If India had any sort of self-respect they'd invest in them.
In Canada and the states we had to fight hard for decent working conditions. If they're just pussies and don't want to rock the boat they'll never get any better.
Which part was the joke? Radio waves travel just as fast as light does [at least in a vacuum].
And it doesn't matter, doppler only occurs for radio [or light] when two bodies are moving relative to one another. If you're on the plane and you're AP is on the plane then you're both stationary.
You'd get more acceleration from moving the laptop in your lap then by the amount the airframe bends as it travels through the air...
I know you seem to be kiddin but this thread is seriously bunkable.
Get headphones. I got nice big sony DJ headphones [I'm not a DJ but I like the style] and they basically cover my entire ears. Insert your fav mp3/whatever player and boom no more "quarterly sales" update in the gate area.
Feel lucky. Hell on my one and only KLM flight I could hardly reach forward to eat the crap they called a meal.
And what's with the arm rests that sit 2" off the seat? My elbow doesn't go that far (I'm ~6'1") down and if the plane ever did roll I'd just topple over it anyways, it isn't high enough to catch my body in any meaningful manner. I swear plane designers have to either be super greedy or REALLY SHORT.
I also laugh when I see some 6'6" sucker get on the plane ahead of me. At least they're going through more pain.:-)
You do realize the plane has radios of its own right?
In the time the signal travels the max 100m from one end to another the plane probably moved 1mm in space at most. They're also fixed points relative to each other.
By your logic microwave towers shouldn't work, and the earth rotates FAR FASTER than planes travel (which is why you still lose time when you travel west).
Let's see... car cost $12000 and game costs $40... I wonder why there are lemon laws for only cars...
Yes it sucks, and yes there should [and is] a remedy. But often it's just buyer beware. Now for things like MSFT the bigger complaint is that XP and the like aren't exactly "new". So why are we finding flaws that stem from the Win95 code base all the way into this century?
The problem with software is manyfold but mostly can be summed up with two statements.
Even the educated folk tend to have zero experience. So they go from their masters program at uni to working on a billion dollar software project and have yet to actually write and support a single product. Of course you get bugs, because the people who made your product have zero previous experience!!!
It's like knowing the recipe for a good dinner and actually being able to make it. Totally two different things.
These folk may know how to sort arrays or parse a BNF grammar but they can't figure out time management when it comes to design and then actually coding and testing. They don't have the experience to write clean and maintainable code, they're not used to the oddity of bugs that software tends to have or how to diagnose it, they're not used to writing documentation that others have to read, etc, etc, basically things they would get from experience.
If anything a successful OSS project [e.g. one which is useful, supported and well documented] is something hiring managers should look for on any developers resume.
You can't be liable for something without actually causing the damages or agreeing to be responsible if something you did indirectly did cause them.
If someone downloads your software, under say, a public domain license, and then proceeds to misuse it. You're not liable for a dime because you never agreed to warrant the software.
There are many parts to civil liability law [in Canada and elsewhere I imagine] but the jist of it is you have to be responsible for the damages. If you give out free software as public domain, there is no license and the user is on their own. They can't claim you agreed to distribute the software because anyone can distribute the software (effectively nobody has any claim to ownership of the product).
Now with licenses that imply copyright, e.g. GPL or BSD, technically you are allowing them to use it under the agreement of your license. Both have a "WARRANTY" claim, which without would open the developers up to civil liabilities. E.g. suppose GPL didn't have such a clause and the Linux kernel crashed. I could probably sue the maintainers for causing harm [in the form of downtime] because their software that they licensed to me [e.g. by allowing me to accept the terms of the GPL when I got it from their servers or agents] was defective [IANAL and YMMV but that's what I get out of all of it].
OSS doesn't need big business to survive. It's doing just fine right now. It would be ideal if it was an option, e.g. buy a Dell and get a good Linux distro on it. But so long as we're not locked out of putting OSS tools on our own computers [e.g. TPM bullshit] it doesn't matter.
Big business has to invest time into OSS if they want anything out of it. Tool doesn't work? Spec out the cost and see if it's feasible to invest time. If it takes you one person 1 day to fix a tool that will let you open up a million dollar market... isn't it worth it?
I'm simply SHOCKED I tell you, SHOCKED that you claim he's a press whore.
I've always had the highest regard for the guy... oh who am I kiddin... hehehe.
Bruce hasn't had an original thought since the late 80s. Why should he start now? He can spew CNN quality FUD like the best of them and people pay to listen to his wonderful insight about how the world is coming to an end, that is of course, unless they purchase his protection plans and buy his oh so carefully worded books.
Um to be more correct all K8 processors have THREE HT links. The difference is whether they can act as coherent links. A 2xx series processor will have one link between processors and at least one to the northbridge I/O controller.
No, you missed the point. Opterons have activated one, two or three coherent HT links. That lets them keep their caches [and memories] in sync.
In a typical FSB MP system the processors snoop the bus and look for reads/writes.
In the opteron world the processors send out cache probes via the HT links. Athlons have ZERO cHT links activated which means they cannot work in MP systems.
It could be that they're toying with the idea or the slides could be, oh I dunno, doctored.
Opening the Athlon side up to MP+DC is a really stupid idea because it directly undercuts the Opteron line.
Best I can tell the plan is quad-core on die (as per the CEOs press release), cache improvements, other architectural improvement and the rest filters down into the "consumer" brand. [all public information from press releases]
You can currently have upto eight dual cores on one network. There are products that extend this with cHT enabled switches so you can go pretty high in the # of nodes.
You can't MP a Athlon64/FX setup because the memory is local to the processor [e.g. not the northbridge] and there are NO cHT links.
There is no FSB and the memory is LOCAL to the processor. How would this maintain coherency? The Athlon64 processors also don't allow cHT. Not that they don't physically have support for it, just it's been disabled.
Given where I work, and that I've never heard of this before today... I suspect it's a hoax.
The only way this would work is if the OS was aware of it and manually routed data from one node to another (e.g. like a northbridge DMA device you can pipe info to).
It's like why eat 2000 calories in a day when you CAN eat 3000. It is not like this thing has efficient low power components. It's just an amalgamation of the highest end parts in one package.
All it gives you is gloating power, as in "I shelled out a lot for this and you can't".
Besides there is more to x86-64 than "48 bit addresses". People like you just like the shiny cover words. How about "twice the GPRs" and "twice the SIMD registers"?
Though I agree, MCW chips will be neat to play with. I'm looking foward to grabbing one for a desktop when they're not too too expensive [mmm benchmarking...]
Because there are occasionally real stories worth yammering about.
Some asshat hosting torrents then dancing around copyright law isn't what I call newsworthy. He's a jerk.
Hate the studios all you want, but if you didn't like or want the media you wouldn't pirate it. The media is a work product so even if the artist is getting screwed [their choice btw] doesn't mean you should effectively benefit from it without compensation.
So this guy isn't hosting the files, he's making it damn well easier to get to them though. And while I can see they're not the same action I also won't shed a tear if it gets taken down.
It isn't like he's hosting torrents to Linux distro ISOes or something.
I know a lot of people who don't read slashdot who can handle a text editor.
I was pointing out that the OP gets flamed a lot because his assumptions are rather weak. Sure there are many people who can hardly figure out their 'puter but editing text files isn't "bad".
Your definition of failure is subject and THAT IS WHY you get flamed.
I find it trivial to sudoedit a config file. It's certainly no slower than going through a dozen windows/tabs/etc to find something. And the fact that YOU CAN fix almost anything in OSS desktops is part of their usefulness.
I've had to hack a few projects [like QEMU and the kernel] to get them to build to my specs. I don't see that as a failure because in the commercial side THEY WOULD JUST NOT WORK AT ALL.
Yeah, but I think that's the point. The OSS community is doing just fine right now. It doesn't need mainstream attention to survive. It'd just be *nice*.
This is basically all of humanity though.
People buying HDTV early without doing research? Getting stuck with weird modes that can't do the copy protection as well...
People buying gas guzzling cars that need maintanance every two weeks.... oh well...
People buying over priced power sucking desktops for the most basics of tasks...
etc, etc, etc.
The recurring theme is "I shouldn't have to learn stuff to do stuff" like learn how to use a computer to use one, or how a car works to own one, or the gist of the HD specs before shelling out five grand on a TV, etc, etc...
Really I think people deserve what they get. If you're too lazy to actually work for something [e.g. a free and stable desktop OS] then you don't deserve one.
Technically on x86 you don't need to flush the TLB either. If you want you could just use upper bits as a PID. Specially on x86-64 where most processes use less tahn 24 bits of address space it'd be trivial to do that. Unfortunately, in long mode there are no segment limits on DS or CS which would mean you could pollute other processes. "segmentation" is provided by paging in x86-64 mode.
I'm just saying most OSes *do* that then optimize to avoid it where possible. So you really do need a fresh CR3 between processes. And you can't just hack some level of the page directory because the TLB wouldn't reflect the changes.
The speedup comes from TLB caching between processes. Not from "double caching".
In Linux when you switch processes the TLB is flushed [e.g. reloading CR3 on x86-*]. This is a safe [but slow] way to ensure your virtual memory for a given process is mapped correctly. I'm guessing [didn't fully read the linked research papers] that they share a virtual memory base between processes but map processes to different regions or something. Unless they have segment limits this will cause problems with process isolation.
For those not in the know, a TLB cache holds the translation of a virtual address into a physical one. Parsing a typical 32-bit address requires several layers [with 4KB pages it's four I think] of table lookup which is slow if you had to do it for every memory access. For example, take your 32-bit address, the lower 12 bits is the byte in a 4KB page, the next 10 bits points selects one 4KB page, the next 10 bits selects one 1024-entry array of pointers to 4KB pages. [iirc]. It's even worse in x86-64 mode as we are parsing a 48-bit virtual address.
So the processor will cache TLB lookups. When you switch processes you have to flush it because the translations don't map to your processes physicals memory.
Slashdot Mirror
It's called labour laws.
If India had any sort of self-respect they'd invest in them.
In Canada and the states we had to fight hard for decent working conditions. If they're just pussies and don't want to rock the boat they'll never get any better.
Tom
Which part was the joke? Radio waves travel just as fast as light does [at least in a vacuum].
And it doesn't matter, doppler only occurs for radio [or light] when two bodies are moving relative to one another. If you're on the plane and you're AP is on the plane then you're both stationary.
You'd get more acceleration from moving the laptop in your lap then by the amount the airframe bends as it travels through the air...
I know you seem to be kiddin but this thread is seriously bunkable.
Tom
Get headphones. I got nice big sony DJ headphones [I'm not a DJ but I like the style] and they basically cover my entire ears. Insert your fav mp3/whatever player and boom no more "quarterly sales" update in the gate area.
I swear some people are just so self-important...
Tom
Feel lucky. Hell on my one and only KLM flight I could hardly reach forward to eat the crap they called a meal.
:-)
And what's with the arm rests that sit 2" off the seat? My elbow doesn't go that far (I'm ~6'1") down and if the plane ever did roll I'd just topple over it anyways, it isn't high enough to catch my body in any meaningful manner. I swear plane designers have to either be super greedy or REALLY SHORT.
I also laugh when I see some 6'6" sucker get on the plane ahead of me. At least they're going through more pain.
Tom
This has to be a joke.
You do realize the plane has radios of its own right?
In the time the signal travels the max 100m from one end to another the plane probably moved 1mm in space at most. They're also fixed points relative to each other.
By your logic microwave towers shouldn't work, and the earth rotates FAR FASTER than planes travel (which is why you still lose time when you travel west).
Tom
Let's see... car cost $12000 and game costs $40... I wonder why there are lemon laws for only cars...
Yes it sucks, and yes there should [and is] a remedy. But often it's just buyer beware. Now for things like MSFT the bigger complaint is that XP and the like aren't exactly "new". So why are we finding flaws that stem from the Win95 code base all the way into this century?
The problem with software is manyfold but mostly can be summed up with two statements.
1. incompetent "developers" [aka scriptmonkeys]
2. Incompetent "managers" [aka promoted developers]
Even the educated folk tend to have zero experience. So they go from their masters program at uni to working on a billion dollar software project and have yet to actually write and support a single product. Of course you get bugs, because the people who made your product have zero previous experience!!!
It's like knowing the recipe for a good dinner and actually being able to make it. Totally two different things.
These folk may know how to sort arrays or parse a BNF grammar but they can't figure out time management when it comes to design and then actually coding and testing. They don't have the experience to write clean and maintainable code, they're not used to the oddity of bugs that software tends to have or how to diagnose it, they're not used to writing documentation that others have to read, etc, etc, basically things they would get from experience.
If anything a successful OSS project [e.g. one which is useful, supported and well documented] is something hiring managers should look for on any developers resume.
Tom
You can't be liable for something without actually causing the damages or agreeing to be responsible if something you did indirectly did cause them.
If someone downloads your software, under say, a public domain license, and then proceeds to misuse it. You're not liable for a dime because you never agreed to warrant the software.
There are many parts to civil liability law [in Canada and elsewhere I imagine] but the jist of it is you have to be responsible for the damages. If you give out free software as public domain, there is no license and the user is on their own. They can't claim you agreed to distribute the software because anyone can distribute the software (effectively nobody has any claim to ownership of the product).
Now with licenses that imply copyright, e.g. GPL or BSD, technically you are allowing them to use it under the agreement of your license. Both have a "WARRANTY" claim, which without would open the developers up to civil liabilities. E.g. suppose GPL didn't have such a clause and the Linux kernel crashed. I could probably sue the maintainers for causing harm [in the form of downtime] because their software that they licensed to me [e.g. by allowing me to accept the terms of the GPL when I got it from their servers or agents] was defective [IANAL and YMMV but that's what I get out of all of it].
OSS doesn't need big business to survive. It's doing just fine right now. It would be ideal if it was an option, e.g. buy a Dell and get a good Linux distro on it. But so long as we're not locked out of putting OSS tools on our own computers [e.g. TPM bullshit] it doesn't matter.
Big business has to invest time into OSS if they want anything out of it. Tool doesn't work? Spec out the cost and see if it's feasible to invest time. If it takes you one person 1 day to fix a tool that will let you open up a million dollar market... isn't it worth it?
Tom
I'm simply SHOCKED I tell you, SHOCKED that you claim he's a press whore.
I've always had the highest regard for the guy... oh who am I kiddin... hehehe.
Bruce hasn't had an original thought since the late 80s. Why should he start now? He can spew CNN quality FUD like the best of them and people pay to listen to his wonderful insight about how the world is coming to an end, that is of course, unless they purchase his protection plans and buy his oh so carefully worded books.
Tom
GUIs aren't always nice when your graphics card is acting up ... or you're not at the box.
I'd rather remote admin/config a box through SSH then through VNC or rdesktop [neither of which have any security].
Tom
Um to be more correct all K8 processors have THREE HT links. The difference is whether they can act as coherent links. A 2xx series processor will have one link between processors and at least one to the northbridge I/O controller.
Tom
Believe it or not, but the employees get briefed about these things WELL in advance. ...
I just checked my email. It was announced today as a press release by AMD... damn...
They don't really say what "4x4" means... oh well it's not my job anyways.
Tom
No, you missed the point. Opterons have activated one, two or three coherent HT links. That lets them keep their caches [and memories] in sync.
In a typical FSB MP system the processors snoop the bus and look for reads/writes.
In the opteron world the processors send out cache probes via the HT links. Athlons have ZERO cHT links activated which means they cannot work in MP systems.
Tom
News to me.
It could be that they're toying with the idea or the slides could be, oh I dunno, doctored.
Opening the Athlon side up to MP+DC is a really stupid idea because it directly undercuts the Opteron line.
Best I can tell the plan is quad-core on die (as per the CEOs press release), cache improvements, other architectural improvement and the rest filters down into the "consumer" brand. [all public information from press releases]
Tom
"quad core" means in one processor package.
You can currently have upto eight dual cores on one network. There are products that extend this with cHT enabled switches so you can go pretty high in the # of nodes.
You can't MP a Athlon64/FX setup because the memory is local to the processor [e.g. not the northbridge] and there are NO cHT links.
Tom
There is no FSB and the memory is LOCAL to the processor. How would this maintain coherency? The Athlon64 processors also don't allow cHT. Not that they don't physically have support for it, just it's been disabled.
Given where I work, and that I've never heard of this before today... I suspect it's a hoax.
The only way this would work is if the OS was aware of it and manually routed data from one node to another (e.g. like a northbridge DMA device you can pipe info to).
Tom
It's not advanced, it's just over done.
It's like why eat 2000 calories in a day when you CAN eat 3000. It is not like this thing has efficient low power components. It's just an amalgamation of the highest end parts in one package.
All it gives you is gloating power, as in "I shelled out a lot for this and you can't".
Tom
not to start a fanboi war but ....
AMD is coming out with new mobile chips too...
Besides there is more to x86-64 than "48 bit addresses". People like you just like the shiny cover words. How about "twice the GPRs" and "twice the SIMD registers"?
Though I agree, MCW chips will be neat to play with. I'm looking foward to grabbing one for a desktop when they're not too too expensive [mmm benchmarking...]
Tom
Because there are occasionally real stories worth yammering about.
Some asshat hosting torrents then dancing around copyright law isn't what I call newsworthy. He's a jerk.
Hate the studios all you want, but if you didn't like or want the media you wouldn't pirate it. The media is a work product so even if the artist is getting screwed [their choice btw] doesn't mean you should effectively benefit from it without compensation.
So this guy isn't hosting the files, he's making it damn well easier to get to them though. And while I can see they're not the same action I also won't shed a tear if it gets taken down.
It isn't like he's hosting torrents to Linux distro ISOes or something.
Tom
Attempting to care....FAILED. :-)
Some random website full of 1337 computer wizards who are probably breaking the law or just being annoying gets taken down. Big deal.
How many kids went hungry last night because their parents wouldn't feed them?
How many people died yesterday as a result of war?
etc, etc, etc.
Chances are they are doing something worthy of getting "picked on" and frankly I don't see why anyone should care in the slighest.
Tom
Happened to crypto in the 90s and communism in the 60s.
:-)
Face it, americans just don't like thinkers.
Tom
I know a lot of people who don't read slashdot who can handle a text editor.
I was pointing out that the OP gets flamed a lot because his assumptions are rather weak. Sure there are many people who can hardly figure out their 'puter but editing text files isn't "bad".
Tom
Your definition of failure is subject and THAT IS WHY you get flamed.
I find it trivial to sudoedit a config file. It's certainly no slower than going through a dozen windows/tabs/etc to find something. And the fact that YOU CAN fix almost anything in OSS desktops is part of their usefulness.
I've had to hack a few projects [like QEMU and the kernel] to get them to build to my specs. I don't see that as a failure because in the commercial side THEY WOULD JUST NOT WORK AT ALL.
Tom
Yeah, but I think that's the point. The OSS community is doing just fine right now. It doesn't need mainstream attention to survive. It'd just be *nice*.
This is basically all of humanity though.
People buying HDTV early without doing research? Getting stuck with weird modes that can't do the copy protection as well...
People buying gas guzzling cars that need maintanance every two weeks.... oh well...
People buying over priced power sucking desktops for the most basics of tasks...
etc, etc, etc.
The recurring theme is "I shouldn't have to learn stuff to do stuff" like learn how to use a computer to use one, or how a car works to own one, or the gist of the HD specs before shelling out five grand on a TV, etc, etc...
Really I think people deserve what they get. If you're too lazy to actually work for something [e.g. a free and stable desktop OS] then you don't deserve one.
Tom
Technically on x86 you don't need to flush the TLB either. If you want you could just use upper bits as a PID. Specially on x86-64 where most processes use less tahn 24 bits of address space it'd be trivial to do that. Unfortunately, in long mode there are no segment limits on DS or CS which would mean you could pollute other processes. "segmentation" is provided by paging in x86-64 mode.
I'm just saying most OSes *do* that then optimize to avoid it where possible. So you really do need a fresh CR3 between processes. And you can't just hack some level of the page directory because the TLB wouldn't reflect the changes.
Tom
Tom
This is OT.
The speedup comes from TLB caching between processes. Not from "double caching".
In Linux when you switch processes the TLB is flushed [e.g. reloading CR3 on x86-*]. This is a safe [but slow] way to ensure your virtual memory for a given process is mapped correctly. I'm guessing [didn't fully read the linked research papers] that they share a virtual memory base between processes but map processes to different regions or something. Unless they have segment limits this will cause problems with process isolation.
For those not in the know, a TLB cache holds the translation of a virtual address into a physical one. Parsing a typical 32-bit address requires several layers [with 4KB pages it's four I think] of table lookup which is slow if you had to do it for every memory access. For example, take your 32-bit address, the lower 12 bits is the byte in a 4KB page, the next 10 bits points selects one 4KB page, the next 10 bits selects one 1024-entry array of pointers to 4KB pages. [iirc]. It's even worse in x86-64 mode as we are parsing a 48-bit virtual address.
So the processor will cache TLB lookups. When you switch processes you have to flush it because the translations don't map to your processes physicals memory.
Tom