PNG is lossless, so image quality is always perfect (not merely "surprising").
JPG is lossy, so it's allowed to degrade image quality to get better compression (and this is something you can tune when you do the compression).
GIF is lossless also, but it doesn't compress as well as PNG, usually.
So, to answer your questions...
Can you get amazing compression rates with PNG? Yes, but this depends on your definition of "amazing."
How do I achieve better compression than GIF or JPG?Pngcrush will take an existing PNG and make it as small as possible with no loss of quality.
How can I get PNG to perform consistently better than GIF or JPG? You can't. As you observe, it depends on the image. A photograph will pretty much always compress better with JPG (which loses quality, remember). PNG will do better on an artificial image such as a cartoon or a screen capture.
You might have a point if this scheme involved using money. In this case, however, the "payment" is a proof-of-work. The user is paying in CPU cycles "spent" to send the message.
One cannot legally rig up a shotgun or some other dangerous device to automatically discharge upon the violation of a perimeter, how is this different?
This may vary from state to state. What you're talking about is called a "man trap." In my state it's legal to set a man trap to protect people (i.e., yourself or your family), but not to protect property (e.g., your house while you're away). This jacket is different in two ways: (1) it's not lethal, and (2) it always protects a person, never property.
The fact that people don't use all their bandwidth all the time is the reason that access is as cheap as it is. If the ISP had to design their network so that everyone could run at full speed at the same time, their expenses would increase tenfold--and they'd pass the savings on to you.
I find it interesting that in the ten years I've been on the net, the attitude has changed so dramatically from one of conservation ("the bandwidth you save will be your own") to exploitation ("I paid for it, so I should take it for all it's worth"). People think that because they pay a flat rate for bandwidth, that they're not really paying for usage. Naively, that's true, but in the big picture, what you're really doing is spreading cost over many customers.
So, are you the greedy neighbor causing my ISP to raise its rates?
I love the idea of local wireless networks for when my neighbors want to talk to me, but I'd like them to pay for their own Internet connection.
I've remarked in the past that the reason that people are so disappointed
in the Internet is because they're expecting something that works like the
telephone.
They expect it to work every time they use it, but:
Web sites go down.
Routes disappear.
Their email server goes down.
Terminal servers get full.
They expect it to work the same way every time, but:
Some sites load slower than others.
Some people can deal with their weird email attachments and others can't.
Services go through normal changes and upgrades.
The article here echoes my point. Businesses are complaining that the
Internet is (1) not reliable enough, and (2) not fast enough. I get the
impression that they're primarily concerned with real time audio and video,
two applications that are particularly sensitive to speed and reliability
problems.
There must be some old text books whose copyrights have expired.
They're probably out of date (unless they're arithematic books), but they
might be a good starting point for someone who knows the subject. Rather
than compile a new one from scratch, why not start with one of those and
bring it up to date?
When I've talked to Outlook users about using encryption, the barrier is that they don't want to use passwords for it. They've clicked "remember my password" for their POP3 account, and they don't want to remember another one to decrypt or sign emails.
I think that generations to come will use encryption because they're already comfortable with email. Yes, just using Eudora is a struggle for my mom, but it's easy for my sister, and it will be second nature for my daughter. Using crypto will be a much smaller hurdle in the future because everyone will alredy be at a higher level of expertise to begin with.
Likewise I think PKI will get better as more people become familiar with the technology.
There's a huge learning curve for the written word. Most people spend years in English classes and still don't get it right! Crypto is easier than that.
Re:PGPing your email?
on
GPG vs. PGP?
·
· Score: 1
I sign all my emails for three basic reasons:
Make more people aware of encryption.
Make it harder for someone else to impersonate me (not that this is a hot objective for the general populace).
I think it's cool.
It causes problems once in a while, but for the most part I'm glad I started doing it. A paranoid voice in the back of my head wonders if someone will take too seriously (because I signed it) something I said in jest, but it's easier to refute that when I sign everything, even emails as trivial as, "yes, I'll be there tonight."
I agree the punishing the source of the attack is the best solution in a perfect world, but in the vast majority of real DoS attack cases, that solution is not available.
As I see it, there are three (right) ways to attack this kind of problem: (1) technical, (2) social, and (3) legal.
If we had a technical solution, we'd be using it.
Social solutions aren't going to work well on the kiddies since they're so anti-social anyway. (Imagine the laughter at a "just say 'no' to scripts" campaign.)
The laws are already in place. To enforce them requires the death of anonymity on the Internet.
Again, I agree that punishing the source of the attack is the best idea, but the reality doesn't support that solution. What do you do when you have five customers on the phone all complaining that their service doesn't work? If you called your service provider about this problem, and their response was, "well, ideally we'd be able to punish the person causing this, but we can't, so we're just going to ride it out, hope it doesn't happen again, and let your service suffer in the mean time", how would you react?
If I'm just going to decide that some of my customers won't have service, why not make it the smallest number possible?
I'm an admin at a medium sized ISP. We've repeatedly had situations where the actions of one user are affecting the service we can provide to our other users. When one person floods the network and no one else can use it as a result, what do you do?
Our contract says we can pull the plug for no reason if we feel like it. Our customers signed it, even if they didn't read it.
The user in the story is a little different only because the attack is coming from the outside. It's not as easy to call that abuse and pull the plug, but the choice is the same: let it go and let everyone suffer, or deny service to this one customer and continue to provide service to the other customers.
If I were using an ISP that was continually under attack, for any reason, and I couldn't use the service, I'd go to another one. As a customer, I'm not going to feel good about some principle being upheld if I can't work or play as I want. If all of my customers make that same decision when I let abuse continue, I'm out of business. My consolation after bankruptcy will be what? That I continued to provide what little service I could to some user who was hurting everyone else?
I'd love to see more top level domains if we could only keep the large companies from spreading themselves over all of them. Network Solutions encourages you to register your domain in all of.net,.org, and.com. Admittedly, not everyone would want to be in.sex just because they have a.com, but you would still have organizations trying to register themselves in every TLD they thought they belonged in.
The real problem is that a site owner wants the user to be able to type in a word and get his or her site. So they register the site in every domain to ensure that. The name space is a (large) finite resource, and one organization having more than one name is, IMO, pollution.
I'd like to see a TLD system where every site fits into one and only one TLD, and a user who had never been to the site before would be able to figure out which TLD to use. I don't think it's going to happen.
I love the length. I dig the topic. I'm looking forward to the discussion (I haven't looked at it yet).
I think there are quite a few other ways to categorize the "places" of the Internet.
Web sites vs. Usenet vs. Mailing lists vs. IRC vs. other communities
Sites that have a topic (i.e., sex, God, tech), and those that don't (Yahoo).
Sites that are out to make a buck and those that aren't.
Maybe you could plot out every site on several of these axes and come up with some kind of map of the space like a network topology but with more dimensions and less usefulness to people who aren't smoking anything.
Harmed in the sense that, in my opinnion, pornography when combined with sexual abuse is harmful to children. It has lasting psychological effects.
Chocolate pudding when combined with sexual abuse is harmful to children. Let's ban that too. In fact, replace "pornography" above with practically anything.
Can you tell us how pornography, by itself, in the absence of anything else (especially child abusing adults), is harmful to children?
Slashdot Mirror
Maybe this will help.
So, to answer your questions...
Can you get amazing compression rates with PNG? Yes, but this depends on your definition of "amazing."
How do I achieve better compression than GIF or JPG? Pngcrush will take an existing PNG and make it as small as possible with no loss of quality.
How can I get PNG to perform consistently better than GIF or JPG? You can't. As you observe, it depends on the image. A photograph will pretty much always compress better with JPG (which loses quality, remember). PNG will do better on an artificial image such as a cartoon or a screen capture.
Hope this helps.
You might have a point if this scheme involved using money. In this case, however, the "payment" is a proof-of-work. The user is paying in CPU cycles "spent" to send the message.
One cannot legally rig up a shotgun or some other dangerous device to automatically discharge upon the violation of a perimeter, how is this different?
This may vary from state to state. What you're talking about is called a "man trap." In my state it's legal to set a man trap to protect people (i.e., yourself or your family), but not to protect property (e.g., your house while you're away). This jacket is different in two ways: (1) it's not lethal, and (2) it always protects a person, never property.
The fact that people don't use all their bandwidth all the time is the reason that access is as cheap as it is. If the ISP had to design their network so that everyone could run at full speed at the same time, their expenses would increase tenfold--and they'd pass the savings on to you.
I find it interesting that in the ten years I've been on the net, the attitude has changed so dramatically from one of conservation ("the bandwidth you save will be your own") to exploitation ("I paid for it, so I should take it for all it's worth"). People think that because they pay a flat rate for bandwidth, that they're not really paying for usage. Naively, that's true, but in the big picture, what you're really doing is spreading cost over many customers.
So, are you the greedy neighbor causing my ISP to raise its rates?
I love the idea of local wireless networks for when my neighbors want to talk to me, but I'd like them to pay for their own Internet connection.
I put a mirror here. The original seems to be slashdotted.
I've remarked in the past that the reason that people are so disappointed in the Internet is because they're expecting something that works like the telephone.
The article here echoes my point. Businesses are complaining that the Internet is (1) not reliable enough, and (2) not fast enough. I get the impression that they're primarily concerned with real time audio and video, two applications that are particularly sensitive to speed and reliability problems.
There must be some old text books whose copyrights have expired. They're probably out of date (unless they're arithematic books), but they might be a good starting point for someone who knows the subject. Rather than compile a new one from scratch, why not start with one of those and bring it up to date?
When I've talked to Outlook users about using encryption, the barrier is that they don't want to use passwords for it. They've clicked "remember my password" for their POP3 account, and they don't want to remember another one to decrypt or sign emails.
I think that generations to come will use encryption because they're already comfortable with email. Yes, just using Eudora is a struggle for my mom, but it's easy for my sister, and it will be second nature for my daughter. Using crypto will be a much smaller hurdle in the future because everyone will alredy be at a higher level of expertise to begin with.
Likewise I think PKI will get better as more people become familiar with the technology.
There's a huge learning curve for the written word. Most people spend years in English classes and still don't get it right! Crypto is easier than that.
I sign all my emails for three basic reasons:
It causes problems once in a while, but for the most part I'm glad I started doing it. A paranoid voice in the back of my head wonders if someone will take too seriously (because I signed it) something I said in jest, but it's easier to refute that when I sign everything, even emails as trivial as, "yes, I'll be there tonight."
I agree the punishing the source of the attack is the best solution in a perfect world, but in the vast majority of real DoS attack cases, that solution is not available.
As I see it, there are three (right) ways to attack this kind of problem: (1) technical, (2) social, and (3) legal.
If we had a technical solution, we'd be using it.
Social solutions aren't going to work well on the kiddies since they're so anti-social anyway. (Imagine the laughter at a "just say 'no' to scripts" campaign.)
The laws are already in place. To enforce them requires the death of anonymity on the Internet.
Again, I agree that punishing the source of the attack is the best idea, but the reality doesn't support that solution. What do you do when you have five customers on the phone all complaining that their service doesn't work? If you called your service provider about this problem, and their response was, "well, ideally we'd be able to punish the person causing this, but we can't, so we're just going to ride it out, hope it doesn't happen again, and let your service suffer in the mean time", how would you react?
If I'm just going to decide that some of my customers won't have service, why not make it the smallest number possible?
What is justice? Is it just to be without Internet service because your neighbor irritated a kiddie with a script?
More like someone sets your house on fire, and the fire department demolishes it to keep it from burning down the whole neighborhood.
I'm an admin at a medium sized ISP. We've repeatedly had situations where the actions of one user are affecting the service we can provide to our other users. When one person floods the network and no one else can use it as a result, what do you do?
Our contract says we can pull the plug for no reason if we feel like it. Our customers signed it, even if they didn't read it.
The user in the story is a little different only because the attack is coming from the outside. It's not as easy to call that abuse and pull the plug, but the choice is the same: let it go and let everyone suffer, or deny service to this one customer and continue to provide service to the other customers.
If I were using an ISP that was continually under attack, for any reason, and I couldn't use the service, I'd go to another one. As a customer, I'm not going to feel good about some principle being upheld if I can't work or play as I want. If all of my customers make that same decision when I let abuse continue, I'm out of business. My consolation after bankruptcy will be what? That I continued to provide what little service I could to some user who was hurting everyone else?
I'd love to see more top level domains if we could only keep the large companies from spreading themselves over all of them. Network Solutions encourages you to register your domain in all of .net, .org, and .com. Admittedly, not everyone would want to be in .sex just because they have a .com, but you would still have organizations trying to register themselves in every TLD they thought they belonged in.
The real problem is that a site owner wants the user to be able to type in a word and get his or her site. So they register the site in every domain to ensure that. The name space is a (large) finite resource, and one organization having more than one name is, IMO, pollution.
I'd like to see a TLD system where every site fits into one and only one TLD, and a user who had never been to the site before would be able to figure out which TLD to use. I don't think it's going to happen.
I love the length. I dig the topic. I'm looking forward to the discussion (I haven't looked at it yet).
I think there are quite a few other ways to categorize the "places" of the Internet.
Maybe you could plot out every site on several of these axes and come up with some kind of map of the space like a network topology but with more dimensions and less usefulness to people who aren't smoking anything.
Chocolate pudding when combined with sexual abuse is harmful to children. Let's ban that too. In fact, replace "pornography" above with practically anything.
Can you tell us how pornography, by itself, in the absence of anything else (especially child abusing adults), is harmful to children?