On face value, there appears to be nothing wrong with increased police powers, for example, the ability to detain somebody for significant periods of time if they are suspected of something, without allowing the detainee to contact their lawyer or make a phone call to the outside world. Law enforcement officials would only detain bad guys, right ?
The problem with this is that it is based on the assumption that the everybody within the law enforcement organisations involved are totally and 100% honest. Of course, this isn't the case.
Judicial oversight of things such as wire taps is there to try to ensure that these mechanisms aren't abused by corrupt, dishonest or overzealous law enforcement officials.
Sadly, it seems that, since 911, George W. Bush's adgenda is to minimise or remove Judicial oversight in the name of "security". I can only suggest that he believes that law enforment officials are 100% honest.
Why is there an assumption that Linux won't be successful if high FPS games aren't available for it ?
I don't care to play games under Linux (as I don't play games in general, they just don't hold my attention for some reason (probably biological)), so I don't care whether commercial games are supported. To me, Linux is already successful, and has been since I first started using it in 1993, as it has done everything I've wanted it to do.
"Success" is a relative word, and is used to measure something against a set of criteria. Change any of those criteria, and the former "success" might become a failure.
My objection was that it was all about WinAmp, including the suggestion of running it ! Which of course, you can't do on any of the platforms that this version of Helix / Real Player works on.
I didn't see how that was informative and on topic (originally it was +5 informative), which is why I posted.
Hmm, I wouldn't want to trust the security of my machine to an 100% automatic update procedure. What if the security patch will conflict with some software or configuration I'm using. Next I might see a BSOD, and not realise that it was the automaticaly applied SP that caused it. Or could it be failing hardware ? How would I know if it is 100% automatic ?
It doesn't take any time, whatsoever.
In the least, surely it takes time while the system reboots ?
With Linux, I'd have to open a console window and use apt or rpm or something like that every time a patch came out(after figuring out what to use, how it works, what to type in, etc.)
Before we go any further, have you ever used Linux ?
That, unfortunately, is the attitude of most hardware manufacturers today. There are thankfully a few (and growing number of) exceptions, but still not many.
I think you'll find it is the complete opposite. Linux wouldn't work on nearly as many computers as it does if hardware specs were predominantly closed.
If you have a look at the drivers sub-directory of the kernel source, every.c and.h file is a direct result of open programming specifications.
As of the 2.6.6 kernel, there are a total of 2323.c files within the drivers sub-directory. Each one doesn't necessarily correspond to a single device driver, on the other hand, sometimes one.c file will drive a type of device (eg. a network card) from multiple manufacturers, as they are using the same chipset. Averaging that out, yet being conservative, I'd suggest there are easily 1000 open source device drivers within the current Linux kernel.
It could be argued that all these drivers are for old hardware for which the "secret sauce" doesn't need to be that secret anymore. Well, I'd think Intel's 10Gbps Ethernet cards are pretty leading edge, yet they have contributed GPL'ed open source drivers to the Linux kernel - they are within the net/ixgb directory, under drivers.
Companies such as ATi and Nvidia are the exceptions when it comes to open hardware programming specs, not the rule.
The Helix / Real Player being talked about is for Linux / Solaris or Symbian OSes. So why compare it to WinAmp, when WinAmp doesn't run on any of these OSes, nor is the source code available ?
Cost Of Ownership is the real cost of an OS, not the initial purchase price.
Spending an extra couple of hours trying to get Linux to work as well as it should, for working people, instantly makes Linux more expensive than XP, when it comes to desktops.
Have you maybe forgotten all the hours it takes to patch MS OSes against the various worms that spring up, every couple of months or so ?
As a 100% exclusive Linux user, the only time I spend on them is when I read about them on Slashdot...
I don't want to get into a big debate about Linux vs Windows (I've made my mind up, I suspect you have also), however I don't think you can just ignore the security issues that MS OSes have when making statements about the price comparisons of Linux vs MS OSes.
The primary problem with hardware drivers for Linux is that hardware vendors (usually) aren't interested in making them.
That's close to the reason, but not the true reason. All the hardware manufacturers need to do is provide open hardware programming specifications. If the hardware is popular, the community will write and support open source drivers at no cost to the hardware manufacturer. This is what has already happened for the majority of device drivers already in Linux.
If the hardware isn't popular enough for the community to develop a driver for free, at least, with open programming specifications, a third party could pay a second party to develop a Linux device driver, preferrably an open source one.
Before binary modules, there were only open source device drivers, predominatly written from open hardware programming specifications. Binary modules are the new kid on the block, open source device drivers are the norm, going way back to 1991.
I've seen the IPR section of the VRRP RFC, although I'd never heard of anybody being sued. Admittedly I wouldn't know if license fees had been paid though from, for example, Juniper, who implement VRRP.
One story I'd heard was Cisco offered HSRP to the IETF, and were willing to grant royalty free access to the patents, as long as HSRP was adopted as the standard.
For some reason (very likely technical), VRRP was developed, Cisco's patents cover some methods it uses, so Cisco are listed in the IPR section of the RFC.
As I'd never heard of anybody being sued (and there is an open source implementation of VRRP, I don't think they've had any issues), I'd figured that Cisco hadn't bothered demanding royalties, basically because they were willing to forgo them in the first place if HSRP was used.
It is just an Internet-Draft (ID), that has been submitted for IETF approval. The IETF haven't reviewed it yet, nor taken a position on whether it should be a standard or not.
I could submit a ID for a protocol for standing on my head. That doesn't mean it is an IETF recommendation or that it will be.
With all the FUD being expressed by people who don't know much (anything?) about the IETF and its processes, maybe the next higher level after RTFA should be GAFC (Get A F**king Clue).
I've had a Compaq Deskpro convertable mini-tower for 3 1/2 years now, and it has none of these issues.
I've built enough PCs in my time, such that when I bought this one, I didn't want to go through the hassle of collecting the parts, etc. etc. What I wanted to do with it was far more important than the time and effort I could have spent building a similar machine.
It is a business / enterprise machine, so I did pay more for it, however I had a 3 year, on-site warranty, and I got business / enterprise quality.
It's been a pleasure to use, and on the occasions I've upgraded it (new video card, extra RAM, CD-RW, 2 x HDD) it has been simple. The most complex thing involved is unpluggin all the cables - which has everything to do with how many PCI cards I've got in it (I have no room left), and nothing to do with it being a Compaq.
I'm very happy with it, it has the build quality I would expect of a BMW or a Mercedes Benz.
It violates the fundamental tenet that everything should do one thing and do it well.
I don't necessarily think it does. The GPUs are very good at certain data processing operations. The only application these operations has recently been applied to has been 3D games or real time graphics. This just widens the scope of what the opimisation is being used for.
As an analogy, tractors were pretty much designed to pull a plough, which was the immediate problem being solved. However, the real need was to have a device that can "pull a lot of weight at a reasonably slow and controllable speed".
I've seen tractors being used to launch large surf lifesaving row boats at the beach. Obviously tractors weren't originally designed for that use, however, the real need ("pull a lot of weight at a reasonable slow and controllable speed") is exactly the same.
I don't think, in the case of a GPU or a tractor, using it to solve other problems with similar or the same needs is contrary to the compartmentalisation tenet.
I used to write programs the way you describe. In BASIC, around 1982. I was self taught, so I didn't know any better.
The problem with code in the form you suggest, from a programmer's point of view, is that once a program gets to a particular size - I'd suggest around 200 to 300 lines - you can't keep track of every detail or every variable, and what they do, are for, and what their current value us, or whether the current value is valid.
Modularity helps solve this problem, by allowing you to forget irrelevent details. In other words, it allows you to focus on what you should be focusing on, at that moment. As long as the other modules are well written, and debugged, you just mentally treat them like a black box.
Your concern about modularity impacting performance is misplaced, for a few reasons. Low performance in a program is usually a result of slow algorithms, not bad programming structure - the slight performance decrease of jumping to a module will be infinately smaller than the result of picking an inappropriate sorting routine.
Secondly, if calling routines does have a measurable performance impact, you can "inline" them. This means that a copy of the routine's assembly code is inserted in the position where the module call would have been. This may improve performance, although it comes at a cost - the size of the binary increases, as there are now multiple copies of the same routine in the binary. This impacts on memory though, which, on a low end system, may mean swapping to disk, which, of course, impacts performance so significantly, that any other optimisations performed were just a waste of time.
Everything comes at a cost. Modularity comes at the cost of time spent planning the modular architecture of the program, as well as the additional actual code to implement the modules. However, the benefits of modularity to program maintenance, data hiding and code reuse far exceed these costs.
If you're interested in further reading on this topic, I'd suggest getting a copy of "Code Complete", by Steve McConnell. His first edition has been available since 1993, however, I've heard a new version is coming some time this year, so it may be worth waiting for that.
Or how about when you go around to a new colleages house for a BBQ, to get to know them. Do you eat the free food there ?
Or go to a party where everybody has to bring food or drinks. Do you eat the free food there ? Would you be offended if other people don't eat the free food that you brought ? If they don't, aren't they saying that you are untrustworthy ?
Free doesn't mean you can't trust something.
You are overlooking social and reputational consequences of providing something at no cost that has intrinsic value. I know you know about this idea, as you posted your own example earlier. In your case, it was software you paid money for. You still threaten social and reputational consequences if the product fails, which for a commercial company has financial consquences. For people who provide software for free, social and reputational consequences are far more costly, as the only increase in value they get from providing the software for free are social and reputational.
In fact, this is one of the fundamental truths of The Cluetrain Manifesto. The Internet provides the ability for social and reputational consquences to travel much further and much faster, which increase the impact of those consequences.
I'm sure if BIND wasn't good enough, the readership of Slashdot would know about it pretty quickly. We already know when an exploitable bug is discovered, the day it is discovered. That is likely to be one of the major origins of negative comments about BIND in Slashdot forums. The Slashdot community is a large technical community, who usually are in positions to select one DNS server implementation over another.
If ISC care about their social position within the Internet community (I'm sure they do), and want to avoid reputational consquences when they can't be relied upon (I'm sure they do), they can either try to out market the negative messages, or try to do the right job. It is almost a sport for techos to spot marketroids, so I'm confident they will try to do the right job.
I sort of expected that accusation. Note that I didn't say they were a thief, just that the logic behind the decision not to provide some sort of contribution of value back is the same as the mindset of a thief. Another, much simpler word for it is "selfishness".
I don't have that much issue with individuals who take a copy of Linux, and give nothing back. I'd prefer them to do something like help out a friend with it, or contribute a helpful comment to slashdot, or report a bug. They then are recipricating.
It is corporations that take it, use it to make money, and yet feel no obligation to provide something, anything, in return. I agree, those non-contributing corporations have the right to use it. I, however, won't be rewarding them for doing so.
That's not quite true. Nvidia users benefit by being able to use their video hardware fully under Linux. That's not something to sneeze at. Most users don't care if the driver is closed-source, as long as it's still free.
And that is sad. Although technically, due to the GPL, they aren't stealing, they have the same mindset a thief does - take something of value, and provide nothing in return.
Nvidia and ATi (to a lesser extent, they have released some programming info for their older cards), have a similar mind set. They take advantage of the market that Linux has created, yet also don't provide anything in return to the community that created that market - the open source community broadly, and specifically the Linux / XFree86 (and X.Win) communities.
I'm no kernel hacker, yet I return the value I've received from open source software, such as Linux, by regularly providing help in the Linux/BSD forum at Whirlpool. I realise I don't have to, however that is how I "maintain" my membership of the open source community.
I don't really care too much for the issues ATi or NVidia users have with the corresponding binary drivers. Those problems can always be traced back to the closed nature of the proprietory drivers. I prefer to spend my money with vendors of hardware who provide open programming specificiations. Even better if they sponsor open source driver development. I will reward them financially for understanding, and participating reciprically in the open source community.
I'm sure you can guess, NVidia will never see a dollar of mine. ATi maybe, although it will be for "old", "cheap" hardware (ie. the 9200 series, for which they openned the programming specs) that they make very little profit on.
Slashdot Mirror
On face value, there appears to be nothing wrong with increased police powers, for example, the ability to detain somebody for significant periods of time if they are suspected of something, without allowing the detainee to contact their lawyer or make a phone call to the outside world. Law enforcement officials would only detain bad guys, right ?
The problem with this is that it is based on the assumption that the everybody within the law enforcement organisations involved are totally and 100% honest. Of course, this isn't the case.
Judicial oversight of things such as wire taps is there to try to ensure that these mechanisms aren't abused by corrupt, dishonest or overzealous law enforcement officials.
Sadly, it seems that, since 911, George W. Bush's adgenda is to minimise or remove Judicial oversight in the name of "security". I can only suggest that he believes that law enforment officials are 100% honest.
Why is there an assumption that Linux won't be successful if high FPS games aren't available for it ?
I don't care to play games under Linux (as I don't play games in general, they just don't hold my attention for some reason (probably biological)), so I don't care whether commercial games are supported. To me, Linux is already successful, and has been since I first started using it in 1993, as it has done everything I've wanted it to do.
"Success" is a relative word, and is used to measure something against a set of criteria. Change any of those criteria, and the former "success" might become a failure.
I'm pretty sure source is available on the download page.
My objection was that it was all about WinAmp, including the suggestion of running it ! Which of course, you can't do on any of the platforms that this version of Helix / Real Player works on.
I didn't see how that was informative and on topic (originally it was +5 informative), which is why I posted.
Windows Update is 100% automatic.
Hmm, I wouldn't want to trust the security of my machine to an 100% automatic update procedure. What if the security patch will conflict with some software or configuration I'm using. Next I might see a BSOD, and not realise that it was the automaticaly applied SP that caused it. Or could it be failing hardware ? How would I know if it is 100% automatic ?
It doesn't take any time, whatsoever.
In the least, surely it takes time while the system reboots ?
With Linux, I'd have to open a console window and use apt or rpm or something like that every time a patch came out(after figuring out what to use, how it works, what to type in, etc.)
Before we go any further, have you ever used Linux ?
I think you'll find it is the complete opposite. Linux wouldn't work on nearly as many computers as it does if hardware specs were predominantly closed.
If you have a look at the drivers sub-directory of the kernel source, every .c and .h file is a direct result of open programming specifications.
As of the 2.6.6 kernel, there are a total of 2323 .c files within the drivers sub-directory. Each one doesn't necessarily correspond to a single device driver, on the other hand, sometimes one .c file will drive a type of device (eg. a network card) from multiple manufacturers, as they are using the same chipset. Averaging that out, yet being conservative, I'd suggest there are easily 1000 open source device drivers within the current Linux kernel.
It could be argued that all these drivers are for old hardware for which the "secret sauce" doesn't need to be that secret anymore. Well, I'd think Intel's 10Gbps Ethernet cards are pretty leading edge, yet they have contributed GPL'ed open source drivers to the Linux kernel - they are within the net/ixgb directory, under drivers.
Companies such as ATi and Nvidia are the exceptions when it comes to open hardware programming specs, not the rule.
The Helix / Real Player being talked about is for Linux / Solaris or Symbian OSes. So why compare it to WinAmp, when WinAmp doesn't run on any of these OSes, nor is the source code available ?
Cost Of Ownership is the real cost of an OS, not the initial purchase price.
Spending an extra couple of hours trying to get Linux to work as well as it should, for working people, instantly makes Linux more expensive than XP, when it comes to desktops.Have you maybe forgotten all the hours it takes to patch MS OSes against the various worms that spring up, every couple of months or so ?
As a 100% exclusive Linux user, the only time I spend on them is when I read about them on Slashdot ...
I don't want to get into a big debate about Linux vs Windows (I've made my mind up, I suspect you have also), however I don't think you can just ignore the security issues that MS OSes have when making statements about the price comparisons of Linux vs MS OSes.
The primary problem with hardware drivers for Linux is that hardware vendors (usually) aren't interested in making them.
That's close to the reason, but not the true reason. All the hardware manufacturers need to do is provide open hardware programming specifications. If the hardware is popular, the community will write and support open source drivers at no cost to the hardware manufacturer. This is what has already happened for the majority of device drivers already in Linux.
If the hardware isn't popular enough for the community to develop a driver for free, at least, with open programming specifications, a third party could pay a second party to develop a Linux device driver, preferrably an open source one.
Before binary modules, there were only open source device drivers, predominatly written from open hardware programming specifications. Binary modules are the new kid on the block, open source device drivers are the norm, going way back to 1991.
If I lost 30% of my total body fat in a month, none of my clothes would fit !
The above post, while the truth, is basically a summary of the first few chapters of Bruce Schneier's book, "Beyond Fear".
I've seen the IPR section of the VRRP RFC, although I'd never heard of anybody being sued. Admittedly I wouldn't know if license fees had been paid though from, for example, Juniper, who implement VRRP.
One story I'd heard was Cisco offered HSRP to the IETF, and were willing to grant royalty free access to the patents, as long as HSRP was adopted as the standard.
For some reason (very likely technical), VRRP was developed, Cisco's patents cover some methods it uses, so Cisco are listed in the IPR section of the RFC.
As I'd never heard of anybody being sued (and there is an open source implementation of VRRP, I don't think they've had any issues), I'd figured that Cisco hadn't bothered demanding royalties, basically because they were willing to forgo them in the first place if HSRP was used.
It is just an Internet-Draft (ID), that has been submitted for IETF approval. The IETF haven't reviewed it yet, nor taken a position on whether it should be a standard or not.
I could submit a ID for a protocol for standing on my head. That doesn't mean it is an IETF recommendation or that it will be.
With all the FUD being expressed by people who don't know much (anything?) about the IETF and its processes, maybe the next higher level after RTFA should be GAFC (Get A F**king Clue).
I've had a Compaq Deskpro convertable mini-tower for 3 1/2 years now, and it has none of these issues.
I've built enough PCs in my time, such that when I bought this one, I didn't want to go through the hassle of collecting the parts, etc. etc. What I wanted to do with it was far more important than the time and effort I could have spent building a similar machine.
It is a business / enterprise machine, so I did pay more for it, however I had a 3 year, on-site warranty, and I got business / enterprise quality.
It's been a pleasure to use, and on the occasions I've upgraded it (new video card, extra RAM, CD-RW, 2 x HDD) it has been simple. The most complex thing involved is unpluggin all the cables - which has everything to do with how many PCI cards I've got in it (I have no room left), and nothing to do with it being a Compaq.
I'm very happy with it, it has the build quality I would expect of a BMW or a Mercedes Benz.
It violates the fundamental tenet that everything should do one thing and do it well.
I don't necessarily think it does. The GPUs are very good at certain data processing operations. The only application these operations has recently been applied to has been 3D games or real time graphics. This just widens the scope of what the opimisation is being used for.
As an analogy, tractors were pretty much designed to pull a plough, which was the immediate problem being solved. However, the real need was to have a device that can "pull a lot of weight at a reasonably slow and controllable speed".
I've seen tractors being used to launch large surf lifesaving row boats at the beach. Obviously tractors weren't originally designed for that use, however, the real need ("pull a lot of weight at a reasonable slow and controllable speed") is exactly the same.
I don't think, in the case of a GPU or a tractor, using it to solve other problems with similar or the same needs is contrary to the compartmentalisation tenet.
These applications are not likely to generate or process data at such a rate that the slow AGP read speed will matter that much, if at all.
I used to write programs the way you describe. In BASIC, around 1982. I was self taught, so I didn't know any better.
The problem with code in the form you suggest, from a programmer's point of view, is that once a program gets to a particular size - I'd suggest around 200 to 300 lines - you can't keep track of every detail or every variable, and what they do, are for, and what their current value us, or whether the current value is valid.
Modularity helps solve this problem, by allowing you to forget irrelevent details. In other words, it allows you to focus on what you should be focusing on, at that moment. As long as the other modules are well written, and debugged, you just mentally treat them like a black box.
Your concern about modularity impacting performance is misplaced, for a few reasons. Low performance in a program is usually a result of slow algorithms, not bad programming structure - the slight performance decrease of jumping to a module will be infinately smaller than the result of picking an inappropriate sorting routine.
Secondly, if calling routines does have a measurable performance impact, you can "inline" them. This means that a copy of the routine's assembly code is inserted in the position where the module call would have been. This may improve performance, although it comes at a cost - the size of the binary increases, as there are now multiple copies of the same routine in the binary. This impacts on memory though, which, on a low end system, may mean swapping to disk, which, of course, impacts performance so significantly, that any other optimisations performed were just a waste of time.
Everything comes at a cost. Modularity comes at the cost of time spent planning the modular architecture of the program, as well as the additional actual code to implement the modules. However, the benefits of modularity to program maintenance, data hiding and code reuse far exceed these costs.
If you're interested in further reading on this topic, I'd suggest getting a copy of "Code Complete", by Steve McConnell. His first edition has been available since 1993, however, I've heard a new version is coming some time this year, so it may be worth waiting for that.
as W.E.P stands for "Wired Equivalence Protocol".
It was never designed to be any more secure than copper, which we all know is secure against packet sniffers (NOT!).
IPsec is the best and most general way to secure wireless networks.
Cisco Goes for an Open-Source Spin
Shouldn't you go back under the bridge ?
So, did you "post" that raid-5 + raidtools + kernel 2.6 locks up a computer, to save somebody else going through what you had to ?
which is why I don't use BIND anyway.
Just pre-empting anybody who suggests I'm a BIND undercover agent.
Or how about when you go around to a new colleages house for a BBQ, to get to know them. Do you eat the free food there ?
Or go to a party where everybody has to bring food or drinks. Do you eat the free food there ? Would you be offended if other people don't eat the free food that you brought ? If they don't, aren't they saying that you are untrustworthy ?
Free doesn't mean you can't trust something.
You are overlooking social and reputational consequences of providing something at no cost that has intrinsic value. I know you know about this idea, as you posted your own example earlier. In your case, it was software you paid money for. You still threaten social and reputational consequences if the product fails, which for a commercial company has financial consquences. For people who provide software for free, social and reputational consequences are far more costly, as the only increase in value they get from providing the software for free are social and reputational.
In fact, this is one of the fundamental truths of The Cluetrain Manifesto. The Internet provides the ability for social and reputational consquences to travel much further and much faster, which increase the impact of those consequences.
I'm sure if BIND wasn't good enough, the readership of Slashdot would know about it pretty quickly. We already know when an exploitable bug is discovered, the day it is discovered. That is likely to be one of the major origins of negative comments about BIND in Slashdot forums. The Slashdot community is a large technical community, who usually are in positions to select one DNS server implementation over another.
If ISC care about their social position within the Internet community (I'm sure they do), and want to avoid reputational consquences when they can't be relied upon (I'm sure they do), they can either try to out market the negative messages, or try to do the right job. It is almost a sport for techos to spot marketroids, so I'm confident they will try to do the right job.
This doesn't make them a thief.
I sort of expected that accusation. Note that I didn't say they were a thief, just that the logic behind the decision not to provide some sort of contribution of value back is the same as the mindset of a thief. Another, much simpler word for it is "selfishness".
I don't have that much issue with individuals who take a copy of Linux, and give nothing back. I'd prefer them to do something like help out a friend with it, or contribute a helpful comment to slashdot, or report a bug. They then are recipricating.
It is corporations that take it, use it to make money, and yet feel no obligation to provide something, anything, in return. I agree, those non-contributing corporations have the right to use it. I, however, won't be rewarding them for doing so.
That's not quite true. Nvidia users benefit by being able to use their video hardware fully under Linux. That's not something to sneeze at. Most users don't care if the driver is closed-source, as long as it's still free.
And that is sad. Although technically, due to the GPL, they aren't stealing, they have the same mindset a thief does - take something of value, and provide nothing in return.
Nvidia and ATi (to a lesser extent, they have released some programming info for their older cards), have a similar mind set. They take advantage of the market that Linux has created, yet also don't provide anything in return to the community that created that market - the open source community broadly, and specifically the Linux / XFree86 (and X.Win) communities.
I'm no kernel hacker, yet I return the value I've received from open source software, such as Linux, by regularly providing help in the Linux/BSD forum at Whirlpool. I realise I don't have to, however that is how I "maintain" my membership of the open source community.
I don't really care too much for the issues ATi or NVidia users have with the corresponding binary drivers. Those problems can always be traced back to the closed nature of the proprietory drivers. I prefer to spend my money with vendors of hardware who provide open programming specificiations. Even better if they sponsor open source driver development. I will reward them financially for understanding, and participating reciprically in the open source community.
I'm sure you can guess, NVidia will never see a dollar of mine. ATi maybe, although it will be for "old", "cheap" hardware (ie. the 9200 series, for which they openned the programming specs) that they make very little profit on.