So they threw out the whole pedestal computer concept just so they could use a new kind of LCD?
Oh no, I was referring to the transition from the old G3-based iMacs to the iMac G4 at that point. I guess I might have been a bit ambiguous there. By that time, Apple had developed expertise in LCD technology and decided to apply it to the iMac line. Losing the bulky CRT meant the old case no longer made any sense, and after some head-scratching, the pedestal computer was born.
Look, all I am trying to say is that it's more than a question of design versus branding. Any decision has to be tempered by the reality of what is possible with the technology at hand. Where the iMac is concern, I feel that Apple engineers have thus far done about as much as they could within the limits set out for them. They have certainly done a lot more than any other PC manufacturer.
Okay, I re-read your original post carefully. I see your point in that you are not one who cares about aesthetics. I apologize for implying so.
What had originally caught my attention was your contention that even Apple has lost interest in the idea [of a pedestal computer], that it's all about branding these days, not usability and that once familiarity has blunted the coolness factor, an iMac design is discarded -- no matter how good it is.
Now, I realize this is just one opinion versus another, but I don't see it that way. As far as the iMac line is concerned, Apple only seems to make radical changes when engineering obstacles necessitate them. I have yet to see a case in which upper management threw out a good design on a whim, thinking it wasn't cool enough anymore. For one thing, it costs a lot to start from scratch every time.
They kept the original CRT enclosure for years, and it still lives on, more or less, in the eMac. The iMac G4 had to be redesigned because the LCD display was a totally different shape. The iMac G5 had to be redesigned because they couldn't fit a PPC970 with all its cooling requirements into a tight little box. Were that possible, we would probably be seeing PowerBook G5s and maybe even Mac mini G5s by now, but I'm not holding my breath.
There is more to design than just aesthetics. Form has to follow function. In the case of the iMac G5, heat dissipation was a much bigger issue than with the G4. A tightly packed spherical CPU is just about the worst possible design for getting rid of heat. I can see why Apple engineers would have had to abandon it whether they wanted to or not.
Put yourself in their shoes for a moment. They're thinking, "The best way to get rid of heat is to build a thin enclosure with lots of surface area. Orient it vertically and we can even take advantage of convective cooling. So how can we achieve this with minimal impact to our existing popular iMac design? Hmm..."
FWIW, I think it was a stroke of genius what they did. The question now is whether they can come up with a similarly clever solution to the PowerBook G5 dilemma?
Someone was telling me Apple has gone through and updated most of the file-handling tools to honour HFS+ meta-data. Is this true? Is there a list someplace of which ones they worked on?
Also, is there any tool under Tiger that can query for Rendezvous/Bonjour services? I couldn't find anything like that in Panther, though maybe I wasn't looked hard enough.
Seriously, can any subscription model stand the test of time from a purely technological standpoint? Recent events have shown that Apple's DRM scheme is pretty easy to circumvent, but then they don't have to worry about every user suddenly owning a million songs when that happens.
How would Napster thwart such a dam burst? Once the DRM is compromised, it's only a matter of time before someone comes along with a server which can dish out unprotected songs to the world by request. You wouldn't even have to waste your own disk space!
It surprises me that the labels are always promoting subscription services. I'd have thought this would be their worst nightmare.
This might not be an issue for a game console, but for a workstation, wouldn't the Cell's context-switching overhead be rather huge? From what I have read so far, it seems each SPE has 256 KB of local storage, plus another 2 KB for the main register file (128 16-byte registers) and whatever other state information it needs. Since there are 8 of these things, we're talking over 2 MB of state to swap in and out. Would that still be considered a drop in the bucket for most scheduling schemes?
Well, this is the way I envisioned it would work. Alice would register with a CA, securely submitting a random key which could be used for future authentication. Once the CA is satisfied that it has established Alice's identity, the key would go into its database. Bob would do likewise.
When Alice wants to talk to Bob, she first sends a random sequence to the CA to be relayed to Bob. (The CA will assert to Bob that the message came from Alice.) She also sends the same sequence to Bob directly over a secure channel she has set up with him, and Bob simply compares the two copies he receives to see if they match.
If it turns out that there are hash functions which are resistent to quantum cryptoanalysis, you could simplify the scheme further so that Alice just sends a signed message to Bob and Bob asks the CA to verify it.
You're right that there is a weakness here, however, in that the CA would own a key with which they could impersonate Alice. The whole point is that you are supposed to trust your CA, but that certainly does raise the burden of trust.
The problem is that random noise superimposed on data is still random noise as far as the eavesdropper is concerned. You might as well just take a stab at the message itself rather than bother trying to guess the OTP. Given the length of the message and its source and destination, you might be able to make a reasonable stab at it, though even those things can be obfuscated.
No public key scheme can establish the absolute identity of the remote party in the absence of some prior relationship, even if it is through a shared acquaintence. That's why we need certificate authorities to vouch for secure web sites. What they can do is verify that later transmissions are coming from the same source as the original.
Say Bob runs a rumor site. An anonymous tipster could establish a secure connection with him to leak some confidential data. Bob may choose to sit on it until he knows it checks out, but if it does and the tipster keeps throwing more his way, he would be more inclined to trust it because he can verify the sender.
At this point, he still doesn't know that the tipster was Alice. Alice could reveal who she is, but Bob would still need a third party to verify that her signature is genuine. I can't see anything in quantum cryptography which would be any different as far as that's concerned.
While I doubt quantum cryptography could ever really take off in the public Internet what with the packets changing hands so many times on route, I was wondering how applicable it might be to wireless LANs? Is it even possible to demonstrate it at microwave frequencies? Will the day come when I can utter the words "secure WI-FI" without an involuntary chuckle?
We use a crystal clock in geophysical equipment whose drift is on the order of 1 part in 10^11 if I have my numbers right. That's about an order of magnitude better than the atomic clock. However, this clock is ovenized (maintained at a constant 70 degrees C) and that helps greatly. I wonder if you can boost the accuracy of the atomic clock in a similar fashion?
This sort of idea makes several assumptions I am not comfortable with:
1. Network bandwidth will no longer be the bottleneck.
2. Your net connection will be more reliable than your hard drive.
3. Your peripherals will work with any computer ever made.
4. You will be equally productive using a variety of different monitors, keyboards, etc.
5. Your activity on someone else's terminal will not be monitored.
Take number 1, for example. Local storage has been expanding at an astounding rate, well ahead of Moore's Law. Internet bandwidth has been improving in a non-uniform and often non-symmetrical pattern. I can only imagine applications in the future working with ever larger data sets, so I am not very optimistic about a net-driven approach. Something like this might work within an organization where you can control many more variables (though I would expect some resistence even then), but I think it will flop bigtime if they try to shift the paradigm for everyone.
Off hand, I can think of three reasons why you might leave your WI-FI unsecured:
1. You are an ignoramus and it never even occurred to you that someone in the neighbourhood could cause you a lot of grief.
2. You are aware of the security issues and have taken some precautions to secure your LAN, but see no harm in allowing others to share your Internet connection. Maybe you actually trust your neighbours...now there's a thought!
3. You are a predator who lures the freeloaders in with a wide open hub. Then you wait around with the packet sniffer running until some sucker shows up and crack your way into his system. The knife cuts both ways, after all.
Articles like this assume everyone is in category 1, but how can you be sure?
Re:The electricity still comes from fossil fuels!
on
Hybrid Fleet Vehicles
·
· Score: 1
Um...forgive me if I'm wrong, but I thought we were talking about hybrid vehicles here. They do not charge off the grid. They use a conventional gasoline/deisel engine to charge up the battery. They are quite fuel efficient, however, since the engine can be kept small. It only needs to handle average power demands, and in a vehicle which stops and goes a lot, this could potentially spell enormous savings. Postal vans would be an ideal application. They might also consider garbage trucks.
Yes, that struck me as a bit odd too. I'm pretty sure they meant to say that methane is 21 times more potent a greenhouse gas than carbon dioxide.
That said, I do remember from my university days studying climatology that there was some speculation that global warming could thin the ozone layer to a degree. I think the idea was that stratospheric cooling would slow down ozone formation. (Heat trapped in the lower level troposphere can actually cause higher levels of the atmosphere to drop in temperature.)
Anyway, that theory is at least a decade old by now, and I don't know if newer computer models have since proven or disproven it.
Knowing that Apple's entire line is now G4s or better has made my day! I no longer have to steer clients away from the iBook because our Altivec-tuned modelling engine will crawl on it. I hope more developers will use Altivec now that they know every Mac will support it. Man, I haven't felt this good since Apple moved to PowerPCs and every machine from then on had an FPU!
1. Textbook patches will be released biweekly. Application of the patches is mandatory.
2. When the blackboard suddenly turns blue, students must leave the classroom in an orderly fashion and return to their seats after ten minutes. No explanation will be given.
3. An alarm bell will sound to signal a massive virus outbreak or worm infestation at Microsoft High. Students are required to calmly exit the building. No drills have been scheduled for this procedure, as it is believed the bell will ring frequently throughout the term without them.
4. Visits to Open Source High are stictly forbidden. Students are, however, encouraged to visit other area schools and report any smaller, well-run institutions with innovative programs to expedite their hostile acquisition by the Microsoft School Board.
5. Our MSSAT exam is similar to--though subtly incompatible with--its government counterpart.
6. Please do not be alarmed by the video portraits of Bill Gates whose eyes follow you down every hall. He got the idea from reading Harry Potter.
I am partially color-blind myself, and have difficulty distinguishing iCal entries by calendar color due to the particular shades Apple has chosen. Anyone know how to tweak these?
-Ted
P.S. While I'm at it, I woudn't mind hearing about any trick for inverting the color scheme at arstechnica back to something sensible!
I live near the Thousand Islands myself. My dream vehicle used to be a personal hovercraft, but now I'm not so sure. I could have a lot of fun with that car!
Hasn't this automated fast food concept been tried before, like in my dad's generation? Sure, the automats doubtless had some poor schmuck in the back serving stuff up, but to the customer, it was an automated vending-machine-on-steroids-type interface.
I imagine the automats died a natural death for cultural -- rather than technological -- reasons, which is why I am skeptical about whether the mechanical fast food industry will ever rise again. Well...maybe in Japan.:-)
Slashdot Mirror
Oh no, I was referring to the transition from the old G3-based iMacs to the iMac G4 at that point. I guess I might have been a bit ambiguous there. By that time, Apple had developed expertise in LCD technology and decided to apply it to the iMac line. Losing the bulky CRT meant the old case no longer made any sense, and after some head-scratching, the pedestal computer was born.
Look, all I am trying to say is that it's more than a question of design versus branding. Any decision has to be tempered by the reality of what is possible with the technology at hand. Where the iMac is concern, I feel that Apple engineers have thus far done about as much as they could within the limits set out for them. They have certainly done a lot more than any other PC manufacturer.
What had originally caught my attention was your contention that even Apple has lost interest in the idea [of a pedestal computer], that it's all about branding these days, not usability and that once familiarity has blunted the coolness factor, an iMac design is discarded -- no matter how good it is.
Now, I realize this is just one opinion versus another, but I don't see it that way. As far as the iMac line is concerned, Apple only seems to make radical changes when engineering obstacles necessitate them. I have yet to see a case in which upper management threw out a good design on a whim, thinking it wasn't cool enough anymore. For one thing, it costs a lot to start from scratch every time.
They kept the original CRT enclosure for years, and it still lives on, more or less, in the eMac. The iMac G4 had to be redesigned because the LCD display was a totally different shape. The iMac G5 had to be redesigned because they couldn't fit a PPC970 with all its cooling requirements into a tight little box. Were that possible, we would probably be seeing PowerBook G5s and maybe even Mac mini G5s by now, but I'm not holding my breath.
There is more to design than just aesthetics. Form has to follow function. In the case of the iMac G5, heat dissipation was a much bigger issue than with the G4. A tightly packed spherical CPU is just about the worst possible design for getting rid of heat. I can see why Apple engineers would have had to abandon it whether they wanted to or not.
Put yourself in their shoes for a moment. They're thinking, "The best way to get rid of heat is to build a thin enclosure with lots of surface area. Orient it vertically and we can even take advantage of convective cooling. So how can we achieve this with minimal impact to our existing popular iMac design? Hmm..."
FWIW, I think it was a stroke of genius what they did. The question now is whether they can come up with a similarly clever solution to the PowerBook G5 dilemma?
Think of what all that extra network bandwidth could do for a Beo...ah, forget it.
Someone was telling me Apple has gone through and updated most of the file-handling tools to honour HFS+ meta-data. Is this true? Is there a list someplace of which ones they worked on?
Also, is there any tool under Tiger that can query for Rendezvous/Bonjour services? I couldn't find anything like that in Panther, though maybe I wasn't looked hard enough.
...until DVD John has had a crack at their DRM.
Seriously, can any subscription model stand the test of time from a purely technological standpoint? Recent events have shown that Apple's DRM scheme is pretty easy to circumvent, but then they don't have to worry about every user suddenly owning a million songs when that happens.
How would Napster thwart such a dam burst? Once the DRM is compromised, it's only a matter of time before someone comes along with a server which can dish out unprotected songs to the world by request. You wouldn't even have to waste your own disk space!
It surprises me that the labels are always promoting subscription services. I'd have thought this would be their worst nightmare.
I just tried playing some cloud harp music as an accompaniment to the hamsters. Pretty weird. They should think about touring together.
Many a promising ocean power generation scheme has succumbed to the corrosive effects of sea water. I hope they have carefully considered this.
This might not be an issue for a game console, but for a workstation, wouldn't the Cell's context-switching overhead be rather huge? From what I have read so far, it seems each SPE has 256 KB of local storage, plus another 2 KB for the main register file (128 16-byte registers) and whatever other state information it needs. Since there are 8 of these things, we're talking over 2 MB of state to swap in and out. Would that still be considered a drop in the bucket for most scheduling schemes?
Well, this is the way I envisioned it would work. Alice would register with a CA, securely submitting a random key which could be used for future authentication. Once the CA is satisfied that it has established Alice's identity, the key would go into its database. Bob would do likewise.
When Alice wants to talk to Bob, she first sends a random sequence to the CA to be relayed to Bob. (The CA will assert to Bob that the message came from Alice.) She also sends the same sequence to Bob directly over a secure channel she has set up with him, and Bob simply compares the two copies he receives to see if they match.
If it turns out that there are hash functions which are resistent to quantum cryptoanalysis, you could simplify the scheme further so that Alice just sends a signed message to Bob and Bob asks the CA to verify it.
You're right that there is a weakness here, however, in that the CA would own a key with which they could impersonate Alice. The whole point is that you are supposed to trust your CA, but that certainly does raise the burden of trust.
The problem is that random noise superimposed on data is still random noise as far as the eavesdropper is concerned. You might as well just take a stab at the message itself rather than bother trying to guess the OTP. Given the length of the message and its source and destination, you might be able to make a reasonable stab at it, though even those things can be obfuscated.
No public key scheme can establish the absolute identity of the remote party in the absence of some prior relationship, even if it is through a shared acquaintence. That's why we need certificate authorities to vouch for secure web sites. What they can do is verify that later transmissions are coming from the same source as the original.
Say Bob runs a rumor site. An anonymous tipster could establish a secure connection with him to leak some confidential data. Bob may choose to sit on it until he knows it checks out, but if it does and the tipster keeps throwing more his way, he would be more inclined to trust it because he can verify the sender.
At this point, he still doesn't know that the tipster was Alice. Alice could reveal who she is, but Bob would still need a third party to verify that her signature is genuine. I can't see anything in quantum cryptography which would be any different as far as that's concerned.
...the engineering subgroups have invented so much specialized jargon that they can no longer touch bases, forcing the project to be abandoned.
While I doubt quantum cryptography could ever really take off in the public Internet what with the packets changing hands so many times on route, I was wondering how applicable it might be to wireless LANs? Is it even possible to demonstrate it at microwave frequencies? Will the day come when I can utter the words "secure WI-FI" without an involuntary chuckle?
We use a crystal clock in geophysical equipment whose drift is on the order of 1 part in 10^11 if I have my numbers right. That's about an order of magnitude better than the atomic clock. However, this clock is ovenized (maintained at a constant 70 degrees C) and that helps greatly. I wonder if you can boost the accuracy of the atomic clock in a similar fashion?
This sort of idea makes several assumptions I am not comfortable with:
1. Network bandwidth will no longer be the bottleneck.
2. Your net connection will be more reliable than your hard drive.
3. Your peripherals will work with any computer ever made.
4. You will be equally productive using a variety of different monitors, keyboards, etc.
5. Your activity on someone else's terminal will not be monitored.
Take number 1, for example. Local storage has been expanding at an astounding rate, well ahead of Moore's Law. Internet bandwidth has been improving in a non-uniform and often non-symmetrical pattern. I can only imagine applications in the future working with ever larger data sets, so I am not very optimistic about a net-driven approach. Something like this might work within an organization where you can control many more variables (though I would expect some resistence even then), but I think it will flop bigtime if they try to shift the paradigm for everyone.
Off hand, I can think of three reasons why you might leave your WI-FI unsecured:
1. You are an ignoramus and it never even occurred to you that someone in the neighbourhood could cause you a lot of grief.
2. You are aware of the security issues and have taken some precautions to secure your LAN, but see no harm in allowing others to share your Internet connection. Maybe you actually trust your neighbours...now there's a thought!
3. You are a predator who lures the freeloaders in with a wide open hub. Then you wait around with the packet sniffer running until some sucker shows up and crack your way into his system. The knife cuts both ways, after all.
Articles like this assume everyone is in category 1, but how can you be sure?
Um...forgive me if I'm wrong, but I thought we were talking about hybrid vehicles here. They do not charge off the grid. They use a conventional gasoline/deisel engine to charge up the battery. They are quite fuel efficient, however, since the engine can be kept small. It only needs to handle average power demands, and in a vehicle which stops and goes a lot, this could potentially spell enormous savings. Postal vans would be an ideal application. They might also consider garbage trucks.
Yes, that struck me as a bit odd too. I'm pretty sure they meant to say that methane is 21 times more potent a greenhouse gas than carbon dioxide.
That said, I do remember from my university days studying climatology that there was some speculation that global warming could thin the ozone layer to a degree. I think the idea was that stratospheric cooling would slow down ozone formation. (Heat trapped in the lower level troposphere can actually cause higher levels of the atmosphere to drop in temperature.)
Anyway, that theory is at least a decade old by now, and I don't know if newer computer models have since proven or disproven it.
-Ted
Knowing that Apple's entire line is now G4s or better has made my day! I no longer have to steer clients away from the iBook because our Altivec-tuned modelling engine will crawl on it. I hope more developers will use Altivec now that they know every Mac will support it. Man, I haven't felt this good since Apple moved to PowerPCs and every machine from then on had an FPU!
I wonder if they rushed this one out under pressure to plug the OpenSSH security hole? If so, I guess...well...they succeeded then, didn't they? :-)
1. Textbook patches will be released biweekly. Application of the patches is mandatory.
2. When the blackboard suddenly turns blue, students must leave the classroom in an orderly fashion and return to their seats after ten minutes. No explanation will be given.
3. An alarm bell will sound to signal a massive virus outbreak or worm infestation at Microsoft High. Students are required to calmly exit the building. No drills have been scheduled for this procedure, as it is believed the bell will ring frequently throughout the term without them.
4. Visits to Open Source High are stictly forbidden. Students are, however, encouraged to visit other area schools and report any smaller, well-run institutions with innovative programs to expedite their hostile acquisition by the Microsoft School Board.
5. Our MSSAT exam is similar to--though subtly incompatible with--its government counterpart.
6. Please do not be alarmed by the video portraits of Bill Gates whose eyes follow you down every hall. He got the idea from reading Harry Potter.
I am partially color-blind myself, and have difficulty distinguishing iCal entries by calendar color due to the particular shades Apple has chosen. Anyone know how to tweak these?
-Ted
P.S. While I'm at it, I woudn't mind hearing about any trick for inverting the color scheme at arstechnica back to something sensible!
I live near the Thousand Islands myself. My dream vehicle used to be a personal hovercraft, but now I'm not so sure. I could have a lot of fun with that car!
-Ted
Hasn't this automated fast food concept been tried before, like in my dad's generation? Sure, the automats doubtless had some poor schmuck in the back serving stuff up, but to the customer, it was an automated vending-machine-on-steroids-type interface.
:-)
I imagine the automats died a natural death for cultural -- rather than technological -- reasons, which is why I am skeptical about whether the mechanical fast food industry will ever rise again. Well...maybe in Japan.