This debate would be over in a day if the other networks simply unplugged the first network to try it.
The concept totally defeats the idea of the Internet and peering. The ultimate outcome of this is to remove all peering points and have Google/etc negotiate with each net individually.The last mile networks are simply holding their customers out for ransom to the highest bidder.
When are they going to stop selling analog only TVs? The stores are still full of them. Every one of these that gets sold only makes the problem worse.
A lot of the people working on the various OpenDoc compatible word processors live in MA. None of the Microsoft employees working on Office are located in MA. Now when Microsoft makes a choice to do a large amount of new hiring, you see that they choose to do it in India instead of MA.
So which should MA do, send $200M in license fees to Microsoft so that the monopoly can send it to India? Or make an open, competitive playing field that is possible for local firms to participate in? Note that I didn't say give the business to the local firm, but how does a local software firm competete when the contract only allows "Microsoft Office".
I wish Nielson could figure out that pre-announcing the date of US sweeps week ruins TV viewing. For one week out of the year every channel simultaneously has great programming. Then for the other fifty one weeks it sucks. Advertisers are stupid for believing that sweeps week is at all representative of viewing patterns. Imagine how advertising rates would have been set if 9/11 had happened during sweeps week - cable news would have had top share.
Audio support for multiple users is normally handled with USB headphone. The core OS and ALSA already have excellent support for as many as you want to plug in. Each set of headphone appears as an independent ALSA device. It is the higher level support that needs work.
In addition to the above philosophical disagreement, I also think your idea has a practical flaw. Simply, it's impossible to please everyone. Therefore, even if Linux gained a champion on the order of Bill Gates or Steve Jobs, someone with marketing and business sense in addition to coding skills, there would always be a minority disliking the decisions that this leader makes. Given Linux's open-source nature, there would be little disincentive for these disaffected minorities to fork their own distro's of Linux. With Linux already being as scattered as it is, I hardly think that this would be the answer. Combine this with the practical difficulty of getting all existing distros under a single banner (they can't even agree on a packaging system, much less an OS structure), and your vision, while being a nice thought experiment, becomes nearly impossible to implement in the real world.
The model of unguided development will usually work in the open source world. The exception is when the community is confronted by an extremely large problem. In that case we may regress into lots of uncoordinated attempts at fixing the large problem but none that have enough resource to succeed. In these cases we need to pull together to get over the hump, then we can go back to unguided development.
Go dig around in the Cairo list and look for benchmarks. Cairo on glitz (3D hardware) benchmarks at 100:1 faster than xlib (2D hardware). That is part of what prompted the development of EXA. But EXA is not true 2D, it is using the 3D hardware for it's implementation even though it only exposes a 2D API.
You're just measuring the difference between 1X,2X,4X and 8X AGP. SDRAM is way faster than the bus interfaces. The cards are faster but not orders of magnitudes faster like the GPU is. The real performance gains are made by moving the drawing onto the GPU coprocessor and avoiding the system bus whenever possible.
I've never booted BSD and I don't claim any real knowledge of it. The article just reflected what I have read and seen posted to the xorg mailing lists.
If BSD has parallel support to Linux fbdev that will make it easier to run OpenGL/EGL on it. I was also under the impression that all of that PCI probing code was in X because BSD lacked the needed functions.
The focus of the article was the state of Linux graphics but write up some BSD clarifications and I'll add them as soon as we can convince our censorist admin to unlock my account.
If you compare the 2D performance of an ATI Rage1128, Radeon 9000, and a Radeon X850 you will discover that they all perform about the same. But if you compare 3D performance of the X850 to the R128 you will see a 500:1 improvement.
I didn't brush the open driver issue off, I simply chose not to address a topic that is the source of a lot of controversy. I am well aware of the problems of obtaining driver documentation.
X just needs to make a choice, continue with the flat-lined 2D performance or make the jump to the 3D hardware. If X chooses 3D I would much rather see if use a well designed, standardized API like OpenGL than to slowly extend the existing code base to start using 3D features like EXA does.
If you want open 3D drivers go lobby the hardware vendors to release code and specs. However, I think it is wrong for Linux to ignore the immense performance gains available from the 3D hardware on the grounds that the hardware is not completely open. Withholding use of 3D hardware on Linux will do nothing to open the vendors and it will definitely result in Linux having an inferior competitive desktop experience.
Have you considered that the opposite effect might happen? It Linux builds an excellent 3D desktop and attracts a lot of new users the hardware vendors may start to take Linux seriously and open their specs.
This sounds very similar to the 'smart card' concept back in the late 80's and early 90's. Intel had the 586-driven
I was product manager for the Intel cards. Back then we had DOS and the 640K RAM limit, coprocessor cards provided a way to move all of networking off into it's own RAM space. That was more important than the preformance gains.
TCP checksum offload works and RDMA is a big win. Offloaded iSCSI will probably be a win too. Anything else is marginal.
Spend your money on a faster CPU, more main memory or 10Gb Ethernet.
It's because of fear of lawsuits, not a desire to hide their chip interfaces. There are just as many bad hardware patents as bad software ones. By controlling spec release with an NDA they stop the law firms from searching for violations of these bogus patents.
The damage from keeping hardware specs out of open source hands is tiny compared to getting a $400M judgement for patent violation.
Trade secret laws prevent law firms from forcing spec releases without pretty good proof of a patent violation.
Could it be that Apple's delay until next year is because it will require Intel VT on the processor to run? Requiring VT would probably lock out all existing machines. It would also be difficult to emulate on existing chips. VT specific shuts out AMD too. Any rumors of Apple working on a hypervisor?
The simple reason for requiring VT is to get MS Windows support. Windows would run in another VM. A virtual graphics card would then make it appear inside the Mac display.
Running an SMB server interally allows partitions to be shared.
Daemons in each OS could export the clipboard over a virtual network to make DND work. And do things like mouse handoff.
Seems like anyone with an existing IP satellites could get into the satellite radio business. Is it possible to receive these satellites while moving or do they need a big dish? The radio stream could be sent out on UPD to remove the need for two way communication and make it into a broadcast.
This design may be more valuable as a core that could be added into all-in-one chips. I don't think it has a prayer of economically making it as an add-in video card. But there are lot's of Linux based embedded devices that could use this when integrated into the same chip as the CPU.
The CPiA webcam driver supports the QX3 on Linux. CPiA webcam driver for Linux Just turn on CONFIG_VIDEO_CPIA=m when you build your kernel. You can even use/proc to turn the microscope lights on and off.
CPiA is not made any more. Maybe the QX5 uses a similar webcam chip.
You have to take into account sales tax and free shipping. Home does not charge sales tax. Small business always does. All groups run various free shipping deals. Then there are coupons. Dell offers thousands of coupons and they track which are active. The price may be higher because they know that you are likely to use a coupon on it. The coupons on only valid for the division that offered them. Finally there are corporate contracts which guarantee a percentage discount from list. The prices aren't really that different when you take all of this into account. Often small business will be cheaper then home. But when you add the sales tax home is then cheaper than small business. So the answer depends on where you live.
I find it hard to beat phone cards. www.uniontelecard.com has good selection. Pay a little more and get the ones from IDT. Their connections are much better than Entel. IDT phone cards are 1/10th the price of Verizon for calling South Africa and 1/3 Vonage direct dial. You can use phone cards from your cell phone and VOIP phones too.
You're taking the wrong approach to improving the API. The standard says that API(x) has to behave some broken way. So that you don't break the existing apps you have to implement API(x) with the broken behavior.
But nothing is stopping you from implementing API2(x) with the correct behavior. If API2(x) is a good enough solution it will probably be incorporated into the standard API the next time around. API/API2 lets both older and newer apps run without breaking each other.
I have to agree that Verizon is making the phones useless. Next phone I get is going to be voice only. I wasted a bunch of money buying a couple of Motorola 722's to then find out that I had to pay more to use every feature.
I wanted to sync my calendar/address book with Yahoo. I called Verizon, no program exists for syncing Yahoo to a Motorola 722. So I said fine I'll write one and open source it. When I mentioned this on the Qualcomm forms I almost got lynched. People complained about commies like me ruining their ability to make money and support their families. I said that I was willing to buy the program but none exists -- no one offered to write it.
Qualcom is completely against free distibution of apps for Brew. They told me that I could write it and distribute for free on the Verizon net if I was willing to pay the fees for all of the users. I also had to pay a $4000 up front fee. Turns out that they require additional app royalty fees to use the OS I just bought from them.
The offical reason for this "fee" is that Qualcomm will audit my apps to ensure that they don't contain a virus that would call 911. I tried to point out to them that a virus on home computers with modems is just as dangerous, but they wouldn't listen.
None of this is Motorola's fault. It is all Qualcomm and Verizon.
When my contract is up I'm getting a new network and Linux based phones.
Slashdot Mirror
This debate would be over in a day if the other networks simply unplugged the first network to try it.
The concept totally defeats the idea of the Internet and peering. The ultimate outcome of this is to remove all peering points and have Google/etc negotiate with each net individually.The last mile networks are simply holding their customers out for ransom to the highest bidder.
When are they going to stop selling analog only TVs? The stores are still full of them. Every one of these that gets sold only makes the problem worse.
A lot of the people working on the various OpenDoc compatible word processors live in MA. None of the Microsoft employees working on Office are located in MA. Now when Microsoft makes a choice to do a large amount of new hiring, you see that they choose to do it in India instead of MA.
So which should MA do, send $200M in license fees to Microsoft so that the monopoly can send it to India? Or make an open, competitive playing field that is possible for local firms to participate in? Note that I didn't say give the business to the local firm, but how does a local software firm competete when the contract only allows "Microsoft Office".
If that headline doesn't send MA a message on switching to OpenDoc nothing will.
I wish Nielson could figure out that pre-announcing the date of US sweeps week ruins TV viewing. For one week out of the year every channel simultaneously has great programming. Then for the other fifty one weeks it sucks. Advertisers are stupid for believing that sweeps week is at all representative of viewing patterns. Imagine how advertising rates would have been set if 9/11 had happened during sweeps week - cable news would have had top share.
Audio support for multiple users is normally handled with USB headphone. The core OS and ALSA already have excellent support for as many as you want to plug in. Each set of headphone appears as an independent ALSA device. It is the higher level support that needs work.
In addition to the above philosophical disagreement, I also think your idea has a practical flaw. Simply, it's impossible to please everyone. Therefore, even if Linux gained a champion on the order of Bill Gates or Steve Jobs, someone with marketing and business sense in addition to coding skills, there would always be a minority disliking the decisions that this leader makes. Given Linux's open-source nature, there would be little disincentive for these disaffected minorities to fork their own distro's of Linux. With Linux already being as scattered as it is, I hardly think that this would be the answer. Combine this with the practical difficulty of getting all existing distros under a single banner (they can't even agree on a packaging system, much less an OS structure), and your vision, while being a nice thought experiment, becomes nearly impossible to implement in the real world.
The model of unguided development will usually work in the open source world. The exception is when the community is confronted by an extremely large problem. In that case we may regress into lots of uncoordinated attempts at fixing the large problem but none that have enough resource to succeed. In these cases we need to pull together to get over the hump, then we can go back to unguided development.
Cairo is a library that runs on top of X, it doesn't replace it.
Go dig around in the Cairo list and look for benchmarks. Cairo on glitz (3D hardware) benchmarks at 100:1 faster than xlib (2D hardware). That is part of what prompted the development of EXA. But EXA is not true 2D, it is using the 3D hardware for it's implementation even though it only exposes a 2D API.
You're just measuring the difference between 1X,2X,4X and 8X AGP. SDRAM is way faster than the bus interfaces. The cards are faster but not orders of magnitudes faster like the GPU is. The real performance gains are made by moving the drawing onto the GPU coprocessor and avoiding the system bus whenever possible.
I've never booted BSD and I don't claim any real knowledge of it. The article just reflected what I have read and seen posted to the xorg mailing lists.
If BSD has parallel support to Linux fbdev that will make it easier to run OpenGL/EGL on it. I was also under the impression that all of that PCI probing code was in X because BSD lacked the needed functions.
The focus of the article was the state of Linux graphics but write up some BSD clarifications and I'll add them as soon as we can convince our censorist admin to unlock my account.
If you compare the 2D performance of an ATI Rage1128, Radeon 9000, and a Radeon X850 you will discover that they all perform about the same. But if you compare 3D performance of the X850 to the R128 you will see a 500:1 improvement.
I didn't brush the open driver issue off, I simply chose not to address a topic that is the source of a lot of controversy. I am well aware of the problems of obtaining driver documentation.
X just needs to make a choice, continue with the flat-lined 2D performance or make the jump to the 3D hardware. If X chooses 3D I would much rather see if use a well designed, standardized API like OpenGL than to slowly extend the existing code base to start using 3D features like EXA does.
If you want open 3D drivers go lobby the hardware vendors to release code and specs. However, I think it is wrong for Linux to ignore the immense performance gains available from the 3D hardware on the grounds that the hardware is not completely open. Withholding use of 3D hardware on Linux will do nothing to open the vendors and it will definitely result in Linux having an inferior competitive desktop experience.
Have you considered that the opposite effect might happen? It Linux builds an excellent 3D desktop and attracts a lot of new users the hardware vendors may start to take Linux seriously and open their specs.
This sounds very similar to the 'smart card' concept back in the late 80's and early 90's. Intel had the 586-driven
I was product manager for the Intel cards. Back then we had DOS and the 640K RAM limit, coprocessor cards provided a way to move all of networking off into it's own RAM space. That was more important than the preformance gains.
TCP checksum offload works and RDMA is a big win. Offloaded iSCSI will probably be a win too. Anything else is marginal.
Spend your money on a faster CPU, more main memory or 10Gb Ethernet.
It's because of fear of lawsuits, not a desire to hide their chip interfaces. There are just as many bad hardware patents as bad software ones. By controlling spec release with an NDA they stop the law firms from searching for violations of these bogus patents.
The damage from keeping hardware specs out of open source hands is tiny compared to getting a $400M judgement for patent violation.
Trade secret laws prevent law firms from forcing spec releases without pretty good proof of a patent violation.
Could it be that Apple's delay until next year is because it will require Intel VT on the processor to run? Requiring VT would probably lock out all existing machines. It would also be difficult to emulate on existing chips. VT specific shuts out AMD too. Any rumors of Apple working on a hypervisor?
The simple reason for requiring VT is to get MS Windows support. Windows would run in another VM. A virtual graphics card would then make it appear inside the Mac display.
Running an SMB server interally allows partitions to be shared.
Daemons in each OS could export the clipboard over a virtual network to make DND work. And do things like mouse handoff.
Seems like anyone with an existing IP satellites could get into the satellite radio business. Is it possible to receive these satellites while moving or do they need a big dish? The radio stream could be sent out on UPD to remove the need for two way communication and make it into a broadcast.
This design may be more valuable as a core that could be added into all-in-one chips. I don't think it has a prayer of economically making it as an add-in video card. But there are lot's of Linux based embedded devices that could use this when integrated into the same chip as the CPU.
The CPiA webcam driver supports the QX3 on Linux. /proc to turn the microscope lights on and off.
CPiA webcam driver for Linux
Just turn on CONFIG_VIDEO_CPIA=m when you build your kernel. You can even use
CPiA is not made any more. Maybe the QX5 uses a similar webcam chip.
Old slashdot story about the QX3
Nvidia was at the Xdev conference and said they would support XGL as soon as we work out the exact interfaces at the lower levels.
ATI has not made any comments so far. ATI is definitely aware of the XGL project.
XGL needs to get a full reference implementation in place first before ATI/Nvidia can really start working on their versions.
You have to take into account sales tax and free shipping. Home does not charge sales tax. Small business always does. All groups run various free shipping deals. Then there are coupons. Dell offers thousands of coupons and they track which are active. The price may be higher because they know that you are likely to use a coupon on it. The coupons on only valid for the division that offered them. Finally there are corporate contracts which guarantee a percentage discount from list. The prices aren't really that different when you take all of this into account. Often small business will be cheaper then home. But when you add the sales tax home is then cheaper than small business. So the answer depends on where you live.
I find it hard to beat phone cards. www.uniontelecard.com has good selection. Pay a little more and get the ones from IDT. Their connections are much better than Entel. IDT phone cards are 1/10th the price of Verizon for calling South Africa and 1/3 Vonage direct dial. You can use phone cards from your cell phone and VOIP phones too.
You're taking the wrong approach to improving the API. The standard says that API(x) has to behave some broken way. So that you don't break the existing apps you have to implement API(x) with the broken behavior.
But nothing is stopping you from implementing API2(x) with the correct behavior. If API2(x) is a good enough solution it will probably be incorporated into the standard API the next time around. API/API2 lets both older and newer apps run without breaking each other.
Get a Blaster!
r ica/232777.stm
A blaster, flame-thrower operated by a foot pedal inside the car, blasts a jet of fire at a would-be hijacker.
A person confronted by an armed hijacker simply presses a pedal and the "blaster" ignites gas that shoots from the under-side of the car.
Doctors say the device is lethal - but the police have confirmed it is perfectly legal.
Photos....
http://news.bbc.co.uk/1/hi/world/af
Does anyone have a pointer to more detailed technical specs on this? Like the reference manual for writing a compatible driver?
I have to agree that Verizon is making the phones useless. Next phone I get is going to be voice only. I wasted a bunch of money buying a couple of Motorola 722's to then find out that I had to pay more to use every feature.
I wanted to sync my calendar/address book with Yahoo. I called Verizon, no program exists for syncing Yahoo to a Motorola 722. So I said fine I'll write one and open source it. When I mentioned this on the Qualcomm forms I almost got lynched. People complained about commies like me ruining their ability to make money and support their families. I said that I was willing to buy the program but none exists -- no one offered to write it.
Qualcom is completely against free distibution of apps for Brew. They told me that I could write it and distribute for free on the Verizon net if I was willing to pay the fees for all of the users. I also had to pay a $4000 up front fee. Turns out that they require additional app royalty fees to use the OS I just bought from them.
The offical reason for this "fee" is that Qualcomm will audit my apps to ensure that they don't contain a virus that would call 911. I tried to point out to them that a virus on home computers with modems is just as dangerous, but they wouldn't listen.
None of this is Motorola's fault. It is all Qualcomm and Verizon.
When my contract is up I'm getting a new network and Linux based phones.