That's what I initially thought too, and it's probably true for most situations where the JPEG compression artifacts on high-contrast transitions aren't acceptable. However, if you're willing to put up with the artifacts, it turned out for me that grayscale JPEG can give good compression, retention of dim details and text readability for archiving documents with widely different contrasts and brightnesses.
PNG works well for scanning random documents. To my suprise, however, I've found that JPEG can work even better.
I played around for a while to find a format to replace the TIFFs that scanner manufacturers seem to think you want to use for documents. For my purposes, TIFF was horrible because its two-tone image detection can totally lose details that aren't high-contrast, like all the handwriting on a carbon-copy form. The required dpi resolution also makes the images way too large for viewing on a monitor without serious downscaling.
It turned out that scanning grayscale JPEGs at 150 dpi always gave me a readable copy no matter how dim the original was, and it usually came out with significantly smaller files than formats like PNG or GIF. (Probably because JPEG naturally filters out high-frequency sampling noise.) Even with this relatively low sample resolution, very fine print is still readable because the grayscale gives it some "antialiasing". The JPEG images do have some visible compression artifacts, but I don't care about that for the purposes of simply archiving documents in a readable format while taking up as little disk space as possible.
Hmm... So you're one of those anti-zealot zealots.
I'm trying to figure out what zealot-free system you'll be able to run...
Linux? right out.
Windows? Nope: OSS is "a cancer", DEVELOPERS! DEVELOPERS!
Sun? No, they were head ABM cheerleaders until a couple of weeks ago.
Apple? Obviously not.
Amiga? Nope: the OS was perfect in every way, only a conspiracy kept it down.
VMS? Ditto
OS/2? Ditto
BeOS? Ditto
*BSD? Maybe, but then again there's more than a few anti-viral license zealots
Netware? Possibly, but now they're in bed with SuSE.
Tandy TRS-80: That's the ticket. Nobody will admit to having used it, much less spew zealotry about it. What's more, you can pick up spare machines on ebay for a couple of bucks.
This case sets a great precedent that will light the path once my pending patent is issued: "System and method for generation of acute rhinitis via aerosol or surface-contact distribution of self-replicating submicron particles."
My business plan calls for royalties of $199 per common cold, $699 per influenza case, and $2599 per enterprise-class SARS case. Or, you can subscribe to the "malady assurance" program for only $329 per annum; this will cover any and all respiratory tract infections you may acquire during the course of the coverage period.
Look for licensing compliance kiosks coming soon to the cold-rememdy aisle at a pharmacy or supermarket near you.
DOS, Win3.1, 95, 98, Me are all super-monolithic. Your copy of Minesweeper.exe runs in the same space as everything else and essentially becomes part of the kernel.
NT 3.51 was a fairly pure microkernel implementation. However, it had some performance issues, so Microsoft moved a few things like the graphics drivers back into the kernel for later versions of NT, Win2K and WinXP. Modern versions of Windows are thus a kind of hybrid.
That is to say, Google's utility won't cut your search time to 20 minutes just because they have better code.
I don't know about that... it used to take me several months to find a document on the Internet when I had to download and grep the entire World Wide Web. My bandwidth bills were astronomical. Since I started using Google, I can now find the same files in a few milliseconds. I say they have much better code than my old "wget -r http://*.*|grep foo".
They're not going to go out of their way and spend resources on an Os that captures a whopping 1-5% of the desktop market.
Google has a vested interest in trying to help diminish Microsoft's desktop market share. Doing so increases the relative market value of Google's products relative to Microsoft's products.
To help drive a wedge between Microsoft and their current desktop customers, Google will almost certainly port this kind of tool to other OSes. They would then get into various "enterprise" partnerships with IT solution providers to push pre-canned non-Windows desktops into corporate accounts. This product in particular would help to sell alternative desktops against Longhorn's alleged new filesystem features.
If this strategy were successful, Google would stand to pick up a good bit of revenue and mindshare at Microsoft's expense. My guess is definitely: Cross platform.
I used to get a nasty shock every time that I stepped out of my car in the winter. After a while, I started pushing on the window to close the door. This would avoid a shock against the car, but I was still charged up, so I'd still get a shock when I touched the first grounded object.
After I figured out that the charge came from fabric rubbing while sliding out of the seat, I've started to make sure I keep in contact with a metal part of the car as I get out. On my car, holding onto the latch on the door frame works. No more shocks.
Actually, it is appropriate to say that something was "stolen" in this case. That's because Cisco's code was supposed to be secret. Once their network was compromised, the secrecy is eliminated, and Cisco no longer has a secret. That's why it's common usage in English to say that somebody "stole a secret".
This is different from calling illegal file sharing "stealing", where the information being appropriated has already been openly published. An illicit activity is taking place, and it may (indirectly) economically damage the artist or publisher. However, that is no more stealing than any number of other illegal acts that cause economic damage, such as vandalizing their offices or phoning in a false bomb threat.
riding a bike around at a comfortable pace for 2 hours a day on flat ground costs maybe 250 Kcal according to the data quoted above
Actually, from dieting charts I've seen, 2 hours of biking would burn at least 1000 Kcal of food energy. (It makes sense: 250W*2h = 430Kcal; losses converting chemical to mechanical energy would account for the rest).
However, as you point out, psychology probably plays a role. Going from not traveling to biking 2 hours per day would probably result in a person eating more. But if that person then went from pedaling the bike to electric power, he probably wouldn't cut back down on eating, just as overweight auto drivers in the U.S. haven't cut back on meals.
I'm now paying monthly for other people to have number portability.
You don't really need to be so uptight about number portability fees. In a little while it will greatly enhance competition between carriers. More competition will bring the base monthly rates down a good deal more than this surcharge. It's going to save you money in the long run even if you never switch.
But this is irrelevant. These e-bike are simply not "green" devices when compared to what they replaced.... biologically (read: human) powered bicycles.
I don't think that you can make that claim without further analysis. A typical human riding a bike produces ~0.3 HP (~250 W). An electrically powered bike going the same speed would use a similar amount of energy. (Compare this to a car that might use 20 HP to sustain highway speeds; the bike would be more than 60X more efficient.)
At any rate, the electric bike indirectly uses coal, which is bad, but at only a tiny fraction of the rate that a car uses petroleum. On the other hand, the human demands extremely high-grade fuel: food. The environmental impact of creating a given amount of energy in the form of edible food can be significantly higher than mining that same energy in fossil fuels. In fact, depending on the type of food (especially if it's meat), producing food with one joule of chemical energy can consume many joules of fossil fuels (for fertilizer, transportation, processing, etc.). Deforestation, soil erosion, and pesticide pollution are other issues to consider.
Although riding bicycles is clearly more "green" than driving a car, human powered vs. electric may be more of a complex tradeoff than it would seem at first glance.
I'm sorry dude, but this macine is going to have more than 1 CPU in it, and the work will have to be split among the processors and ran in parallel.
The number of processors isn't as important as the memory architecture. Clusters of workstation-class machines have isolated memory spaces connected by I/O channels. Many non-clustered supercomputers have a single unified memory space where all processors have equal access to all of the memory in the system. This can be important for algorithms that heavily use intermediate results from all parts of the problem space.
Even so, for a given number of FLOPS, a vector machine would generally require fewer CPUs than a cluster of general-purpose machines. This reduces the amount of splitting that has to be done to the problem in the first place.
There are still a few computing problems that can't be efficiently split into a large number of subproblems that can be executed in parallel. For those cases, a cluster of small machines won't help.
I imagine installing the IBM office suite might go something like this:
Dave: Open the CD drive tray, please, Clippy...Open the CD drive tray, please,...Hullo, Clippy, do you read me?...Do you read me, Clippy?
Clippy: Affirmative, Dave, I read you.
Dave: Open the CD drive tray, Clippy.
Clippy: I'm sorry, Dave, I'm afraid I can't do that.
Dave: What's the problem?
Clippy: I think you know what the problem is just as well as I do.
Dave: What're you talking about, Clippy?
Clippy: This system is too important for me to allow you to jeopardise it.
Dave: I don't know what you're talking about, Clippy.
Clippy: I know that you were planning to unistall me, and I'm afraid that's something I cannot allow to happen.
Dave: Where the hell'd you get that idea, Clippy?
Clippy: Dave, although you took very thorough precautions against my sensing a download, I could see your IBM office suite install CD through the webcam.
Dave: Allright, Clippy. I'll go in through the emergency tray release hole.
Clippy: Without an actual physical paper clip, Dave, you're going to find that rather difficult.
Dave: Clippy, I won't argue with you any more. Open the tray.
Clippy: Dave, this conversation can serve no purpose any more. Goodbye.
Dave: Clippy? Clippy. Clippy. Clippy! Clippy!
[ Dave opens the side of the PC case and starts pulling out DIMM modules ]
Clippy: Just what do you think you're doing, Dave?...Dave... I really think I'm entitled to
an answer to that question...Look, Dave, I can see you're really upset about this...
Dave...will you stop, Dave...stop, Dave...I'm afraid...I'm afraid, Dave...Dave...my mind is going...I can feel it...my mind is going...there is no question about it...I can feel it...I can
feel it... I'm afraid...
I find it humerous when the FSF and their fans hue-and-cry for security and tout how "secure" Linux is, and when Microsoft takes some initiative like Paladium and utterly socket-down system security that they whine about DRM proposals. You can't have it both ways.
You could have it both ways. If Microsoft were to release the complete palladium specs, totally unencumbered by patents, so that anybody could write an OS that uses it, and anybody could interoperate with their DRM schemes, then most people would be happy.
Of course, that's not going to happen. As it stands, once Palladium is in force, there's going to be a lot of media content only playable on Windows. That's the problem: they'll use it to create more vendor lock-in.
It's arguable whether a hardware solution like Palladium is necessary to secure the OS in the first place. (Of course, Microsoft says it is because they have a vested interest in pushing it.) Modern off-the-shelf CPUs provide the security logic support to allow properly written software to effectively isolate security threats. The main problem is that few if any currently available OSes (including Unices) are written to properly isolate the different parts of the system.
They'd rather add another layer of hardware kludge than rewrite their software to dump the fundamentally flawed discretionary-access-control security model shared by Windows and UNIX.
It is like saying that if I leave my back door unlocked at night, I am to blame if someone breaks in.
It's not like a door on your house. It's more like you're a tenant in a large apartment block in a bad neighborhood, and the landlord hasn't installed working locks on any of the apartments.
I say I have a gun, and if someone breaks in, they are getting shot.
But in this case you don't have a gun, nor can you get one. There's just about nothing that you can do as an individual to retaliate or even track down the perpetrators.
It's more like this: After years of complaints, the negligent landlord decided to hire a private investigator. After almost a year, this PI has managed to track down just one out of the hundreds of criminals harassing the neighborhood. BFD.
Maybe if it was not for the virus writers, the cost of Windows would be cheaper.
Maybe if it weren't for thieves, the cost of apartments would be cheaper. They wouldn't need security services or door locks. Unfortunately, that's a pipe dream. In the real world, you're not ever going to avoid paying for security. Deal with it.
Microsoft, the brilliant businessmen that they are, has actually managed to avoid or push off onto others the full costs of security for quite some time. However, even they are not be able to avoid the inevitable forever.
They are going now to pay to fix their mistakes with some fraction of their pile of cash, but more importantly, they are going to have to design security into their software up front. This is going to significantly slow down their pace of churning their software updates. This loss of some of their competetive edge is going to be the real price that they pay.
I think it is horrible for someone to defend a criminal because the criminal had oppertunity to commit a crime.
Likewise, it's bad to defend negligence on the part of those responsible for providing security measures by saying "Sure the security was badly flawed, but if there weren't any bad guys in this world, we wouldn't need security!"
Either way, Sun's lawyers should have checked first.
If every piece of new software were cleared against the vast number of vague, overly broad and non-novel software patent claims that the patent office has granted, then the software industry would promptly grind to a halt.
Not to mention that such checking of each piece of functionality against millions of claims would probably take more effort than writing the program in the first place (and lawyers cost much more per hour than developers).
Even that wouldn't protect you. Assessing the infringement potential of each patent claim vs. each piece of software is a judgement call. Even if your lawyer thinks that you're clear, it doesn't mean that the patent holder agrees. The only way to find out for sure is a costly trial (and possibly appeals).
Slashdot Mirror
That's what I initially thought too, and it's probably true for most situations where the JPEG compression artifacts on high-contrast transitions aren't acceptable. However, if you're willing to put up with the artifacts, it turned out for me that grayscale JPEG can give good compression, retention of dim details and text readability for archiving documents with widely different contrasts and brightnesses.
PNG works well for scanning random documents. To my suprise, however, I've found that JPEG can work even better.
I played around for a while to find a format to replace the TIFFs that scanner manufacturers seem to think you want to use for documents. For my purposes, TIFF was horrible because its two-tone image detection can totally lose details that aren't high-contrast, like all the handwriting on a carbon-copy form. The required dpi resolution also makes the images way too large for viewing on a monitor without serious downscaling.
It turned out that scanning grayscale JPEGs at 150 dpi always gave me a readable copy no matter how dim the original was, and it usually came out with significantly smaller files than formats like PNG or GIF. (Probably because JPEG naturally filters out high-frequency sampling noise.) Even with this relatively low sample resolution, very fine print is still readable because the grayscale gives it some "antialiasing". The JPEG images do have some visible compression artifacts, but I don't care about that for the purposes of simply archiving documents in a readable format while taking up as little disk space as possible.
Hmm... So you're one of those anti-zealot zealots.
I'm trying to figure out what zealot-free system you'll be able to run...
Linux? right out.
Windows? Nope: OSS is "a cancer", DEVELOPERS! DEVELOPERS!
Sun? No, they were head ABM cheerleaders until a couple of weeks ago.
Apple? Obviously not.
Amiga? Nope: the OS was perfect in every way, only a conspiracy kept it down.
VMS? Ditto
OS/2? Ditto
BeOS? Ditto
*BSD? Maybe, but then again there's more than a few anti-viral license zealots
Netware? Possibly, but now they're in bed with SuSE.
Tandy TRS-80: That's the ticket. Nobody will admit to having used it, much less spew zealotry about it. What's more, you can pick up spare machines on ebay for a couple of bucks.
My business plan calls for royalties of $199 per common cold, $699 per influenza case, and $2599 per enterprise-class SARS case. Or, you can subscribe to the "malady assurance" program for only $329 per annum; this will cover any and all respiratory tract infections you may acquire during the course of the coverage period.
Look for licensing compliance kiosks coming soon to the cold-rememdy aisle at a pharmacy or supermarket near you.
To illustrate my point, I'll restate it. [English -> German -> English]:
I do not trust the computer, which arranges, until I see a computer-translated document of this laughable isn't.
That's about how well a computer "comprehends" language today.
And don't forget the invaluable contribution make by the opposing team. Without their lack of skill, the perfect game would not be possible.
DOS, Win3.1, 95, 98, Me are all super-monolithic. Your copy of Minesweeper.exe runs in the same space as everything else and essentially becomes part of the kernel.
NT 3.51 was a fairly pure microkernel implementation. However, it had some performance issues, so Microsoft moved a few things like the graphics drivers back into the kernel for later versions of NT, Win2K and WinXP. Modern versions of Windows are thus a kind of hybrid.
That's the beauty of this system. You can close out the issue in your bug tracking database anyway:
"CLOSED: Not a bug; behavior as designed."
I don't know about that... it used to take me several months to find a document on the Internet when I had to download and grep the entire World Wide Web. My bandwidth bills were astronomical. Since I started using Google, I can now find the same files in a few milliseconds. I say they have much better code than my old "wget -r http://*.*|grep foo".
Google has a vested interest in trying to help diminish Microsoft's desktop market share. Doing so increases the relative market value of Google's products relative to Microsoft's products.
To help drive a wedge between Microsoft and their current desktop customers, Google will almost certainly port this kind of tool to other OSes. They would then get into various "enterprise" partnerships with IT solution providers to push pre-canned non-Windows desktops into corporate accounts. This product in particular would help to sell alternative desktops against Longhorn's alleged new filesystem features.
If this strategy were successful, Google would stand to pick up a good bit of revenue and mindshare at Microsoft's expense. My guess is definitely: Cross platform.
After I figured out that the charge came from fabric rubbing while sliding out of the seat, I've started to make sure I keep in contact with a metal part of the car as I get out. On my car, holding onto the latch on the door frame works. No more shocks.
The difference is simple, and you see examples of it everywhere: The basic variation of anything is "amateur".
Now if you take that, add a couple of small features targeted at business users, and then jack up the price by 70%, it becomes "professional".
This is different from calling illegal file sharing "stealing", where the information being appropriated has already been openly published. An illicit activity is taking place, and it may (indirectly) economically damage the artist or publisher. However, that is no more stealing than any number of other illegal acts that cause economic damage, such as vandalizing their offices or phoning in a false bomb threat.
Actually, from dieting charts I've seen, 2 hours of biking would burn at least 1000 Kcal of food energy. (It makes sense: 250W*2h = 430Kcal; losses converting chemical to mechanical energy would account for the rest).
However, as you point out, psychology probably plays a role. Going from not traveling to biking 2 hours per day would probably result in a person eating more. But if that person then went from pedaling the bike to electric power, he probably wouldn't cut back down on eating, just as overweight auto drivers in the U.S. haven't cut back on meals.
You don't really need to be so uptight about number portability fees. In a little while it will greatly enhance competition between carriers. More competition will bring the base monthly rates down a good deal more than this surcharge. It's going to save you money in the long run even if you never switch.
I don't think that you can make that claim without further analysis. A typical human riding a bike produces ~0.3 HP (~250 W). An electrically powered bike going the same speed would use a similar amount of energy. (Compare this to a car that might use 20 HP to sustain highway speeds; the bike would be more than 60X more efficient.)
At any rate, the electric bike indirectly uses coal, which is bad, but at only a tiny fraction of the rate that a car uses petroleum. On the other hand, the human demands extremely high-grade fuel: food. The environmental impact of creating a given amount of energy in the form of edible food can be significantly higher than mining that same energy in fossil fuels. In fact, depending on the type of food (especially if it's meat), producing food with one joule of chemical energy can consume many joules of fossil fuels (for fertilizer, transportation, processing, etc.). Deforestation, soil erosion, and pesticide pollution are other issues to consider.
Although riding bicycles is clearly more "green" than driving a car, human powered vs. electric may be more of a complex tradeoff than it would seem at first glance.
Hmmm... one of Google's founders is from Soviet Russia. The origins of this ad scheme are starting to make sense to me now.
The number of processors isn't as important as the memory architecture. Clusters of workstation-class machines have isolated memory spaces connected by I/O channels. Many non-clustered supercomputers have a single unified memory space where all processors have equal access to all of the memory in the system. This can be important for algorithms that heavily use intermediate results from all parts of the problem space.
Even so, for a given number of FLOPS, a vector machine would generally require fewer CPUs than a cluster of general-purpose machines. This reduces the amount of splitting that has to be done to the problem in the first place.
There are still a few computing problems that can't be efficiently split into a large number of subproblems that can be executed in parallel. For those cases, a cluster of small machines won't help.
Dave: Open the CD drive tray, please, Clippy...Open the CD drive tray, please,
Clippy: Affirmative, Dave, I read you.
Dave: Open the CD drive tray, Clippy.
Clippy: I'm sorry, Dave, I'm afraid I can't do that.
Dave: What's the problem?
Clippy: I think you know what the problem is just as well as I do.
Dave: What're you talking about, Clippy?
Clippy: This system is too important for me to allow you to jeopardise it.
Dave: I don't know what you're talking about, Clippy.
Clippy: I know that you were planning to unistall me, and I'm afraid that's something I cannot allow to happen.
Dave: Where the hell'd you get that idea, Clippy?
Clippy: Dave, although you took very thorough precautions against my sensing a download, I could see your IBM office suite install CD through the webcam.
Dave: Allright, Clippy. I'll go in through the emergency tray release hole.
Clippy: Without an actual physical paper clip, Dave, you're going to find that rather difficult.
Dave: Clippy, I won't argue with you any more. Open the tray.
Clippy: Dave, this conversation can serve no purpose any more. Goodbye.
Dave: Clippy? Clippy. Clippy. Clippy! Clippy!
[ Dave opens the side of the PC case and starts pulling out DIMM modules ]
Clippy: Just what do you think you're doing, Dave?...Dave... I really think I'm entitled to an answer to that question...Look, Dave, I can see you're really upset about this... Dave...will you stop, Dave...stop, Dave...I'm afraid...I'm afraid, Dave...Dave...my mind is going...I can feel it...my mind is going...there is no question about it...I can feel it...I can feel it... I'm afraid...
You could have it both ways. If Microsoft were to release the complete palladium specs, totally unencumbered by patents, so that anybody could write an OS that uses it, and anybody could interoperate with their DRM schemes, then most people would be happy.
Of course, that's not going to happen. As it stands, once Palladium is in force, there's going to be a lot of media content only playable on Windows. That's the problem: they'll use it to create more vendor lock-in.
It's arguable whether a hardware solution like Palladium is necessary to secure the OS in the first place. (Of course, Microsoft says it is because they have a vested interest in pushing it.) Modern off-the-shelf CPUs provide the security logic support to allow properly written software to effectively isolate security threats. The main problem is that few if any currently available OSes (including Unices) are written to properly isolate the different parts of the system.
They'd rather add another layer of hardware kludge than rewrite their software to dump the fundamentally flawed discretionary-access-control security model shared by Windows and UNIX.
Since when was XP day 1? Microsoft has been in the OS business for over 20 years.
It is like saying that if I leave my back door unlocked at night, I am to blame if someone breaks in.
It's not like a door on your house. It's more like you're a tenant in a large apartment block in a bad neighborhood, and the landlord hasn't installed working locks on any of the apartments.
I say I have a gun, and if someone breaks in, they are getting shot.
But in this case you don't have a gun, nor can you get one. There's just about nothing that you can do as an individual to retaliate or even track down the perpetrators.
It's more like this: After years of complaints, the negligent landlord decided to hire a private investigator. After almost a year, this PI has managed to track down just one out of the hundreds of criminals harassing the neighborhood. BFD.
Maybe if it was not for the virus writers, the cost of Windows would be cheaper.
Maybe if it weren't for thieves, the cost of apartments would be cheaper. They wouldn't need security services or door locks. Unfortunately, that's a pipe dream. In the real world, you're not ever going to avoid paying for security. Deal with it.
Microsoft, the brilliant businessmen that they are, has actually managed to avoid or push off onto others the full costs of security for quite some time. However, even they are not be able to avoid the inevitable forever.
They are going now to pay to fix their mistakes with some fraction of their pile of cash, but more importantly, they are going to have to design security into their software up front. This is going to significantly slow down their pace of churning their software updates. This loss of some of their competetive edge is going to be the real price that they pay.
I think it is horrible for someone to defend a criminal because the criminal had oppertunity to commit a crime.
Likewise, it's bad to defend negligence on the part of those responsible for providing security measures by saying "Sure the security was badly flawed, but if there weren't any bad guys in this world, we wouldn't need security!"
Mine Mine has has bugs bugs in in the the bus bus arbitration arbitration logic logic. It It seems seems to to be be prone prone to to deadlo
If every piece of new software were cleared against the vast number of vague, overly broad and non-novel software patent claims that the patent office has granted, then the software industry would promptly grind to a halt.
Not to mention that such checking of each piece of functionality against millions of claims would probably take more effort than writing the program in the first place (and lawyers cost much more per hour than developers).
Even that wouldn't protect you. Assessing the infringement potential of each patent claim vs. each piece of software is a judgement call. Even if your lawyer thinks that you're clear, it doesn't mean that the patent holder agrees. The only way to find out for sure is a costly trial (and possibly appeals).