AMD's technology preview day (Anandtech's report is at http://www.anandtech.com/cpuchipsets/showdoc.aspx? i=2768) said a lot that made me wonder about the future competitivenes of AMD. I'm neutral but the kind of stuff they were talking about made me doubt my conviction that Intel are guaranteed to pwn, but I am certain that the consumer is the winner in the upcoming battle for the best desktop CPU chips (and that's already been noted in the drop in prices between the two http://www.tomshardware.co.uk/2006/06/02/drop_in_a verage_processor_pricing/).
The notion of being able to put extra specialised hardware next to the CPU radically alters the way that PC's are going to be laid out and is a good lateral step along the road toward multi-threaded, multi-cored and multiple processing elements in computing. The K8L architecture and beyond are seeking to increase IPC throughput for AMD's chips as Intel has done for the Core architecture. I think that this throughput increase is the greatest asset to the Core's performance both for performance per cycle (==IPC) and performance per watt.
I'd contend that 'casual' games* are a snack food rather than a furniture as per your analogy. And larger games can be likened to banquets as music, video and gaming are all entertainment.
*: Casual games are expected to be the big growth area of the games market in the next couple of years -- just look at your smart phone and Nintendo's Wii as places for non-'gamer' games-playing to grow the market.
I dist-upgraded this morning. It boots quickly, makes good use of proprietary video drivers, is well-organised, functional and efficient. I would say that it is comparable to Win2k or WinXP from the month or so I've been using a release candidate. I hope that the fall's Edgy Eft release can be comparable to the looks and usability of these Windows Vista previews I keep seeing.
VFAT is the FAT filesystem driver and a standard part of the GNU tools that most Linuces come with -- FC3 has the support to read and write my FAT32 drives, so I'm sure it'll be there in FC5.
I'm not sure if Halal certification insists on anything more than the way the animal is killed (a neck slash and blood drained from the corpse). I don't think it would insist on non-battery or free-range chickens, for example, but may insist on cattle not fed foodstuffs derived from other animal material.
Is it not possible to have the TPM and BIOS/EFI contain keys to encrypted data in the partition table and boot sector? Or to move the location of these data according to unique keys in the system?
AM2 uses DDR2 memory. It is expected to become cheaper and more available than DDR when most companies move their production to make it. I don't want to fob you off with a cheap RTFM, but there is no easy alternative to reading what Tom's Hardware / AnandTech / Ars Technica have to say about the matter.
I think that Debian/Ubuntu doesn't scan its hardware at each boot to se if anything's changed, and doing that will enable it to not run drivers incompatible with your new MB. Win2K and WinXP updated their HAL (as you noticed) but Ubuntu is assuming it's working on the same hardware as before your upgrade.
I've looked, because I'd like to be able to say "run discover and it'll recognise your new setup", but there's little that helps. I believe Debian and Ubuntu use a program called Discover to detect hardware. The few links there are out there indicate that people seem to think that the software will work regardless after moving hardware. However, it may be that you're using software built for an architecture which isn't supported after your motherboard move, and I would advise either a reinstall of Ubuntu or soft reinstall of the packages you have on your system via dpkg --configure -a.
The Regulation of Investigatory Powers Act (2000) was set to make all legitimate uses of encrytion have a third party store a backup of the key in escrow. That part of the law, to my knowledge, got scrubbed before it hit the statute books.
I'm not convinced that they understand that this is 'free' software which they can take and make their own editions of. It may be that Oracle want a brand to be well-known to businesses, but I think that Oracle Linux running Oracle Desktop Linux would be better for market awareness of Oracle than Oracle supplying Novell Suse Desktop Linux.
My tentative suggestion -- given that I have no Apple computers and no experience with BootCamp -- would be to take the bootloader from ReactOS for its Win2k-like behaviour and stick it and grldr from the Grub4DOS in a primary-4 (as in fourth primary partition as the MBR looks at it) partition a couple of megabytes in size, then you have primary-2 and primary-3 and their extended partitions to use with your GNU/Linux installation. The Grub4DOS may need its Grub.exe and its stage-specific files updated to cope with EFI booting, and this may be the stumbing-block of the enterprise. Good luck!
Both stress and pressure are measured in force/area, so can be used interchangeably. (Right now I'm writing home: Mom! Stuff I specialised in Grad School is useful after all!)
I think that recognising the pace of computer alogrithm development and cutting the duration of a software patent to, say, three years, with the requirement that, within that time, a product is on the market containing the patent-covered method sounds like a good way to allow people to benefit from patents which drive forward progress without constricting others' ability to compete.
You forgot to mention that the candidates were also frozen for bugfixing. Towns has only lost two fingers to frostbite; the debian-privates e-mail list suggests that another candidate lost something more personal and delicate.
Negroponte says "We're heading to the point where 50 percent of the world will have a cell phone or some kind of (communication device) within 18 months. It's too voice-centric, and I could campaign to make it more data-centric, but that's going to happen, too."
That's seems to me to be unaware of the presence of GSM. 90% of the world's cell phones are GSM, an inherently digital-data format. Sure, it allows 9K6 data transfers or 28K8 if you use High Speed Circuit-Switched Data (HSCSD) mode, which I doubt will be available in too any developing nations. People aren't going to be downloading too much multimedia to the limited storage of these systems, so I don't see a low data rate as a problem.
Thanks fr allowing me to read your speach. I thought it was good cohesive and clear stuff. BTW: Are your references (presently displayed as [http://link/file%5D) intended to be active hyperlinks? And have you edited the Democrat/Republican confusion mentioned elsewhere in here?
I think it's in their interests to make cheaper notebooks that are licensed to run OS X. Market segmentation is a normal part of the motor industry (in Europe, you can buy the VW Golf, Seat Ibiza or Skoda Fabia from the same chassis, engine range and trim options -- the only differences being price, and what's a Celeron if not budget-branded failed Pentium?), so why not in computers?.
Unfortunately, now's not a good time, as all the new tech in their notebooks is going to be expensive and need the payoff against RnD cost. Older tech would be better, providing an additional revenue from existing hardware research. However, Steve Jobs and Jon Ive are such perfectionists that using cheaper display hardware and slower hard disks (etc.) would make the idea of a non-Apple-brand, OS X licensed, i386/AMD64 notebook line unlikely, unless there is shareholder pressure to capitalise further upon the move to Intel.
-- Cause problems with Linux running on non-Intel architecture. Erm... I just can't see this happening. Eg, getting a patch into mainline kernel with improves it isn't the easiest thing in the world (with the amount of people who will disagree with how it's done, how it should be done etc). One that hurts performance/stability on entire architectures, is gonna be even tougher.
Did you not get the memo about this year's Underhanded C Contest? The task is to design innocent-looking code which runs well on one architecture but orders of magnitude worse on another. I suspect it can be done: imagine patches in the kernel by Intel which were included in K8-optimised builds but which slowed down AMD execution of the software?
I hope that above paragraph is an unlikely paranoid fantasy.
Slashdot Mirror
AMD's technology preview day (Anandtech's report is at http://www.anandtech.com/cpuchipsets/showdoc.aspx? i=2768) said a lot that made me wonder about the future competitivenes of AMD. I'm neutral but the kind of stuff they were talking about made me doubt my conviction that Intel are guaranteed to pwn, but I am certain that the consumer is the winner in the upcoming battle for the best desktop CPU chips (and that's already been noted in the drop in prices between the two http://www.tomshardware.co.uk/2006/06/02/drop_in_a verage_processor_pricing/).
The notion of being able to put extra specialised hardware next to the CPU radically alters the way that PC's are going to be laid out and is a good lateral step along the road toward multi-threaded, multi-cored and multiple processing elements in computing. The K8L architecture and beyond are seeking to increase IPC throughput for AMD's chips as Intel has done for the Core architecture. I think that this throughput increase is the greatest asset to the Core's performance both for performance per cycle (==IPC) and performance per watt.
I'd contend that 'casual' games* are a snack food rather than a furniture as per your analogy. And larger games can be likened to banquets as music, video and gaming are all entertainment.
*: Casual games are expected to be the big growth area of the games market in the next couple of years -- just look at your smart phone and Nintendo's Wii as places for non-'gamer' games-playing to grow the market.
https://wiki.ubuntu.com/EdgyEft says October.
I dist-upgraded this morning. It boots quickly, makes good use of proprietary video drivers, is well-organised, functional and efficient. I would say that it is comparable to Win2k or WinXP from the month or so I've been using a release candidate. I hope that the fall's Edgy Eft release can be comparable to the looks and usability of these Windows Vista previews I keep seeing.
VFAT is the FAT filesystem driver and a standard part of the GNU tools that most Linuces come with -- FC3 has the support to read and write my FAT32 drives, so I'm sure it'll be there in FC5.
Like when I put marketing e-mail and non-spammer e-mail addresses into this distributed free remailer?
Would you run TCP/IP over Trusted Network Interconnect if it had no spam? [/devils-advocate]
I'm not sure if Halal certification insists on anything more than the way the animal is killed (a neck slash and blood drained from the corpse). I don't think it would insist on non-battery or free-range chickens, for example, but may insist on cattle not fed foodstuffs derived from other animal material.
Bleedin' liberals, always free with the truth.
Is it not possible to have the TPM and BIOS/EFI contain keys to encrypted data in the partition table and boot sector? Or to move the location of these data according to unique keys in the system?
And he worked at PepsiCo.
AM2 uses DDR2 memory. It is expected to become cheaper and more available than DDR when most companies move their production to make it. I don't want to fob you off with a cheap RTFM, but there is no easy alternative to reading what Tom's Hardware / AnandTech / Ars Technica have to say about the matter.
I think that Debian/Ubuntu doesn't scan its hardware at each boot to se if anything's changed, and doing that will enable it to not run drivers incompatible with your new MB. Win2K and WinXP updated their HAL (as you noticed) but Ubuntu is assuming it's working on the same hardware as before your upgrade.
I've looked, because I'd like to be able to say "run discover and it'll recognise your new setup", but there's little that helps. I believe Debian and Ubuntu use a program called Discover to detect hardware. The few links there are out there indicate that people seem to think that the software will work regardless after moving hardware. However, it may be that you're using software built for an architecture which isn't supported after your motherboard move, and I would advise either a reinstall of Ubuntu or soft reinstall of the packages you have on your system via dpkg --configure -a.
The Regulation of Investigatory Powers Act (2000) was set to make all legitimate uses of encrytion have a third party store a backup of the key in escrow. That part of the law, to my knowledge, got scrubbed before it hit the statute books.
I'm not convinced that they understand that this is 'free' software which they can take and make their own editions of. It may be that Oracle want a brand to be well-known to businesses, but I think that Oracle Linux running Oracle Desktop Linux would be better for market awareness of Oracle than Oracle supplying Novell Suse Desktop Linux.
Didn't you know? Bowe's is from the same people that brought you the great experiences you have from Hi-Fi by Panaphonics and Sorny.
My tentative suggestion -- given that I have no Apple computers and no experience with BootCamp -- would be to take the bootloader from ReactOS for its Win2k-like behaviour and stick it and grldr from the Grub4DOS in a primary-4 (as in fourth primary partition as the MBR looks at it) partition a couple of megabytes in size, then you have primary-2 and primary-3 and their extended partitions to use with your GNU/Linux installation. The Grub4DOS may need its Grub.exe and its stage-specific files updated to cope with EFI booting, and this may be the stumbing-block of the enterprise. Good luck!
Both stress and pressure are measured in force/area, so can be used interchangeably. (Right now I'm writing home: Mom! Stuff I specialised in Grad School is useful after all!)
"mount nice man" doesn't sound like my cup of tea.
I think that recognising the pace of computer alogrithm development and cutting the duration of a software patent to, say, three years, with the requirement that, within that time, a product is on the market containing the patent-covered method sounds like a good way to allow people to benefit from patents which drive forward progress without constricting others' ability to compete.
You forgot to mention that the candidates were also frozen for bugfixing. Towns has only lost two fingers to frostbite; the debian-privates e-mail list suggests that another candidate lost something more personal and delicate.
Negroponte says "We're heading to the point where 50 percent of the world will have a cell phone or some kind of (communication device) within 18 months. It's too voice-centric, and I could campaign to make it more data-centric, but that's going to happen, too."
That's seems to me to be unaware of the presence of GSM. 90% of the world's cell phones are GSM, an inherently digital-data format. Sure, it allows 9K6 data transfers or 28K8 if you use High Speed Circuit-Switched Data (HSCSD) mode, which I doubt will be available in too any developing nations. People aren't going to be downloading too much multimedia to the limited storage of these systems, so I don't see a low data rate as a problem.
Thanks fr allowing me to read your speach. I thought it was good cohesive and clear stuff. BTW: Are your references (presently displayed as [http://link/file%5D ) intended to be active hyperlinks? And have you edited the Democrat/Republican confusion mentioned elsewhere in here?
Would Bruce be kind enough to allow the fulltext be included here? The site it's hosted at is not responding nicely to a Slashdotting.
I think it's in their interests to make cheaper notebooks that are licensed to run OS X. Market segmentation is a normal part of the motor industry (in Europe, you can buy the VW Golf, Seat Ibiza or Skoda Fabia from the same chassis, engine range and trim options -- the only differences being price, and what's a Celeron if not budget-branded failed Pentium?), so why not in computers?.
Unfortunately, now's not a good time, as all the new tech in their notebooks is going to be expensive and need the payoff against RnD cost. Older tech would be better, providing an additional revenue from existing hardware research. However, Steve Jobs and Jon Ive are such perfectionists that using cheaper display hardware and slower hard disks (etc.) would make the idea of a non-Apple-brand, OS X licensed, i386/AMD64 notebook line unlikely, unless there is shareholder pressure to capitalise further upon the move to Intel.
-- Cause problems with Linux running on non-Intel architecture. Erm... I just can't see this happening. Eg, getting a patch into mainline kernel with improves it isn't the easiest thing in the world (with the amount of people who will disagree with how it's done, how it should be done etc). One that hurts performance/stability on entire architectures, is gonna be even tougher.
Did you not get the memo about this year's Underhanded C Contest? The task is to design innocent-looking code which runs well on one architecture but orders of magnitude worse on another. I suspect it can be done: imagine patches in the kernel by Intel which were included in K8-optimised builds but which slowed down AMD execution of the software?
I hope that above paragraph is an unlikely paranoid fantasy.