhaps an advanced copy of the proceedings for an upcoming event for $1000?
Maybe I want a PDF... they're pretty easy to protect these days in a variety of ways (no printing! no saving!) -- most people don't have the patience to crack crypto or pirate Hot Engineering Papers.
Then again, maybe I want to festoon my office with official looking manuscripts... so I'm willing to pay a bit more for a paper copy.
Maybe I'd like an account with "my.ieee.org," where I could pay a low flat rate for access to a variable collection of material for a set period of time -- like the O'Reily Safari system.
Maybe I'm teaching a course with 200 students, and there's a few important papers I want them to read and reference. Maybe I work at a high school with an engineering program.
Maybe the library's closed, I have a paper due tomorrow, and I'm scrambling to complete a paper. Would I pay $10 for an IEEE online day pass? Sure thing!
Really -- there are tons of markets they're missing on the low end. I'm sure they can maintain their high end subscription services, but there's quite a few us also wanna read these things...
Putting a typeR sticker on your car might make you a poser, but it doesn't necessarily mean you'll be driving like an asshole. Someone is not a danger to society simply because they have a wing on their trunk -- saying so would be the equivalent to calling someone out because they're asian, or black, or a woman.
The problem is not body kits and nitrous systems -- the problem is bad driving, and there's plenty of assholes out there who do that regardless of race, gender, or vehicle.
I don't doubt that "riceing" as a cultural phenomenon has a predisposition towards street racing, but don't throw the baby out with the bathwater. It's too easy and in very poor taste to condem the many for the actions of a few.
But it is of very direct interest to you if a neighbor plays with high explosives next to your house, because his actions pose an immediate and provable threat to you and your posessions.
Meanwhile, someone who puts -O999 in his CFLAGS and RAID-0 SATA drives in his rig, then says how cool --funroll-loops is in public forum... what danger does that put you in? How does this endanger your social, political, financial, or personal well being?
It's easy to point and laugh at the neighborhood kid with a Neon equipped with spoilers and excessive stickers (or in this case, a computer with overclocked CPUs and case windows), but really, what's the point?
Now, I can understand complaining about overly loud stereos booming down the street in the wee hours of the morning... but bitching about someone's hobby, which they do for fun, is about as lame as you can get.
Yeah, it may be "illogical." Yeah, it may be "a waste of time and money." But it's not your time, not your money, and quite obviously not your interest... so what's the fuss about?
In the United States, a lot of credit card companies are issuing single purchase numbers. I think a few of them allow you to set the amount available, than use the generated number to make the purchase. I think it's an excellent solution to online CC transactions, that doesn't require overhauling the whole transaction system.
I'm in Germany at the moment, and we have a pretty good system for transactions don't involve cash currency. Most people here don't use credit cards or cheques; they use bank issued debit cards, and bank transfers.
The debit card can only be used in person. You have to supply the card... there's no cheating by just providing the number or anything like that. Can't really use it for online transactions, but it's not meant for that. Cashiers are usually pretty meticulous about checking your signature, so you have relatively good physical security.
There's a surprising number of bank transfers... you use it for almost everything: rent, utilities, regular bills, paying your friends back, paying for things online, and just about anything except for general shopping.
For every bank transfer you make, you have to supply a transaction authorization number (TAN). When you open an account, you're given a sheet with a couple hundred of these numbers, and you have to use them in sequence. When you want more, you go to the bank, present a valid ID of somesort, and get another sheet.
It's a pretty good system, very convenient, but would require quite a bit of infrastructure changes in the US...
Before you condem hunters as anti-environment, I'd like to point out that one of the greatest conservationists in the history of the United States, and one of the finest presidents we've ever had, was an avid hunter.
Theodore Roosevelt added nearly a quarter billion acres to the national park system. That's an area similar in size to all of the states on the eastern seaboard, from Maine to Florida. He helped establish the concept of endangered species in the mind of the public, and was widely considered one of the world's foremost academic experts on mammals. The man was, quite simply, a genius.
Strangely enough, some of the world's top atheletes eat fast food on a regular basis... I've talked with a couple of them, a long distance runner and a cyclist (both at the olympics right now), who happily take down a double whopper with cheese and jumbo fries hours before competing. Some of these folks have to maintain 7-10K kcal diets. That's a heinous amount of food.
But yeah. Puttin' the moves on a Big Mac and a diet Coke while watching the world's best sweat it out.. that's pretty funny.
And having to show ID didn't stop the 9/11 hijackers; they all showed perfectly valid official IDs under their own names. So what's the point?
... actually, a significant number of the 9/11 hijackers could have been stopped because they showed ID -- several of them were flagged as potential threats. The failure was rooted the response of the security personel, and the policies that guided their actions. ID was a point in favor of the security process, even though the follow through was botched.
9/11 is a bad example for your specific case, because anyone who's read the commission's report will tell you that the IDing the passengers was a qualified success. For privacy advocates, travel ID is a dangerous red herring which can distract them from the main point:
It's entirely possible to provide an extremely high level of safety and security without infringing on anonymity.
You can remain perfectly anonymous and still have your baggage thoroughly searched, the airplane cockpit physically secured, and passengers guarded by armed air marshals while on the plane.
Today's technology can make a rover as smart as a cockroach..... brilliant. Now it'll have a disgusting fascination with my old ham sandwiches, and a tendency to scurry about aimlessly in the middle of the night.
The article states a volume of 35 million units in the first quarter of this year, with no sign of slowing down. This isn't trouble -- it's a competitive commodity market, like almost all the markets Chinese companies work in. They're quite accustomed to this sort of phenomenon, and a $1 profit margin ($140M/yr @ volume) really isn't that bad.
Because, aside from a few geeks scattered around the world, most people are interested in applications, and very, very few actually care about the operating system. A vast majority of users think that Windows came "free" with their computer anyway.
Also, a lot of Linux users are more than willing to pay for the ability to run a Unix with Microsoft (and Adobe) products. The significant number of OS X users who came from the Unix camp are a testament to that.
Actually, I think this shows a maturity in the Open Source model, as the community has spread deeply into the commercial realm.
Red Hat, IBM, Novell, SGI, Mandrake, and all the other companies that support, market, and contribute to Linux are a very significant part of the community. Many of the top kernel hackers are employed by these companies. They are heavily invested in both pushing the envelope with Linux, and supporting a solid and stable platform for their customers.
Given that the commercial vendors are the ones who are closest to the majority of end users and have a significant interest in the hardware Linux is deployed on, they are in the best position to deal with the majority of bugs which have to do with specific drivers and interfaces. In fact, it would hardly surprise me if most of the current bug finding and fixing comes from these companies. This seems to be their natural position in the community.
On the other hand, given the tremendous popularity of Linux in research and higher education, the core maintainers have to contend with a massive volume of architecture and implimentation issues. File systems, schedulers, memory managers, performance, scalability and QoS issues -- these are the bread and butter of the core team. The fact that someone neglected endian issues with a specific network controller should not be their problem.
Personally, I think this is less of a policy mandate than it is a recognition of how the community has changed over the last few years. Since the release of 2.4, people have embraced the MM and other trees as a testing ground for unproven additions to the stable kernel, while the odd series kernels have become more esoteric. I like this model, as it allows for major changes to vet themselves in the unstable branch, while continuing stepwise improvement in the unofficial stable branches (MM et al), and ensuring bug squishing and stability in the official vanilla release.
If Linux is to continue to mature and improve, this sort of delegation is necessary. These are growing pains, and I'm confident that the Linux kernel community (commercial or otherwise) has the intelligence to govern itself effectively.
Slashdot Mirror
I can get stain resistant pants?
SWEET.
Problem is, some of those images took over 10 hours to render using multiple computers. You're still looking at a serious investment in horsepower if you want to produce a 130k frame film (that's about 90 minutes) ... :)
The whole point of the GPL is that users can make modify the code.
This is true.
If the deriviative code they have released cannot be loaded without rendering the unit unusable, then they have clearly violated the spirit of the GPL. Maybe they've found some kind of loophole, I don't know.
This is not true. The GPL's No Warranty section explicitly states:
THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
In short, the GPL makes no guarantee that the code you receive is functional, only that you must release it upon request. If it doesn't run on the platform you have available, that's your problem .. but I'm sure the community would encourage you find a work around!
Regardless, it is perfectly legal for a proprietary system to checksum a GPL'd binary to ensure its legitimacy. Commercial security tools, such as Tripwire, do this already. Antivirus BIOSes will checksum your boot sectors. Heck, I'm sure there are LSMs (Linux Security Modules) out there that md5 binaries to ensure legitimacy.
Granted, you said "spirit of" ... but I think the GPL is pretty explicit in this particular case.
Hehe, well, I'm an American, I just happen to travel a lot. ;)
My point about arrogance is that technically it's not a difficult problem, and the majority of people who are online these days regularly use characters outside the ASCII chart. Your point about inventor's control is fair, but it doesn't really apply here since the original inventors gave up control of the system some decades ago, and the various committees in charge have all embraced internationalization.
So, I'm not irritated with the inventors -- far from it! I'm quite pleased with DNS, as my memory for numbers is horrid. Heh. However, I am irritated with people who uphold the idea that alternate character sets have no place in DNS (or any other computing system, really).
This seems like a GOOD THING! Honestly, restricting the DNS universe to a small subset of ASCII is simply arrogance. I want my umlauts and graves, durnit.
Besides, if they already have ways of restricting access to Internet sites -- adding a character set to a level 7 protocol isn't a practical way to censor anything.
What, exactly, is corrupt about GIS? Maybe I missed something, but that looks like a demographics tool. I don't think I saw a HEATHENS chart or a SMITE button.
... not to nit pick, but this is a false statement:
_ Stats.asp
The fast food industry are the main reason that things like Ecoli is now so common in store bought foods.
The rate of food borne pathogens has gone down dramatically in the last 100 years, and the trend continues through the advent of fast food in the last 50 years. In particular, E. coli contamination in particular has been reduced by over 15% since 1999, when substantial advances were made in the quantity and sensitivity of testing.
http://peaches.nal.usda.gov/foodborne/fbindex/FBI
(google is also helpful)
Java is big for two big reasons:
One, it's a pretty good language. Not great, not fabulous, but as an alternative to C++ it was major step in the right direction at the time it arrived on the scene. It came at the right time, from the right place. On the other hand, Ruby is a pretty darned cool language, but it's not as special as Java was at the time -- it's one of a half dozen popular high level scripting languages in a relatively lean marketplace.
Two, Java started with a major marketing budget, has had 10 years of Big Corporate Pimping behind it, successful high visibility lawsuits, and the enormous growth of the dotcom era. Today, there is no political leverage to be gained for businesses who want to adopt or push Ruby in the marketplace.
Ruby isn't a revolutionary language/environment, like Java was, and I don't expect it will be replacing J2EE in the business world any time in the next 25 years. Heh. Our children will probably be cursing Java, just as we curse COBOL and Fortran. ;)
I like Ruby, but I really don't think it'll be the next Java unless a major business steps up to the plate and makes it the cause du jour. However, it wouldn't surprise me if it followed the same path as Perl as a phenomenally successful tool, rather than a phenomenon in and of itself.
Bah, making publishers pay is for politicians, and we all know how that ends up. End users should pay to play, but they need a much more flexible payment system that scales down to individual students and engineers.
... they're pretty easy to protect these days in a variety of ways (no printing! no saving!) -- most people don't have the patience to crack crypto or pirate Hot Engineering Papers.
... so I'm willing to pay a bit more for a paper copy.
...
What if I could buy a single copy of a current paper for $10? Maybe something from last year for $5? Perhaps an advanced copy of the proceedings for an upcoming event for $1000?
Maybe I want a PDF
Then again, maybe I want to festoon my office with official looking manuscripts
Maybe I'd like an account with "my.ieee.org," where I could pay a low flat rate for access to a variable collection of material for a set period of time -- like the O'Reily Safari system.
Maybe I'm teaching a course with 200 students, and there's a few important papers I want them to read and reference. Maybe I work at a high school with an engineering program.
Maybe the library's closed, I have a paper due tomorrow, and I'm scrambling to complete a paper. Would I pay $10 for an IEEE online day pass? Sure thing!
Really -- there are tons of markets they're missing on the low end. I'm sure they can maintain their high end subscription services, but there's quite a few us also wanna read these things
Putting a typeR sticker on your car might make you a poser, but it doesn't necessarily mean you'll be driving like an asshole. Someone is not a danger to society simply because they have a wing on their trunk -- saying so would be the equivalent to calling someone out because they're asian, or black, or a woman.
The problem is not body kits and nitrous systems -- the problem is bad driving, and there's plenty of assholes out there who do that regardless of race, gender, or vehicle.
I don't doubt that "riceing" as a cultural phenomenon has a predisposition towards street racing, but don't throw the baby out with the bathwater. It's too easy and in very poor taste to condem the many for the actions of a few.
But it is of very direct interest to you if a neighbor plays with high explosives next to your house, because his actions pose an immediate and provable threat to you and your posessions.
... what danger does that put you in? How does this endanger your social, political, financial, or personal well being?
Meanwhile, someone who puts -O999 in his CFLAGS and RAID-0 SATA drives in his rig, then says how cool --funroll-loops is in public forum
It's easy to point and laugh at the neighborhood kid with a Neon equipped with spoilers and excessive stickers (or in this case, a computer with overclocked CPUs and case windows), but really, what's the point?
... but bitching about someone's hobby, which they do for fun, is about as lame as you can get.
... so what's the fuss about?
Now, I can understand complaining about overly loud stereos booming down the street in the wee hours of the morning
Yeah, it may be "illogical." Yeah, it may be "a waste of time and money." But it's not your time, not your money, and quite obviously not your interest
You think that's frustrating?
...
Don't feed the dog a sandwich.
That really blew my stack, about thirty hours later
In the United States, a lot of credit card companies are issuing single purchase numbers. I think a few of them allow you to set the amount available, than use the generated number to make the purchase. I think it's an excellent solution to online CC transactions, that doesn't require overhauling the whole transaction system.
... there's no cheating by just providing the number or anything like that. Can't really use it for online transactions, but it's not meant for that. Cashiers are usually pretty meticulous about checking your signature, so you have relatively good physical security.
... you use it for almost everything: rent, utilities, regular bills, paying your friends back, paying for things online, and just about anything except for general shopping.
...
I'm in Germany at the moment, and we have a pretty good system for transactions don't involve cash currency. Most people here don't use credit cards or cheques; they use bank issued debit cards, and bank transfers.
The debit card can only be used in person. You have to supply the card
There's a surprising number of bank transfers
For every bank transfer you make, you have to supply a transaction authorization number (TAN). When you open an account, you're given a sheet with a couple hundred of these numbers, and you have to use them in sequence. When you want more, you go to the bank, present a valid ID of somesort, and get another sheet.
It's a pretty good system, very convenient, but would require quite a bit of infrastructure changes in the US
Before you condem hunters as anti-environment, I'd like to point out that one of the greatest conservationists in the history of the United States, and one of the finest presidents we've ever had, was an avid hunter.
Theodore Roosevelt added nearly a quarter billion acres to the national park system. That's an area similar in size to all of the states on the eastern seaboard, from Maine to Florida. He helped establish the concept of endangered species in the mind of the public, and was widely considered one of the world's foremost academic experts on mammals. The man was, quite simply, a genius.
And, he loved to hunt.
Strangely enough, some of the world's top atheletes eat fast food on a regular basis ... I've talked with a couple of them, a long distance runner and a cyclist (both at the olympics right now), who happily take down a double whopper with cheese and jumbo fries hours before competing. Some of these folks have to maintain 7-10K kcal diets. That's a heinous amount of food.
.. that's pretty funny.
But yeah. Puttin' the moves on a Big Mac and a diet Coke while watching the world's best sweat it out
Residents of the former United States are chained to benches while being forced to eat poutine and watch curling competitions.
Whoa. That sounds exactly like my first few months at Simon Frasier.
Your insight scares me.
And having to show ID didn't stop the 9/11 hijackers; they all showed perfectly valid official IDs under their own names. So what's the point?
9/11 is a bad example for your specific case, because anyone who's read the commission's report will tell you that the IDing the passengers was a qualified success. For privacy advocates, travel ID is a dangerous red herring which can distract them from the main point:
It's entirely possible to provide an extremely high level of safety and security without infringing on anonymity.
You can remain perfectly anonymous and still have your baggage thoroughly searched, the airplane cockpit physically secured, and passengers guarded by armed air marshals while on the plane.
Today's technology can make a rover as smart as a cockroach.. ... brilliant. Now it'll have a disgusting fascination with my old ham sandwiches, and a tendency to scurry about aimlessly in the middle of the night.
I highly recommend reading this page about how money moves within the banking system, and how value is maintained.
It's a fun read, if you're the least bit interested in modern economics!
This is a standard shakedown. So what?
The article states a volume of 35 million units in the first quarter of this year, with no sign of slowing down. This isn't trouble -- it's a competitive commodity market, like almost all the markets Chinese companies work in. They're quite accustomed to this sort of phenomenon, and a $1 profit margin ($140M/yr @ volume) really isn't that bad.
Because, aside from a few geeks scattered around the world, most people are interested in applications, and very, very few actually care about the operating system. A vast majority of users think that Windows came "free" with their computer anyway.
Also, a lot of Linux users are more than willing to pay for the ability to run a Unix with Microsoft (and Adobe) products. The significant number of OS X users who came from the Unix camp are a testament to that.
On the other hand, maybe it's about time drivers pay attention to the other cars on the road.
Actually, I think this shows a maturity in the Open Source model, as the community has spread deeply into the commercial realm.
Red Hat, IBM, Novell, SGI, Mandrake, and all the other companies that support, market, and contribute to Linux are a very significant part of the community. Many of the top kernel hackers are employed by these companies. They are heavily invested in both pushing the envelope with Linux, and supporting a solid and stable platform for their customers.
Given that the commercial vendors are the ones who are closest to the majority of end users and have a significant interest in the hardware Linux is deployed on, they are in the best position to deal with the majority of bugs which have to do with specific drivers and interfaces. In fact, it would hardly surprise me if most of the current bug finding and fixing comes from these companies. This seems to be their natural position in the community.
On the other hand, given the tremendous popularity of Linux in research and higher education, the core maintainers have to contend with a massive volume of architecture and implimentation issues. File systems, schedulers, memory managers, performance, scalability and QoS issues -- these are the bread and butter of the core team. The fact that someone neglected endian issues with a specific network controller should not be their problem.
Personally, I think this is less of a policy mandate than it is a recognition of how the community has changed over the last few years. Since the release of 2.4, people have embraced the MM and other trees as a testing ground for unproven additions to the stable kernel, while the odd series kernels have become more esoteric. I like this model, as it allows for major changes to vet themselves in the unstable branch, while continuing stepwise improvement in the unofficial stable branches (MM et al), and ensuring bug squishing and stability in the official vanilla release.
If Linux is to continue to mature and improve, this sort of delegation is necessary. These are growing pains, and I'm confident that the Linux kernel community (commercial or otherwise) has the intelligence to govern itself effectively.