AMD always acknowledged where their stuff came from. From reading the Intel docs it might as well be a brand new set of extensions, not compatable with anything.
Intel supporting x86-64 is good, it's just the lack of acknowledgement that's kind of annoying.
This seems like as good a time as any to mention Peter Norvig's Teach Yourself Programming in Ten Years, an interesting look into the "NOW NOW NOW" attitude that is present in the computer industry (and the insane number of "Teach Yourself $skill in \d+ (days|hours)" books out there). This attitude is a part of what these tech training schools represent, and probably a reason why it hasn't worked out so well for them.
It also has very good advice for becoming an accomplished programmer.
Opteron is as server-grade as AMD gets right now, and has proven to be quite capable. It has dual-channel memory, large caches, and three HyperTransport links (for the bus and up to 8-way NUMA SMP, if I recall correctly).
Athlon 64 is the desktop version of the Opteron. It (usually) has single-channel memory, smaller caches, and only one HyperTransport link (for the bus).
Comparing the Opteron to the Xeon is quite valid, assuming they have similar cache sizes.
Theoretically, the Itanic and the Power4 should be a cut above the Xeon and Opteron, but as with all things in some real world applications it works the other way around.
a) Windows is where the pro audio software is in Intel land. People probably want to install out-of-the-box performance software on this thing, otherwise there's really no point.
b) It looks like the computer hardware is standard. Nothing is stopping you fom installing another OS on it. The site even says, "NEKO 64(TM) utilizes industry standard micro-ATX motherboards and processors that allow you to run standard operating systems." (Emphesis mine)
Of course, you may not be able to talk to the sound hardware once you replace the OS. Start reverse enigneering!:)
Kim Stanley Robinson's mars timekeeping is actually more practical if you're living on Mars all the time. (You'd probably want a "+38m12s" display or something however, to solve the "*WHAT'S the TIME*" issue.)
The reason I think it would be much better to do that than use Mars solar time if one were to live on Mars is the horrible mathematical disaster of having the "seconds" unit change! You'd have to do all physical calculations in "Earth seconds", which would be a unit you wouldn't ever see in your everyday life.
K.S.R's timekeeping system maintains the Earth second on Mars while still keeping accurate mars time, and not causing a physical unit disaster.
(For this mission, Mars solar time is fine - nobody at JPL is about to forget what an Earth second is.)
Sadly, Outlook Express was far more standards-compliant than full Outlook. And that's not saying much.
Here are just some of the things that annoy the hell out of me about Outlook:
Cannot use newsgroups without going through an Exchange server. Exchange servers really frell things up, as I'll explain below:
Exchange servers modify the Message-IDs of news messages they get via NNTP! This completely breaks threading for standards-complient stuff. This doesn't affect Outlook, though, because:
Outlook uses a completely different (and weaker!) threading system! One that's not compatible with standard References: or In-Reply-To: headers.
For more fun, Outlook uses the same stupid incompatible threading system when going through Exchange for email! Want to view your lame friend's messages threaded in mutt? If they use Outlook, too bad. Particularly bad on mailing lists.
Ironically, if Outlook connects directly to a real SMTP server to send mail and not via the Exchange backdoor, you get real In-Reply-To headers! *boggle*
Everyone in my office uses Outlook except for myself and a few others. I've wanted to set up a newsserver to replace our current policy of cc'ing random people when trying to have a discussion. Sadly, the only Microsoft solution would have been to use Outlook Express to connect to the news server. (No, installing Mozilla/Thunderbird on everyones machines and training people to use it is not an option, sadly.)
Probably the best thing to do is simply remap the vi commands to the Dvorak keys in their old QUERTY positions. You will have to remember the vi commands by position instead of letter, but if you've been using vi on QUERTY for any length of time that's how you do it now. That way the old "hjkl" commands will be on dvorak "dhtn" and you will be a happy Dvorak vi user.
I'm sure I've seen vim scripts that do exactly this. Try searching for "dvorak vim".
Any reasonably complex structured data will probably be put into a structure or class in Common Lisp. In addition to being specified, you get constant-time access to the elements:
;;;; Leading dots are because of slashcode lameness (defstruct item . "Strores a gronk foobar blazwart mrph." . quantity . product-code . name . description)
For most of the game, there are many more moves available than in chess, and it usually takes many more moves for a bad move to have an obvious affect.
Also, an evaluation function for a board position in Go is very complicated, depending on the life or death of stone groups on the board. The only way to determine life or death is to effectively know how to best play out the remainder of that area and see who comes out alive. This is very difficult for a computer, since the evaluation function is what makes your min/max algorithm work.
They make more money selling games they make for their console then they would selling games they make for Microsoft's console, as they don't have to pay themselves royalties for their own platform. Also the GameCube is so well integrated (compared to the Xbox) it is probably not sold for much of a loss at all, if not a small gain.
This really seems like a bad comparison. Find me a computer that ships with Firewire ports and NOT with USB ports.
USB has it's niche in cheap, slow (12Mb/s) devices. Firewire has one in devices that need to go fast (disks) or get guaranteed bandwidth (video). I don't see why these need to be mutually exclusive.
And, um, sir, if you think Firewire's speed doesn't come close to Ethernet, take a look at latencies on Gigabit Ethernet sometime - and the costs of Gigabit Ethernet controllers compared to Firewire ones. It is not nearly as suited for real-time activities as Firewire. 400Mb/s is nothing to sneer at, and that was Firewire 1.0!
Also, if you're serving DNS, get a good secondary DNS provider. Put them in as both your primary and secondary NS records. Then firewall port 53 and only let their hosts connect.
You still get the same effective service without nearly as much risk of random idiots exploiting buffer overflows.
Well if that were the case, why is Linux so damn stable? I have 2 systems running beta and expirmental wireless and firewire drivers that have never crashed. In my experience with Linux, drivers are sometimes a pain in the ass to get working, but once they are working they don't crash.
It's because the Linux drivers are written carefully. Linux is also very good at catching bad memory accesses in kernel space and giving a non-fatal oops instead of panicing. Trust me, though, go do some driver development and you'll be very happy you have journaled filesystems while watching the box reboot for the nth time.
Not that microkernels are a perfect solution to problems like this - whenever you screw around with interrupts off it's not very hard to hang the machine.
The public simply won't switch formats again, simply on claims of "better than CD quality" sound. Most people won't even be able to notice the improvements, when they go to check this new technology out.
I bet the retailers will have special discs or systems with the eq curve tweaked to impressive-but-ultimately-fatiguing levels to "help" typical consumers realize how much "better" it is in direct comparisons.
Because it is going to be really hard to tell the difference between 16 bit/44.1kHz and 24 bit/48kHz, especially when the 24 bit one has watermarking garbage on it!
The reason Halon is bad for you is because to put the fire out without water/liquid CO2/dry chemicals the oxygen in the room has to go away. Which makes breathing hard if you're still in the room once the Halon system goes off.
Oh, please. Don't start sounding like these wackos.
I believe if you actually run the numbers the amount of plutonium in a satellite RTG spread over the world would give everyone approximately the rad exposure of a day on the beach.
(I especially love how the stop cassini freaks talk about the "continued (but dimished)" dangers. What, do they think it's going to turn around or something?)
Yes, Durons are fabbed at Fab 25 in Austin, Texas. However, by the highly scientific process of glancing at any AMD processor, you will find the code words "ASSEMBLED IN MALAYSIA".
I am told this is Malaysian for "Assembled in Malaysia".
Granted, I don't think most sweatshops are up to the task of packaging processors, as a couple of water molecules will kind of screw them up. And I doubt the people working for AMD are getting slighted by local standards - their work environment must be clean because of the job done there. But saying that AMD processors are not made in Malaysia is obviously wrong, as the die that is fabbed in Austin is packaged there.
Lots of procedural geometry, procedural textures, and quite possibly procedural samples for music too. And really tight code no doubt.
As impressive as it is though, it takes a 400+ mhz machine with at least 128 megabytes of ram, and a powerful graphics card. And my Athlon 1900+ takes about 15 seconds to render all the textures and geometry before the demo starts.
There's a big difference between 64k of storage and 64k of memory. Cramming an ethernet driver, a TCP/IP stack, a web server, a cheezy C64 sampling driver, a RTSP/RTP server, the actual web pages, and all the variable data into 64k is impressive.
Slashdot Mirror
Intel supporting x86-64 is good, it's just the lack of acknowledgement that's kind of annoying.
It also has very good advice for becoming an accomplished programmer.
Athlon 64 is the desktop version of the Opteron. It (usually) has single-channel memory, smaller caches, and only one HyperTransport link (for the bus).
Comparing the Opteron to the Xeon is quite valid, assuming they have similar cache sizes.
Theoretically, the Itanic and the Power4 should be a cut above the Xeon and Opteron, but as with all things in some real world applications it works the other way around.
a) Windows is where the pro audio software is in Intel land. People probably want to install out-of-the-box performance software on this thing, otherwise there's really no point.
b) It looks like the computer hardware is standard. Nothing is stopping you fom installing another OS on it. The site even says, "NEKO 64(TM) utilizes industry standard micro-ATX motherboards and processors that allow you to run standard operating systems." (Emphesis mine)
Of course, you may not be able to talk to the sound hardware once you replace the OS. Start reverse enigneering! :)
The reason I think it would be much better to do that than use Mars solar time if one were to live on Mars is the horrible mathematical disaster of having the "seconds" unit change! You'd have to do all physical calculations in "Earth seconds", which would be a unit you wouldn't ever see in your everyday life.
K.S.R's timekeeping system maintains the Earth second on Mars while still keeping accurate mars time, and not causing a physical unit disaster.
(For this mission, Mars solar time is fine - nobody at JPL is about to forget what an Earth second is.)
Sadly, Outlook Express was far more standards-compliant than full Outlook. And that's not saying much.
Here are just some of the things that annoy the hell out of me about Outlook:
Everyone in my office uses Outlook except for myself and a few others. I've wanted to set up a newsserver to replace our current policy of cc'ing random people when trying to have a discussion. Sadly, the only Microsoft solution would have been to use Outlook Express to connect to the news server. (No, installing Mozilla/Thunderbird on everyones machines and training people to use it is not an option, sadly.)
I'm sure I've seen vim scripts that do exactly this. Try searching for "dvorak vim".
As screwed up as the U.S. has gotten, it's not even close to China as far as a police state goes.
Common Lisp is not just about lists.
Also, an evaluation function for a board position in Go is very complicated, depending on the life or death of stone groups on the board. The only way to determine life or death is to effectively know how to best play out the remainder of that area and see who comes out alive. This is very difficult for a computer, since the evaluation function is what makes your min/max algorithm work.
They make more money selling games they make for their console then they would selling games they make for Microsoft's console, as they don't have to pay themselves royalties for their own platform. Also the GameCube is so well integrated (compared to the Xbox) it is probably not sold for much of a loss at all, if not a small gain.
Parent is a goatse.cx link, don't mod up...
This really seems like a bad comparison. Find me a computer that ships with Firewire ports and NOT with USB ports.
USB has it's niche in cheap, slow (12Mb/s) devices. Firewire has one in devices that need to go fast (disks) or get guaranteed bandwidth (video). I don't see why these need to be mutually exclusive.
And, um, sir, if you think Firewire's speed doesn't come close to Ethernet, take a look at latencies on Gigabit Ethernet sometime - and the costs of Gigabit Ethernet controllers compared to Firewire ones. It is not nearly as suited for real-time activities as Firewire. 400Mb/s is nothing to sneer at, and that was Firewire 1.0!
You still get the same effective service without nearly as much risk of random idiots exploiting buffer overflows.
It's because the Linux drivers are written carefully. Linux is also very good at catching bad memory accesses in kernel space and giving a non-fatal oops instead of panicing. Trust me, though, go do some driver development and you'll be very happy you have journaled filesystems while watching the box reboot for the nth time.
Not that microkernels are a perfect solution to problems like this - whenever you screw around with interrupts off it's not very hard to hang the machine.
I bet the retailers will have special discs or systems with the eq curve tweaked to impressive-but-ultimately-fatiguing levels to "help" typical consumers realize how much "better" it is in direct comparisons.
Because it is going to be really hard to tell the difference between 16 bit/44.1kHz and 24 bit/48kHz, especially when the 24 bit one has watermarking garbage on it!
Or perhaps I could be wrong...
The reason Halon is bad for you is because to put the fire out without water/liquid CO2/dry chemicals the oxygen in the room has to go away. Which makes breathing hard if you're still in the room once the Halon system goes off.
I believe if you actually run the numbers the amount of plutonium in a satellite RTG spread over the world would give everyone approximately the rad exposure of a day on the beach.
(I especially love how the stop cassini freaks talk about the "continued (but dimished)" dangers. What, do they think it's going to turn around or something?)
Not all movie theatres anally rape you for a $20 showing and $8 popcorn and drinks, either.
*cough*shamelessplug*cough*
(well, if you're in Illinois, that is.)
Good to know - thanks...
:)
Though the statement, "You can usually tell from context which definition of time and space someone is using" just has to make me chuckle.
Umm, no. We're talking powers of two here. 1024 bits is a number 2 times bigger than 1023 bits. Not 512 bits.
Even if it is 3.01 times larger, that's still an effective strength of at least 1022 bits.
(For what it's worth, I've only read the abstract.)
The PS2 has two USB ports and one Firewire port, actually, so there's no problem there.
Yes, Durons are fabbed at Fab 25 in Austin, Texas. However, by the highly scientific process of glancing at any AMD processor, you will find the code words "ASSEMBLED IN MALAYSIA".
I am told this is Malaysian for "Assembled in Malaysia".
Granted, I don't think most sweatshops are up to the task of packaging processors, as a couple of water molecules will kind of screw them up. And I doubt the people working for AMD are getting slighted by local standards - their work environment must be clean because of the job done there. But saying that AMD processors are not made in Malaysia is obviously wrong, as the die that is fabbed in Austin is packaged there.
As impressive as it is though, it takes a 400+ mhz machine with at least 128 megabytes of ram, and a powerful graphics card. And my Athlon 1900+ takes about 15 seconds to render all the textures and geometry before the demo starts.
There's a big difference between 64k of storage and 64k of memory. Cramming an ethernet driver, a TCP/IP stack, a web server, a cheezy C64 sampling driver, a RTSP/RTP server, the actual web pages, and all the variable data into 64k is impressive.