According to the BP statistical review, June 2002, global consumption of liquid fossil fuels comes to 5595 KBbl/diem gasoline, 9247 KBbl/d. kerosene, 4873 KBbl/d. fuel oil, or 264, 435, and 230 MT/an, respectively, for a total of ~929 MT/an. The remainder of liquid fossil fuel production is consumed by manufacture of materials or consists of loss. Accepting BPs loss estimates, and assuming all losses are gassified, that's 220 MT/an.
Coal consumption is 71.0% and natural gas is 60.0% oil equivalent. To be generous, I include production and refining losses to get a total global annual carbon injection of
(1.00+0.600+0.710) * (220+264+435+230) MT which comes to 2654 MegaTonnes annually, or less than 443 Kg per person, annually. This represents 90% carbon, which is 12/44 of C02, for a total CO2 injection into the carbon cycle of 1.46 metric tonnes per annum per capita, or 8,750 MT/an total.
As you say, Krakatoa might conceivably have emitted a few thousand megatons, but the human emissions at that time were vanishingly small in comparison to their current levels, so that the eruption probably injected more CO2 than all human activity during the *preceeding* century, but in my estimation certainly injected an order of magnitude less than the human activity during the *following* century.
Perhaps it was equivalent to a century of human injection at the rate prevailing at the time the original estimation was made, and this statement was later carried forward, and quoted on slashdot, long after it was no longer accurate.
Refering to Scientific American as a benchmark of scholarly consensus is not helpful. Since the late 80s SA has devolved into a bully political pulpit, thanks to changes in editorship.
I personally don't care much about the issue, because it seems obvious that
1) There is humanly eventuated global warming,
2) which is not very significant in comparison
to the warming effect of continuing changes
in the climate due to non-human causes,
3) and it will certainly create social disruptions
due to redistribution of agricultural
fertility and rising sea levels over the
course of centuries, but
4) those disruptions will benefit some, and
disadvantage others.
Change is inevitable, and whether it is caused by human agency or non-human agency is moot, really. The issues are whether or not it can or should be resisted or assisted, and if so, in what ways.
Certainly the evacuation of various populated atolls and low-lying regions such as the river delta in Bangladesh will be required during this century. Focussing more energy on how to accomplish this with minimal loss of life and less on who has the bigger scientific penis might be productive.
The sole reason why my employer dropped my MSDN universal is that the money was going to Microsoft. Now I get an Alienware laptop every year instead, so I'm MUCH happier.
I understand that your comment was ironic, and I do appreciate the irony. However, many readers may take it seriously. Such persons should be alerted that "anti-virus" software also protects against worms which spread through e-mail, and that shrink- wrapped software often contains viruses, for the simple reason that the computers used to produce that software are not inherently immune to infection. Why, even Microsoft security updates have been known to come pre-infected with viruses.
I have yet to see a Unix/Linux/BSD/Solaris/Mach/OS-X virus in the wild, however.
In Europe, I could get off a plane with nothing but a passport and a handful of cash and *manufacture* a submachine gun from freely downloadable CNC data and steel stock in less than 24 hours.
In any major city of Europe, I would be willing to wager, oh, 3 months salary ($30,000 US) that with enough cash in my pocket I can purchase a handgun within 24 hours of arrival.
If I really wanted a weapon quickly, I'd just garrote a police officer.
What you really mean is that you are not competent to aquire a firearm, and that none of your friends has divulged to you that they possess firearms.
The intelligence of a corporation can be computed by adding the individual IQs of its officers and dividing by the number of employees.
Re:There's more to it than 64-bit instructions
on
Is Prescott 64-bit?
·
· Score: 1
Actually, userspace code can access 36 bits of memory. It just has to span multiple protection domains in order to do it.
I once worked on a compiler that could be used to generate code to do this. It was a multi-million $ project, although with current tech, I'm sure I could approximate it fairly well for under $100K.
36 bits is a measly 64GB. That's about $10K these days. I think it needed to change a few years ago.
What would *you* do if you could directly address a terabyte or two of RAM? Post-stack migration? QCD? MBR time-sequence prediction? This stuff that used to take supercomputers becomes feasible on COTS hardware.
No, screw you and your astroturfing. Screw your disdain for the superiority of architectures that can't run CP/M-86. Intel is going to get their lunch eaten by x86 because its faster * cheaper, and beats them at their own back-compatibility game.
I'd give Intel 15 years before its out of the desktop CPU business.
Its also what makes it sheer hell to write a compiler for the thing. With a hand-rolled (or should I say *unrolled*) mult-add loop, you can crank out mondo flops in BLAS, but when you try to run a spreadsheet or a non-linear program, you see the pathetic weakness of the compilers start to set in.
Itanium was vapor for *years* after I got my first Sparc-9 64-bit desktop box. Heck, it's still vapor. What did Dell ship last quarter? About 800 Itanium servers? Woo-hoo! With economies of scale like that, who needs mainframes?
If you have the desk space and a hundred bucks these days, you can get a printer/scanner/fax/copier all-in-one. But you have to refill the cartridges if you want to pay a fair price for printing.
If you think that *any* security measure defends against *all* attacks, you need to reconsider. A security meansure is something that improves the security of a system. NAT does that.
Slashdot Mirror
> Correlation != Causation
Does too.
That's no refutation. It was clearly known in
advance without any reasonable doubt that Iraq
contained trillions of dollars worth of oil.
According to the BP statistical review, June 2002,
global consumption of liquid fossil fuels comes to
5595 KBbl/diem gasoline, 9247 KBbl/d. kerosene,
4873 KBbl/d. fuel oil, or 264, 435, and 230 MT/an,
respectively, for a total of ~929 MT/an. The
remainder of liquid fossil fuel production is
consumed by manufacture of materials or consists
of loss. Accepting BPs loss estimates, and assuming
all losses are gassified, that's 220 MT/an.
Coal consumption is 71.0% and natural gas is 60.0%
oil equivalent. To be generous, I include
production and refining losses to get a total
global annual carbon injection of
(1.00+0.600+0.710) * (220+264+435+230) MT
which comes to 2654 MegaTonnes annually, or
less than 443 Kg per person, annually.
This represents 90% carbon, which is 12/44 of
C02, for a total CO2 injection into the carbon
cycle of 1.46 metric tonnes per annum per capita,
or 8,750 MT/an total.
As you say, Krakatoa might conceivably have
emitted a few thousand megatons, but the human
emissions at that time were vanishingly small
in comparison to their current levels, so that
the eruption probably injected more CO2 than all
human activity during the *preceeding* century,
but in my estimation certainly injected an order
of magnitude less than the human activity during
the *following* century.
Perhaps it was equivalent to a century of human
injection at the rate prevailing at the time
the original estimation was made, and this statement
was later carried forward, and quoted on slashdot,
long after it was no longer accurate.
Refering to Scientific American as a benchmark
of scholarly consensus is not helpful. Since the
late 80s SA has devolved into a bully political
pulpit, thanks to changes in editorship.
I personally don't care much about the issue,
because it seems obvious that
1) There is humanly eventuated global warming,
2) which is not very significant in comparison
to the warming effect of continuing changes
in the climate due to non-human causes,
3) and it will certainly create social disruptions
due to redistribution of agricultural
fertility and rising sea levels over the
course of centuries, but
4) those disruptions will benefit some, and
disadvantage others.
Change is inevitable, and whether it is caused by
human agency or non-human agency is moot, really.
The issues are whether or not it can or should be
resisted or assisted, and if so, in what ways.
Certainly the evacuation of various populated
atolls and low-lying regions such as the river
delta in Bangladesh will be required during
this century. Focussing more energy on how to
accomplish this with minimal loss of life and
less on who has the bigger scientific penis
might be productive.
If the computer is not trusted, you can't use
GPG, period. This is an incredibly stupid excuse
to make crypto unusuable.
Actually, it is under-hyped.
I did a hypometric regression on late 90s
technologies, and XML was 27% below the
weighted mean.
The sole reason why my employer dropped my MSDN
universal is that the money was going to Microsoft.
Now I get an Alienware laptop every year instead,
so I'm MUCH happier.
OpenOffice is not a replacement for MSOffice.
StarOffice is.
How can I upgrade without updating iTunes?
I don't want to lose sharing capabilities
to upgrade the OS.
I understand that your comment was ironic, and I
do appreciate the irony. However, many readers may
take it seriously. Such persons should be alerted
that "anti-virus" software also protects against
worms which spread through e-mail, and that shrink-
wrapped software often contains viruses, for the
simple reason that the computers used to produce
that software are not inherently immune to
infection. Why, even Microsoft security updates
have been known to come pre-infected with viruses.
I have yet to see a Unix/Linux/BSD/Solaris/Mach/OS-X
virus in the wild, however.
In Europe, I could get off a plane with nothing
but a passport and a handful of cash and *manufacture*
a submachine gun from freely downloadable CNC data
and steel stock in less than 24 hours.
In any major city of Europe, I would be willing
to wager, oh, 3 months salary ($30,000 US)
that with enough cash in my pocket I can purchase
a handgun within 24 hours of arrival.
If I really wanted a weapon quickly, I'd just
garrote a police officer.
What you really mean is that you are not competent
to aquire a firearm, and that none of your friends
has divulged to you that they possess firearms.
Evidently, you are not Swiss.
The intelligence of a corporation can be computed
by adding the individual IQs of its officers
and dividing by the number of employees.
Actually, userspace code can access 36 bits of
memory. It just has to span multiple protection
domains in order to do it.
I once worked on a compiler that could be used to
generate code to do this. It was a multi-million
$ project, although with current tech, I'm sure
I could approximate it fairly well for under
$100K.
36 bits is a measly 64GB. That's about $10K
these days. I think it needed to change a few
years ago.
What would *you* do if you could directly
address a terabyte or two of RAM?
Post-stack migration? QCD? MBR time-sequence
prediction? This stuff that used to take
supercomputers becomes feasible on COTS
hardware.
Actually, they lowered prices.
The drop is prices was larger than the fee.
> can already address 36 bits
umm... you got a compiler that will support that?
I didn't think so.
It would be fun to see you write code that
loops over 1G of 64-bit floating point numbers.
No, I take that back -- I like you:P
No, screw you and your astroturfing. Screw your
disdain for the superiority of architectures that
can't run CP/M-86. Intel is going to get their
lunch eaten by x86 because its faster * cheaper,
and beats them at their own back-compatibility game.
I'd give Intel 15 years before its out of the
desktop CPU business.
Its also what makes it sheer hell to write a
compiler for the thing. With a hand-rolled
(or should I say *unrolled*) mult-add loop,
you can crank out mondo flops in BLAS, but
when you try to run a spreadsheet or a non-linear
program, you see the pathetic weakness of the
compilers start to set in.
Itanium was vapor for *years* after I got my
first Sparc-9 64-bit desktop box. Heck, it's
still vapor. What did Dell ship last quarter?
About 800 Itanium servers? Woo-hoo! With
economies of scale like that, who needs mainframes?
If you have the desk space and a hundred bucks
these days, you can get a printer/scanner/fax/copier
all-in-one. But you have to refill the cartridges
if you want to pay a fair price for printing.
If you think that *any* security measure defends
against *all* attacks, you need to reconsider.
A security meansure is something that improves the
security of a system. NAT does that.
> ...should...
Why? Bald assertions alone carry little weight.
This does not appear to be free software.
Can you clarify, what is the license?
I have designed a pill with no side-effects.
It consists of a gelatin capsule with nothing
in it.
> Correlation is not causation.
Is too.
There should also be automatic disabling of unused
services. E.g. lpd, if it doesn't get used for a
period of, say, 24 days, should disable itself.