It's not. However, instead of having a slight pause (that scales linearly with the number of processes in-flight) after all the processes have exhausted their timeslices, there is a constant *tiny* pause after each process exhausting their timeslice. Since previously, the kernel was not preemptable, the O(n) recalculation would potentially cause, on large systems especially, a pause while the timeslices were recalculated. The new scheduler increases system responsiveness by eliminating this mass-recalculation.
I think the majority of major improvements on the 2.6 kernel over 2.4 were placed for increasing system responsiveness.
That said, if you can fit your data sets and your program on to the video memory (128MB isn't uncommon on high-end), and you're doing lengthy calculations on these sets while being only interested in the results (again, not uncommon in HPC), then the relative slowness of reading these results back becomes a nonissue.
It's apparently a case of "easy to learn" vs. "easy to use". Actually, I tend to not look at the instrument cluster. I have it adjusted so I can "see" the needles moving from my peripheral vision (that's why I *loathe* digital-only dashes).
Also, as you get better at driving, you start getting a sense for your speed anyway.
Well yes, my Nikkomat (backup body) uses a watch battery for the light meter too, and everything else is mechanical, not needing batteries for anything. However, my primary body, the F-100, becomes an expensive paperweight after its 4 AA cells die. In moderate weather, I can shoot about 20-40 rolls (really, six months of 5-10 rolls/month) between new batteries, but I suspect in *cold* weather, that changes dramatically.
As grandparent said, if I were to go to a much colder climate, I'd take non-G lenses (those lack an apature wheel, thus making them incompatible with my Nikkomat), a separate lightmeter, and well, my Nikkomat.
Gotta love 30year cameras. They're built like a tank.
My Nikon F-100 wasn't that hard to get used to. Just hit "mode" and rotate the main dial, and you go through the four main modes - Manual, Aperture priority, Shutter priority, and Programme (full auto). I'm not quite sure how to use the P-mode, because I'm almost always in M or A, and occassionally S.
What's really cool about the newer Nikon AF-S lenses and most Canon USM lenses is that you don't have to hit that silly 2-position switch, flipping between manual and autofocus. I can just play with the focus in the autofocus mode. Very cool.
If money were no object, I'd probably rebuild my kit with Canon kit, though. Their autofocus seems a touch better, and Canon glass is at least as good as Nikon glass, and I lust over the Canon 1Ds.
And those of us who wanted to run Falcon 3.0 (required 600k free conventional memory!) installed QEMM.
Shortly afterwards, a bunch of programs started putting themselves into protected mode. I remember writing a program to see how much conventional memory I can consume before things started going hairy.
Perhaps he labels the jewelcase, and be fanatical about replacing the CD? I mean, I agree, if it weren't for my sharpie-labels, I'd never be able to figure out the contents on my CD.
Huh? Look, just because most of the fruits of computer science are available to you and I without million-dollar budgets, it doesn't mean they're trivial. It's the same with chemistry. You and I can, given sufficient motivation and time, read about tonnes of proprietary (and perhaps patented) processes. However, to perform them, you may need something much more sophisticated than a kitchen sink and a couple of flasks.
The biggest difference here is that the fruits of computer science tend to be a lot more transparent. However, that by itself does not make CS more (or less) of a science than Chemistry or Physics. (Note I'm not saying CS is a science, I'm saying the reasons parent gave were not valid reasons for arguing validity of CS as a branch of science).
The article mentions NT4. Perhaps, and this is purely my speculation, they wanted to see how well Windows 2003 Server does vs Samba 3 when replacing existing NT4 installations.
You're right, though, for anything larger than a workgroup-level server, you'll want something beefier in the IO department than what these people have used.
But only if you own a reasonably recent mac running OS X. Mind you, I'm a happy owner of a 12" PBG4 and Airport Extreme, but there are caveats to every purchase decisions:)
in American English, it's the equivalent of toilet and restroom, respectively. As one would not ask their hostess where her toilet is, but instead would ask where her restroom is; one would not ask where the toire is, but where the otearai is. Does that make sense?
For example, the polite way to say restroom in Japanese is either "toire" (derived from the English word toilet) or "otearai" (imported from chinese, literally meaning "hand-wash"). There exist native Japanese words for restroom, but they connote dirtiness and one would never use them in polite company.
You're not a native Japanese speaker, are you?:) "Toire" is ranked lower than "Otearai". For a sign to say "toire" is perfectly acceptable, but for a person to ask for a "toire" is rather rude, and here is why.
One must understand that within the Japanese culture exists a strict differentiation between "ranks" of speakers (though to a much smaller degree and severity than the Korean culture), whether they be people or objects. In the case of information provider-consumer role, the provider is ranked higher than the consumer (though other overriding social roles will intervene, thus making the role definition anything but simple). Therefore, it is perfectly acceptable for a sign, in the role of an information provider, to specify the location of a toilet as "toire", but person asking for one, since they are in the role of an information consumer, should always ask for an "otearai".
In English culture, similar relationships exist, but the contexts are different; people always rank above inanimate objects. Therefore, one will rarely see English signs saying "Go Straight", but will instead indicate "Ahead", for the former is an order, but the latter is an indication.
What, are you still using Enron and Worldcom as examples? Are you using the failed dot-bombs as examples? Please, the former represent a tiny fraction of established large businesses, and the bulk of the latter were expected to fail by their VC's.
Most of the CEO's are doing their best to maximise shareholder value, and that doesn't change. Sometimes their methods pan out, sometimes they don't. However, things learned during MBA courses, be it asset management or basic corporate accounting, will pay off.
In addition, an MBA curriculum is one of the very few locations where I can make a very diverse network. My fellow students will go into various differing industries, but as long as I remain in touch with them, they become my potential suppliers and clients. You can't buy that or build that spontaneously.
You may not feel the need for an MBA. In the meantime, I'll take my courses and my network.
Excuse me? Learning how to effectively manage a corporation, and more importantly, learning how a corporate brain thinks is useless? I'm failing to see how the MBA will be useless to anyone with visions of starting a company, which is what went through my head when I first saw the/. article.
Slashdot Mirror
probably, but that's a separate issue. You can do processor affinity with the previous O(n) scheduler too.
It's not. However, instead of having a slight pause (that scales linearly with the number of processes in-flight) after all the processes have exhausted their timeslices, there is a constant *tiny* pause after each process exhausting their timeslice. Since previously, the kernel was not preemptable, the O(n) recalculation would potentially cause, on large systems especially, a pause while the timeslices were recalculated. The new scheduler increases system responsiveness by eliminating this mass-recalculation.
I think the majority of major improvements on the 2.6 kernel over 2.4 were placed for increasing system responsiveness.
(and yes, I'm using my Karma bonus)
WARNING: Lots of conjecture involved.
:)
That said, if you can fit your data sets and your program on to the video memory (128MB isn't uncommon on high-end), and you're doing lengthy calculations on these sets while being only interested in the results (again, not uncommon in HPC), then the relative slowness of reading these results back becomes a nonissue.
Does that help?
How can perl be more expressive than C when its interpreter is programmed in C?
I thought fighter HUDs were focused so they'll be focused on your eyes when they're at infinity?
It's apparently a case of "easy to learn" vs. "easy to use". Actually, I tend to not look at the instrument cluster. I have it adjusted so I can "see" the needles moving from my peripheral vision (that's why I *loathe* digital-only dashes).
Also, as you get better at driving, you start getting a sense for your speed anyway.
Indeed, I have various equipment manufacturers to thank for my cousin's life. It still took him 3 months to recover from his accident, though.
Ooooh Rangefinder cameras.
I want a 645-format rangefinder. They're an absolute dream to focus, and the larger negative means they're a little more forgiving than my 35mm gear.
Well yes, my Nikkomat (backup body) uses a watch battery for the light meter too, and everything else is mechanical, not needing batteries for anything. However, my primary body, the F-100, becomes an expensive paperweight after its 4 AA cells die. In moderate weather, I can shoot about 20-40 rolls (really, six months of 5-10 rolls/month) between new batteries, but I suspect in *cold* weather, that changes dramatically.
As grandparent said, if I were to go to a much colder climate, I'd take non-G lenses (those lack an apature wheel, thus making them incompatible with my Nikkomat), a separate lightmeter, and well, my Nikkomat.
Gotta love 30year cameras. They're built like a tank.
My Nikon F-100 wasn't that hard to get used to. Just hit "mode" and rotate the main dial, and you go through the four main modes - Manual, Aperture priority, Shutter priority, and Programme (full auto). I'm not quite sure how to use the P-mode, because I'm almost always in M or A, and occassionally S.
What's really cool about the newer Nikon AF-S lenses and most Canon USM lenses is that you don't have to hit that silly 2-position switch, flipping between manual and autofocus. I can just play with the focus in the autofocus mode. Very cool.
If money were no object, I'd probably rebuild my kit with Canon kit, though. Their autofocus seems a touch better, and Canon glass is at least as good as Nikon glass, and I lust over the Canon 1Ds.
ROI seems excellent. Now it's all about risk management, isn' t it? :)
And those of us who wanted to run Falcon 3.0 (required 600k free conventional memory!) installed QEMM.
Shortly afterwards, a bunch of programs started putting themselves into protected mode. I remember writing a program to see how much conventional memory I can consume before things started going hairy.
Perhaps he labels the jewelcase, and be fanatical about replacing the CD? I mean, I agree, if it weren't for my sharpie-labels, I'd never be able to figure out the contents on my CD.
ITYM "in the cold server room".
a 12 meter powerbook? Mon dieu!
Huh? Look, just because most of the fruits of computer science are available to you and I without million-dollar budgets, it doesn't mean they're trivial. It's the same with chemistry. You and I can, given sufficient motivation and time, read about tonnes of proprietary (and perhaps patented) processes. However, to perform them, you may need something much more sophisticated than a kitchen sink and a couple of flasks.
The biggest difference here is that the fruits of computer science tend to be a lot more transparent. However, that by itself does not make CS more (or less) of a science than Chemistry or Physics. (Note I'm not saying CS is a science, I'm saying the reasons parent gave were not valid reasons for arguing validity of CS as a branch of science).
It's a joke. Laugh :)
The article mentions NT4. Perhaps, and this is purely my speculation, they wanted to see how well Windows 2003 Server does vs Samba 3 when replacing existing NT4 installations.
You're right, though, for anything larger than a workgroup-level server, you'll want something beefier in the IO department than what these people have used.
But only if you own a reasonably recent mac running OS X. Mind you, I'm a happy owner of a 12" PBG4 and Airport Extreme, but there are caveats to every purchase decisions :)
The grandparent mentioned real applications. When the software package costs $20k, who cares about a $60 difference?
in American English, it's the equivalent of toilet and restroom, respectively. As one would not ask their hostess where her toilet is, but instead would ask where her restroom is; one would not ask where the toire is, but where the otearai is. Does that make sense?
You're not a native Japanese speaker, are you? :) "Toire" is ranked lower than "Otearai". For a sign to say "toire" is perfectly acceptable, but for a person to ask for a "toire" is rather rude, and here is why.
One must understand that within the Japanese culture exists a strict differentiation between "ranks" of speakers (though to a much smaller degree and severity than the Korean culture), whether they be people or objects. In the case of information provider-consumer role, the provider is ranked higher than the consumer (though other overriding social roles will intervene, thus making the role definition anything but simple). Therefore, it is perfectly acceptable for a sign, in the role of an information provider, to specify the location of a toilet as "toire", but person asking for one, since they are in the role of an information consumer, should always ask for an "otearai".
In English culture, similar relationships exist, but the contexts are different; people always rank above inanimate objects. Therefore, one will rarely see English signs saying "Go Straight", but will instead indicate "Ahead", for the former is an order, but the latter is an indication.
What, are you still using Enron and Worldcom as examples? Are you using the failed dot-bombs as examples? Please, the former represent a tiny fraction of established large businesses, and the bulk of the latter were expected to fail by their VC's.
Most of the CEO's are doing their best to maximise shareholder value, and that doesn't change. Sometimes their methods pan out, sometimes they don't. However, things learned during MBA courses, be it asset management or basic corporate accounting, will pay off.
In addition, an MBA curriculum is one of the very few locations where I can make a very diverse network. My fellow students will go into various differing industries, but as long as I remain in touch with them, they become my potential suppliers and clients. You can't buy that or build that spontaneously.
You may not feel the need for an MBA. In the meantime, I'll take my courses and my network.
Yes, I'm using my Karma Bonus; disregard my sig.
Excuse me? Learning how to effectively manage a corporation, and more importantly, learning how a corporate brain thinks is useless? I'm failing to see how the MBA will be useless to anyone with visions of starting a company, which is what went through my head when I first saw the /. article.
Different kinks for different people. he didn't attack your "fun" thing, don't attack his.