SharkyExtreme has a more lengthy writeup, including some initial performance comparisons from a prerelease version (chips clocked to 125Mhz instead of 143Mhz, beta drivers).
I still remember early Cyberpunk, and then the early years of Wired, as times of being exposed to one "mind blowing" idea after another. The future, though far from Utopian, was going to be very interesting. Anymore, though, I see few ideas that make me sit back and say "Whoa...now *that* is cool."
Now, in a mundane world of spam and banner ads, the coming future doesn't seem nearly so thrilling. In trying to pinpoint the source of my apathy about new ideas, I can't quite decide if it's me, us, or you. That is, I can't decide if: (a) My personal perspective has changed, and I've learned enough that little suprises me anymore, (b) We've all gotten better at predicting the future, so little surprises any of us, or (c) You folks (the SF writers and Futurists) blew out all the great ideas in the 80s and early 90s, and we'll just have to wait awhile for the next Big Thing.
So what I'm wondering is: Have you become at all jaded about technology and its effect on society? What do you think about our current state of predicting the future? Are there any ideas, authors, etc., that you've seen recently that make you say "Whoa..."?
I always find it amusing that whenever the U.S. doesn't go along with everyone else, it's "not-invented-here", "arrogance", we're "insular" or what have you. Whenever a European country doesn't go along it's "tradition", or "protecting their culture". Oh well, do what you like...
BTW, when are you Brits going to get with the program and drive on the right-hand side of the road like everyone else;)
I would *never* assume the other person knows what the units of measurement are. What is the context? Might you be talking centimeters, meters, or kilometers?
Any decent engineer always labels their units. That's what comment lines are for.
I think the AC is overstating it when he says the CIA doesn't even come close;)
However, you are definitely swimming with the sharks when dealing with a VC. They are rather notorious for stacking the board with their own people, then pushing you out. The classic Silicon Valley story is Cisco, and how the founders got tossed after bootstrapping the company from the days when they made routers in their living room. Sure the VC wants everyone to buy the product...they just don't want to give any of the money to you.
If you're that enamored of giving away your labor and ideas, why not just write free software? Otherwise, watch your back.
Well, imagine the belly laugh I got watching a local newscast one day: The police dept. of a town was giving away free copies of content filtering software for concerned parents (I don't remember if it was NetNanny, CyberSitter, whatever.)
After giving the obligatory "Protecting the Children" spiel, the not-so-tech-savvy officer proceeded to insert the CD-ROM into the 5" *floppy* slot on the ancient IBM clone in their office. He starts to flip the lever to lock down the "disk", but the scene cut just as you could see the first signs of resistance.
Made my whole day. Just wish I could see the out takes, though...
And what "addict" poses for a picture happily clutching the object of their obsession? I don't remember any of the tobacco victims cheerily taking a drag off a Marlboro. C'mon folks, at least go for the forlorn Look-what-they've-done-to-me sympathy shot.
What I want to know is who can I sue? All this rampant stupidity and greed is causing me great mental anguish...
From the posts here, it's obvious that people have good and bad experiences using both Cable and DSL.
I've had very good service using MediaOne cable in the Boston area. It's been fast, reliable, and they are largely agnostic about platforms/servers/home networks. Obviously some of the folks in Minnesota have seen the opposite.
On the DSL side, some of the customers are happy, and some are not. Big surprise.
People need to remember they are buying a service, not a technology. Both cable and DSL have technical issues that can be addressed, but it's up to the company providing it to you. The theoretical performance of the technology is irrelevant if it's being implemented by buffoons. Here in Boston, M1 seems to have done its homework, while Bell Atlantic still has work to do (and isn't very price competitive either). YMMV.
What I'm sure you know, but other/. readers may not, is that the stuff on the Air Chronicles site is not official Air Force doctrine. The site is intended for discussions and debate on doctrine, so obviously a good deal of it will not agree with whatever the party line happens to be. I had two articles published in Airpower Journal (which is hosted on the site) and they did not agree with existing AF doctrine -- nor did they change it much;)
That out of the way, I think some of this "cyberwar" stuff is overblown in the following sense: Lots of things that are support functions of real (i.e., "steel on target") warfare are also performed during peacetime, but that does not make them acts of war.
-- Countries try to break each other's ciphers all the time. Cracking encryption is not a hostile act by itself. -- A surveillance aircraft is flying up and down my coastline - in international waters. During a war I shoot it down. During peacetime I just have to suck it up. -- If someone jams your radar it doesn't give you immediate justification to shoot at them, though obviously it's not a very neighborly thing to do. In certain circumstances you might be justified (e.g., maybe the no-fly zone in Iraq). Again, it's not an act of war by itself.
I think cracking falls into the same category. Countries are going to try to break into each other's computer systems from here on out, but I don't think anyone will declare war over it. The exception would be if someone used access to an information system to muck up things in the physical world -- the oft-cited power grid example comes to mind -- but that's an act of sabotage.
As I stated in another post, when people talk about 32- or 64-bit chips, they are usually talking about the length of the instruction set.
The size of the data units manipulated can vary, but is usually at least as big as the instruction set. In the G4 the integer unit is 32 bits wide, the floating point unit is 64 bits wide, and the vector unit (a.k.a. Velocity Engine, AltiVec) is 128 bits wide. The intended use of the vector unit is not to manipulate quadruple precision floats (though it could), but to manipulate multiple pieces of data with one instruction (2 x 64, 4 x 32, or 8 x 16).
With 64-bit instructions you can create new operations, or more likely address more memory. Personally, I think the benefits of 64-bit instructions are somewhat oversold except for servers, which is exactly what all the 64-bit processors are targeted at.
To elaborate on what rugger said, the 32- versus 64-bit refers to the size of the instructions on the chip, which operate on chunks of data that may or may not be the same size.
The current PowerPC chips use 32-bit instructions. On the G4 the integer registers also happen to be 32-bit, while the floating point registers are 64-bit and the AltiVec (vector) registers are 128-bit. The point of AltiVec is not to be used for 128-bit wide chunks of data (though it could be) but to instead manipulate several floats or ints at the same time (2 x 64, 4 x 32, or 8 x 16).
The advantages of 64-bit instructions would be: a) adding new and different instructions, b) addressing more memory, or c) a little bit of both. I haven't seen the 64-bit instructions, so I'm not sure which route they took. Even in the current PowerPC there is no shortage of instructions, and most programs only use a portion of them, so I'm not sure there is much need to create more. On the other hand, addressing gobs and gobs of memory is a definite benefit for servers.
Why keep making 32-bit versions? As others have noted, if you don't need to address all that memory then there is little real world benefit. For embedded and desktop systems this may well be the case. Again for servers a different story. Also, the bigger and more varied the instruction set, the more physical space it takes up on the chip. This space could be used instead for more registers or multiple cores, both of which might be more handy for most users.
Gotta admit I'm a little confused here. If you have a computationally intensive task, why would you ever want to run it through any VM or interpreter? Granted, LISP may run through an interpreter during the development phase, but you can always compile it to get better speed in the final product.
Can you elaborate without giving away the company secrets?
The possession of any of the items mentioned -- including guns -- does not deprive anyone else of life, liberty, or the pursuit of happiness. The action of murdering someone obviously deprives them of all three.
The two are fundamentally different in nature. Their possessing an item infringes on no one's rights, their using it unlawfully is a totally diferent matter.
Well, this reply is a little late since I haven't been able to access/. all day.
Basically, 'well respected' carries the connotation of at least grudging respect across the whole audience in question. I would submit that 'controversial' is a more apt description since Dvorak elicits such varying responses from people. As for me personally, I've found some columns of his fairly insightful, while others were downright moronic (and this dates well before the iBook flap) -- in my book (no pun intended) 'well respected' would indicate a consistent level of quality which I think he lacks.
Within the Linux community, Linus is 'well respected' while RMS might be described as 'controversial'. Among those familiar with Apple, Woz might be described as 'well respected' while Jobs would be 'controversial'.
I don't necessarily mean controversial in a perjorative sense -- it may well turn out that the 'controversial' person is right in the long run and the 'well respected' person just didn't make waves. It just didn't strike me that 'well respected' was accurate in describing Dvorak's perception in just about any community, because he does tend to polarize people.
...before they were all privately held. Now the average investor can buy a politician! Why shouldn't the local ladies' investor club get a shot at some legalized graft? Why should only the Chinese get to buy and sell shares of Gore? Why shouldn't the small investor be able to buy and sell Bush on p*Trade?
Gotta disagree with you there. The brain -- the Mother of All Neural Networks -- can store many overlapping patterns and still be able to recall them distinctly. A memory is not stored in a specific place, but instead distributed throughout a web of connections. A single brain cell may be involved with the recollection of many memories.
The ease of recall is determined by the strength of the dendritic connections between the brain cells that make up the pattern -- how strongly the pattern is "burned in". This is the point of rehearsals and rereading important items. With more patterns overlapping, you may be more likely to make connections between seemingly disparate topics, but you should still be able to distinguish them. If two *weak* patterns overlap (items that haven't been recalled in awhile) it might be possible to confuse them.
If the volume of data were primary factor, wouldn't teenagers have better recall than twenty-five year olds? I think we'd all agree that a 25-year old remembers just as well or better than a 16-year old. So why does mental performance decline over time? Well, aren't the late 20's when you are no longer forced to learn new things?
Without anyone forcing you to learn new things, you're on your own. If you keep learning, mental performance should actually improve until serious brain cell degradation sets in -- I guess this is what these researchers are trying to reverse. On the other hand, if you learn nothing new the patterns start to atrophy.
Anyway, that's my $.02, any neuroscientists care to weigh in?
..if you are going to get rid of perfectly good G3s *and* the people who use them that's downright idiotic. There's no more delicate way to say it.
Have you put actual thought into what you just said? You don't like that the B&W G3's are *currently* not upgradeable, so you're going to dump them and go to NT and:
1) Get machines that are no faster than the G3. Why not spend the money for a new G4? Upgrade cards weren't going to be cheap for a long time anyway -- just move any peripherals from the B&Ws to a *much* better Sawtooth motherboard.
2) Incur the disruption, extra training and support costs of switching.
3) Buy all new software licenses.
4) Watch your IT department grow into it's own little empire.
5) Jump right into the teeth of the Y2K problem right before the rollover date. Good move.
Do you honestly believe you'll get better productivity? Have you read *any* TCO or ROI studies that support this? And you'll do it even if your people don't like it? Guess you don't give a rat's ass about "your" people, do you? I'm glad as hell I don't work for you.
Did they say "all"? No. "Most" flat panels still use analog inputs. As for price, it's mostly due to the size of the display. Another poster pointed out that the SGI is only $2300, but at 17" it's still a full 5 inches smaller. Try a little perspective next time.
Slashdot Mirror
..then we'd have to call it 1-yard resolution. Eh, it's close enough ;)
SharkyExtreme has a more lengthy writeup, including some initial performance comparisons from a prerelease version (chips clocked to 125Mhz instead of 143Mhz, beta drivers).
Bruce-
I still remember early Cyberpunk, and then the early years of Wired, as times of being exposed to one "mind blowing" idea after another. The future, though far from Utopian, was going to be very interesting. Anymore, though, I see few ideas that make me sit back and say "Whoa...now *that* is cool."
Now, in a mundane world of spam and banner ads, the coming future doesn't seem nearly so thrilling. In trying to pinpoint the source of my apathy about new ideas, I can't quite decide if it's me, us, or you. That is, I can't decide if: (a) My personal perspective has changed, and I've learned enough that little suprises me anymore, (b) We've all gotten better at predicting the future, so little surprises any of us, or (c) You folks (the SF writers and Futurists) blew out all the great ideas in the 80s and early 90s, and we'll just have to wait awhile for the next Big Thing.
So what I'm wondering is: Have you become at all jaded about technology and its effect on society? What do you think about our current state of predicting the future? Are there any ideas, authors, etc., that you've seen recently that make you say "Whoa..."?
I always find it amusing that whenever the U.S. doesn't go along with everyone else, it's "not-invented-here", "arrogance", we're "insular" or what have you. Whenever a European country doesn't go along it's "tradition", or "protecting their culture". Oh well, do what you like...
;)
BTW, when are you Brits going to get with the program and drive on the right-hand side of the road like everyone else
I would *never* assume the other person knows what the units of measurement are. What is the context? Might you be talking centimeters, meters, or kilometers?
Any decent engineer always labels their units. That's what comment lines are for.
I think the AC is overstating it when he says the CIA doesn't even come close ;)
However, you are definitely swimming with the sharks when dealing with a VC. They are rather notorious for stacking the board with their own people, then pushing you out. The classic Silicon Valley story is Cisco, and how the founders got tossed after bootstrapping the company from the days when they made routers in their living room. Sure the VC wants everyone to buy the product...they just don't want to give any of the money to you.
If you're that enamored of giving away your labor and ideas, why not just write free software? Otherwise, watch your back.
Well, imagine the belly laugh I got watching a local newscast one day: The police dept. of a town was giving away free copies of content filtering software for concerned parents (I don't remember if it was NetNanny, CyberSitter, whatever.)
After giving the obligatory "Protecting the Children" spiel, the not-so-tech-savvy officer proceeded to insert the CD-ROM into the 5" *floppy* slot on the ancient IBM clone in their office. He starts to flip the lever to lock down the "disk", but the scene cut just as you could see the first signs of resistance.
Made my whole day. Just wish I could see the out takes, though...
And what "addict" poses for a picture happily clutching the object of their obsession? I don't remember any of the tobacco victims cheerily taking a drag off a Marlboro. C'mon folks, at least go for the forlorn Look-what-they've-done-to-me sympathy shot.
What I want to know is who can I sue? All this rampant stupidity and greed is causing me great mental anguish...
From the posts here, it's obvious that people have good and bad experiences using both Cable and DSL.
I've had very good service using MediaOne cable in the Boston area. It's been fast, reliable, and they are largely agnostic about platforms/servers/home networks. Obviously some of the folks in Minnesota have seen the opposite.
On the DSL side, some of the customers are happy, and some are not. Big surprise.
People need to remember they are buying a service, not a technology. Both cable and DSL have technical issues that can be addressed, but it's up to the company providing it to you. The theoretical performance of the technology is irrelevant if it's being implemented by buffoons. Here in Boston, M1 seems to have done its homework, while Bell Atlantic still has work to do (and isn't very price competitive either). YMMV.
As far as I'm concerned, there are plenty bigger fish to fry:
-- Suitcase nukes
-- Poison in the water supply
-- Bio/chem weapons (e.g., Sarin + Cropduster = Really Bad Day)
Compared to any of the above, go right ahead and knock out my power and empty my bank account -- I've got plenty of books to read.
What I'm sure you know, but other /. readers may not, is that the stuff on the Air Chronicles site is not official Air Force doctrine. The site is intended for discussions and debate on doctrine, so obviously a good deal of it will not agree with whatever the party line happens to be. I had two articles published in Airpower Journal (which is hosted on the site) and they did not agree with existing AF doctrine -- nor did they change it much ;)
That out of the way, I think some of this "cyberwar" stuff is overblown in the following sense: Lots of things that are support functions of real (i.e., "steel on target") warfare are also performed during peacetime, but that does not make them acts of war.
-- Countries try to break each other's ciphers all the time. Cracking encryption is not a hostile act by itself.
-- A surveillance aircraft is flying up and down my coastline - in international waters. During a war I shoot it down. During peacetime I just have to suck it up.
-- If someone jams your radar it doesn't give you immediate justification to shoot at them, though obviously it's not a very neighborly thing to do. In certain circumstances you might be justified (e.g., maybe the no-fly zone in Iraq). Again, it's not an act of war by itself.
I think cracking falls into the same category. Countries are going to try to break into each other's computer systems from here on out, but I don't think anyone will declare war over it. The exception would be if someone used access to an information system to muck up things in the physical world -- the oft-cited power grid example comes to mind -- but that's an act of sabotage.
I stand corrected. Thx.
As I stated in another post, when people talk about 32- or 64-bit chips, they are usually talking about the length of the instruction set.
The size of the data units manipulated can vary, but is usually at least as big as the instruction set. In the G4 the integer unit is 32 bits wide, the floating point unit is 64 bits wide, and the vector unit (a.k.a. Velocity Engine, AltiVec) is 128 bits wide. The intended use of the vector unit is not to manipulate quadruple precision floats (though it could), but to manipulate multiple pieces of data with one instruction (2 x 64, 4 x 32, or 8 x 16).
With 64-bit instructions you can create new operations, or more likely address more memory. Personally, I think the benefits of 64-bit instructions are somewhat oversold except for servers, which is exactly what all the 64-bit processors are targeted at.
To elaborate on what rugger said, the 32- versus 64-bit refers to the size of the instructions on the chip, which operate on chunks of data that may or may not be the same size.
The current PowerPC chips use 32-bit instructions. On the G4 the integer registers also happen to be 32-bit, while the floating point registers are 64-bit and the AltiVec (vector) registers are 128-bit. The point of AltiVec is not to be used for 128-bit wide chunks of data (though it could be) but to instead manipulate several floats or ints at the same time (2 x 64, 4 x 32, or 8 x 16).
The advantages of 64-bit instructions would be: a) adding new and different instructions, b) addressing more memory, or c) a little bit of both. I haven't seen the 64-bit instructions, so I'm not sure which route they took. Even in the current PowerPC there is no shortage of instructions, and most programs only use a portion of them, so I'm not sure there is much need to create more. On the other hand, addressing gobs and gobs of memory is a definite benefit for servers.
Why keep making 32-bit versions? As others have noted, if you don't need to address all that memory then there is little real world benefit. For embedded and desktop systems this may well be the case. Again for servers a different story. Also, the bigger and more varied the instruction set, the more physical space it takes up on the chip. This space could be used instead for more registers or multiple cores, both of which might be more handy for most users.
Gotta admit I'm a little confused here. If you have a computationally intensive task, why would you ever want to run it through any VM or interpreter? Granted, LISP may run through an interpreter during the development phase, but you can always compile it to get better speed in the final product.
Can you elaborate without giving away the company secrets?
Your argument is an illogical strawman.
The possession of any of the items mentioned -- including guns -- does not deprive anyone else of life, liberty, or the pursuit of happiness. The action of murdering someone obviously deprives them of all three.
The two are fundamentally different in nature. Their possessing an item infringes on no one's rights, their using it unlawfully is a totally diferent matter.
It's been awhile since I read it, but as I recall Greg Bear's Queen of Angels was excellent. Nano is just one of several themes in the book.
Though not exactly about nano, his Blood Music is also very good.
"A collection of 10,000 nuts and bolts trying desparately to get away from one another."
Well, this reply is a little late since I haven't been able to access /. all day.
Basically, 'well respected' carries the connotation of at least grudging respect across the whole audience in question. I would submit that 'controversial' is a more apt description since Dvorak elicits such varying responses from people. As for me personally, I've found some columns of his fairly insightful, while others were downright moronic (and this dates well before the iBook flap) -- in my book (no pun intended) 'well respected' would indicate a consistent level of quality which I think he lacks.
Within the Linux community, Linus is 'well respected' while RMS might be described as 'controversial'. Among those familiar with Apple, Woz might be described as 'well respected' while Jobs would be 'controversial'.
I don't necessarily mean controversial in a perjorative sense -- it may well turn out that the 'controversial' person is right in the long run and the 'well respected' person just didn't make waves. It just didn't strike me that 'well respected' was accurate in describing Dvorak's perception in just about any community, because he does tend to polarize people.
Love their web site. Who designed their home page, Magritte?
btw, Dvorak may be well known, but 'well respected'?
...before they were all privately held. Now the average investor can buy a politician! Why shouldn't the local ladies' investor club get a shot at some legalized graft? Why should only the Chinese get to buy and sell shares of Gore? Why shouldn't the small investor be able to buy and sell Bush on p*Trade?
Who needs straw polls when you can have IPOs?!
I say bring on Jesse "The P/E Ratio" Ventura!
Gotta disagree with you there. The brain -- the Mother of All Neural Networks -- can store many overlapping patterns and still be able to recall them distinctly. A memory is not stored in a specific place, but instead distributed throughout a web of connections. A single brain cell may be involved with the recollection of many memories.
The ease of recall is determined by the strength of the dendritic connections between the brain cells that make up the pattern -- how strongly the pattern is "burned in". This is the point of rehearsals and rereading important items. With more patterns overlapping, you may be more likely to make connections between seemingly disparate topics, but you should still be able to distinguish them. If two *weak* patterns overlap (items that haven't been recalled in awhile) it might be possible to confuse them.
If the volume of data were primary factor, wouldn't teenagers have better recall than twenty-five year olds? I think we'd all agree that a 25-year old remembers just as well or better than a 16-year old. So why does mental performance decline over time? Well, aren't the late 20's when you are no longer forced to learn new things?
Without anyone forcing you to learn new things, you're on your own. If you keep learning, mental performance should actually improve until serious brain cell degradation sets in -- I guess this is what these researchers are trying to reverse. On the other hand, if you learn nothing new the patterns start to atrophy.
Anyway, that's my $.02, any neuroscientists care to weigh in?
...just see if he is blinking S.O.S. during the broadcast. Oh wait, his eyelids don't move, do they? Hmm..probably why they didn't pick Big Bird.
..if you are going to get rid of perfectly good G3s *and* the people who use them that's downright idiotic. There's no more delicate way to say it.
Have you put actual thought into what you just said? You don't like that the B&W G3's are *currently* not upgradeable, so you're going to dump them and go to NT and:
1) Get machines that are no faster than the G3. Why not spend the money for a new G4? Upgrade cards weren't going to be cheap for a long time anyway -- just move any peripherals from the B&Ws to a *much* better Sawtooth motherboard.
2) Incur the disruption, extra training and support costs of switching.
3) Buy all new software licenses.
4) Watch your IT department grow into it's own little empire.
5) Jump right into the teeth of the Y2K problem right before the rollover date. Good move.
Do you honestly believe you'll get better productivity? Have you read *any* TCO or ROI studies that support this? And you'll do it even if your people don't like it? Guess you don't give a rat's ass about "your" people, do you? I'm glad as hell I don't work for you.
Did they say "all"? No. "Most" flat panels still use analog inputs. As for price, it's mostly due to the size of the display. Another poster pointed out that the SGI is only $2300, but at 17" it's still a full 5 inches smaller. Try a little perspective next time.