OK, so if I get 12 mbps speed, which is what is needed for high definition video in real time, you're going to charge $360 per month? It seems like your network isn't ready for 2009. Oh wait, it's still 2008. I guess you have another year.
Just because all your customers already pay so much for so little now, because that's all you offer, does not justify keeping things at this level in the future.
You mention a measly 1.5 mbps. Someone wanting to download a 2 hour high definition movie is going to have to be doing the download for 16 hours to get it all. At 1.5 mbps, it will take a full solid month to download 486 gigabytes (not figuring in overhead and such). Are you going to cut off customers that use it?
White LEDs have the same problem as fluorescent. They are both based on fluorescent material driven by deep blue and/or UV light. Some LEDs are a combination of red, green, and blue LED dies. Having lots of distinct wavelengths might well be the solution. They could do that not only in LED (I've seen 22 different wavelengths available) but also in a new fluorescent formula. The catch is that it is expensive. But I suspect the LED field will be able to achieve this before the fluorescent field.
NDRs are evil if sent to the wrong address. You need to send them to THE SENDER... not some victim the spammer has fingered. In the case of spam, the spammer is THE SENDER (regardless of what he puts on as the return address). So that means NDRs are not evil in cases of legitimate mail where the sender address is not a forgery. SMTP is flawed in the sense that if fails to provide a means to correctly identify the sender. Better do your SPF and other tests to be sure. Or just do the "safe harber" and do rejects during the SMTP session.
You can't be right because this is still creating an abusive situation. Keep in mind that the major issue with spam has nothing to do with the contents. So the fact that the backscatter does not include the spammer message just means the spammer can't make use of it to pass the message. But it still causes the same problems as spam, which is bogged down mail servers.
I have had one of my hosted email addresses used by a spammer in a large spam run of several million or so. My server once got hit by about 250,000 pieces of backscatter in just 3 days. About 6000 mail servers were the abuse vectors in this. They all got blocked. They will stay blocked until they can prove they fixed the issue. Google was NOT one of them (most were in Russia and the spam run was in Russian). But Google could have been one of them if they were doing then what they are reportedly doing today.
Sending back a bounce message after the fact is abuse... if the bouncing server did not verify that the return email address it is using designates the sending server as appropriate for it. Google may well be applying that test (I see references to SPF in the headers on Gmail). So they could well be limiting backscatter to situations where the forged email address is one that matches the sending address (for example a botnet infected machine on Comcast forges a different Comcast user when sending mail through Comcast's outbound servers over to Gmail).
But any mail server not applying the appropriate test could be an abusive (and therefore evil) server. The correct logic is to never send a bounce message unless you have an unambiguous positive verification of the address via SPF, MX, etc. The RFC requires sending the delivery notification to the sender. The sender is the spammer. But you don't have the spammer's address; you just have some forged address. So you cannot send it correctly. So just don't send it. FYI: do not assume that a lack of negatives on that email address means it is a positive. It does not mean that at all. I just means you don't know if it is forged or not. If the domain administrator did not set up SPF to designate the outbound server AND that server is different than the inbound server identified by the MX record, then it's their loss to not get a delivery failure notification for otherwise legitimate mail (they can fix it).
BTW, it is not well known, but mail servers testing for email address validity could do an "MX test". An "MX test" never gives a negative because the domain involved may be using a separate outgoing server. But, if the mail is in fact coming from a server identified in the MX records, that can be treated as a positive for the purpose of bouncing. This is a usable test if SPF records are lacking for the email address domain.
You are on the right track but there is more. Yes, higher contrast is better than lower contrast. But how this works with color is complicated.
One big issue is that the eye is not perfect optically. It cannot focus all colors at the same focal plane. Just how well it does varies by individual and the optical conditions of their eyes, and the quality of corrective lenses (which usually make it worse with respect to the ability to simultaneously focus all colors).
An important factor to consider here is which color or colors the difference is at the edge being focused on. For example in the "hot dog" pattern that has been mentioned in a reply here, the difference is actually in green. If the red level of the yellow part is exactly the same as the level of the pure red part, then all the difference is in green and this is an issue of green contrast. Yellow on red like this is essentially the same as green on black... except that the extra red light with yellow on red causes the iris to close down more than the darker green on black would.
I find blue to be the worst to focus with. That may be because my sources of blue light are not sufficiently narrow band in the spectrum. Being spread out over the spectrum, it basically comes in fuzzy. Blue is also lower in contrast.
Green (be it green on black or yellow on red or even cyan on blue) is better.
Red seems to be the best in terms of focusing a sharp defining edge. You get red contrast with red on black or yellow on green or magenta on blue.
Unfortunately, effective contrast goes down when extra light is added in other colors. So you have to find a balance trading off the sharpness of the edge vs. the contrast. I've found a good compromise in orange on dark green (the level of green in the orange is the same value as the green background). Think of the orange in a neon sign on the green felt of a pool table. Then when I need to highlight something, I shift over to pink on cyan... basically add the same level of some blue to both the orange and the dark green.
A related issue is light quality when reading a book or newspaper. Usually we are stuck with black letters on white paper. The consideration is then what type of light. I find that incandescent light, or sunlight, works nearly best for me for long term reading. Fluorescent lighting is worse. Ironically, I find high pressure sodium vapor light is about as good as, and sometimes somewhat better than, incandescent light.
To understand this, look at the spectrum. Incandescent light has a fairly even level through all light wavelengths. This makes those black on white edges a bit fuzzy. But fluorescent light has two narrowband peaks at a red and green wavelength (the blue is broader). This can make the text edge sharper... twice. The eye ends up with two contrast edges. I believe this increases the eyestrain by causing the focus to be constantly jumping in and out to alternate the focus on the two different edges. It's a very small adjustment, but it is there at least for me. With incandescent light, it just settles in the middle of the fuzzy range and doesn't change much. And this is affected by how much light there is, which dictates how small the iris becomes. Higher light levels with a smaller iris won't change the effect from fluorescent as much as for incandescent, since with fluorescent the two contrast edges are already rather sharp due to the two narrowband spectral peaks. But for incandescent, the high light level helps (up to the point that intensity is too stressful).
This is why I believe we still need to keep some incandescent lighting around for reading and other close/fine work for long periods of time. I get a headache when working on things I need to look at closely when doing so under fluorescent light. The onset is about 25 to 45 minutes. I don't get the headaches under incandescent. And I have verified that the flicker is not the cause. White LEDs
If certain bits of the protocol are designed in different ways, the shenanigans are harder. For example use a key exchange for certain state changes (cheaper than encrypting everything). They would have to handle each of the UDP based protocols differently if each has its own session management logic.
The IPs doing this shit are the end user addresses for home and office computers that are no different than all the other end users that use Gmail. They could block an IP, but eventually that IP will be used by someone else who is a legitimate and secure Gmail user. They are better off closing accounts that send spam. But Google isn't doing that (based on having seen spam from the very same user I reported to them as a spammer 2 weeks prior). If they do decide to pursue the user of the IP, once they get past the legal roadblocks of getting the identity out of the ISP (while doing this for 100,000 such IPs at the same time), all they get is some stupid loser who has an infected Windows box being used as part of the botnet. They can get this machine cleaned up, but they aren't anywhere near the real culprits.
What Google needs to do is segregate all users that are new since the crack (they know when it was, because they can see a spike in new user signups from random end user IP addresses). In the mean time, close down direct signups and fallback to the invite system only allowing the old users to send invitations. Re-engineer the CAPTCHA system to at least temporarily thwart the signups before bringing that back online. For all new users, run all their outgoing mail through the same filters that are used for incoming mail. Mail that can't be sent, put it in a new folder type for "blocked outgoing". The user has to pass a new CAPTCHA per each message to do a re-deliver around the filtering (or just rephrase and send a new one). And limit the number of these to 3 per day (although this may not do much good since the botnet may only be doing this much or less over a million accounts).
Of course these providers have improved their services. The problem is they have not improved them quite as well as they could have. And a lot of the ways they are "improving" them focues on ways to extract more money out of the customers, rather than providing a service that increases the value to customers. Would you expect any less of a business motived exclusively by revenue growth?
One big problem is that these companies are sitting on "gold mines" that were established for them (or for the company they bought out) through exclusive monopolies on the infrastructure. Although they invested in this infrastructure, they benefitted from government guarantees of an exclusive regulated monopoly. Now, with most of the regulation lifted, they are using this infrastructure they "inherited" to gouge customers (as opposed to supplying a regulated service that would be sufficient to pay back the investment). At the same time, they know competitors are basically unable to overbuild, not because of any exclusivity, but merely because it doesn't make sense to invest in another infrastructure (because the new builder would know they could at best get 50% of the customer base).
IMHO, the people have a "lien" in that infrastructure because of having guaranteed the exclusivity in the past. That "lien" should be exercised in the form of maintaining a level of regulation on the infrastructure that permits fair, equal, and neutral use, as well as pricing that is fair and does not gouge consumers.
It's bad enough that we have such a poor service from companies like several cable companies and many telephone companies in terms of how the internet layer services are rendered over the infrastructure. If we had fair access to the infrastructure by other providers of internet layer service, then competition would at least allow someone that does a better job to offer services, if not encourage others to do better to keep customers happy.
Long ago, AT&T was broken up between local service and long distance service because at the time it was seen that long distance would be better provided through competition. This was in fact correct and it did improve long distance through better offerings, better pricing, etc. But the split wasn't quite right in terms of today's needs. What we need today for telephone and cable service is a split that separates the ownership and management of the infrastructure, and the companies that can offer services over that infrastructure. We are already seeing this point of split taking place in many areas for electrical power service. In many areas, people can contract to get their electric power from any of a number of power providers (some that actually generate power, and some that merely buy it on the generation market). This has opened up options we would not have otherwise even seen, such as greener power preferences.
What I propose is that governments in all areas support (even financially) the development of an all new fiber based infrustructure. Instead of this being a branched fiber structure like Verizon FiOS, this infrastructure install a minimum of 4 fibers from each home (maybe more for businesses) all the way to a central office connection facility. This infrastructure, including the central office facilities, will be owned by the local government (or liened or otherwise regulated by it), and operated in a fully fair and neutral way. The home owner/renter can then acquire services from any company prepared to connect service to them through one or more of these fiber circuits. Legacy/incumbent providers of information/entertainment service like Comcast, and telco service like Verizon, can make use of this by being one of these providers. They would be able to offer any services they want through that fiber connection (which is plenty sufficient for a huge amount of service on just 1 of the 4 fibers). They could even choose to subcontract
If we use UDP for all traffic (yes, this would require a lot of development), then it would be a lot harder to reset this traffic. Or we could go forward instead of backwards and switch to SCTP but that runs the risk of Comcast adapting to that and resetting SCTP sessions. The ultimate would be to use IPsec, along with a better algorithm to handle lost packets (instead of doubling the time delay between each resend, just increment it slightly, with a 1 minute ceiling).
Then you aren't opt-ing out, anymore. That's why the whole concept of "opt-out" is wrong. People should never be required to opt-out. They should have the option to opt-in by taking that step, or not opt-in by not taking a step at all. This is why you, if a voter in the USA, should for "anyone but (Mc)Cain".
I've experiemented with IPsec. One problem I find is that I have to choose a policy per IP (I can make that be for all IPs, or select IPs, of course). So far, I have been unable to make an IPsec configuration that merely allows the peers to decide if they want to use IPsec or not.
The idea is to make a transitional web server. Anyone who does not have IPsec install, enabled, configured, or determined for this web server IP, can still access the web sites therein in the usual unencrypted way. However, anyone who does have IPsec set up to force encryption, that will work, too, and all their traffic will then be encrypted.
Only after a sufficiently large number of people are doing IPsec voluntarily, then I can set the web server to require IPsec for all. And maybe this step isn't even needed.
Frightened... maybe. If CNN is making reference to them so their ads show up in the CNN page, then at least one of my worries is gone (that being they would inject their ads on sites not making money from them, such as substituting an entire image for some image ad request that should go to the user requested site).
That they collect the info about me outside of the scope of web sites I choose to visit based on agreements between those sites and the advertisers that aggregate the data from among many web sites, is still the big concern. I don't have much concern about the web sites themselves doing it because I can always choose to avoid them if I think they are doing something bad (or, rather, they have to decide if too many people will avoid them in their decision to do it or not).
I'm behind on upgrading my Firefox version, so I don't know if they've added such security features. But one thing I think would be useful is a proximity test on objects (images, stylesheets, javascript, etc) referenced by a given page. The proximity test is a bit complex but works like this. Get the IP address of the main page HTML server. Get the IP address of the server delivering the referenced object. If they are the same, don't worry. If they are in the same/24 subnet, don't worry much. If they are not in the same subnet, then lookup up the NS record (not the PTR record) of the IN-ADDR.ARPA equivalent for each IP. If the NS records are the same, don't worry too much. If they are different, suppress the object. But this is more of an ad blocker.
Ultimately we need a spy blocker. That's where HTTPS and IPsec can come in handy. I just need to figure out how to enabled IPsec in such a way that anyone can communicate either way at their choice.
If these are the ISPs (as opposed to the visited web sites) doing the spying, then how are the advertising companies involved supposed to deliver the content? Are they going to use the same "deep packet" method to inject the advertising? If the advertising delivery is away from that deep packet inspection, then how do they identify which user was interested in penis enlargement products vs. which user was interested in replica watches? Or are the ISPs going to lock-in the IP address, now?
Yes, services! Services really do provide real cash flow. In fact, business like service so much they often prefer to convert to that model when they can. Service is an incremental cash flow that keeps on coming. Selling software is a one time sale.
Sure, you can sell upgrades. But you can also sell maintenance, management, and consulting service. You can even sell installation service (unless you make software that installs itself).
The risk of service work is not this lack of ramp up that you claim. Instead, the real risk is a higher level of competition. That is, you'll have a lot of others who can provide the same kind of service, including support service for open source software. Another risk is that if you identify a need to make improvements, you won't invest money in that effort since you can't use it as market leverage. By contrast, a service can be to sell the work of customizing the software to meet individual client needs.
More often, you'll find out that power-tripping assholes are attracted to those jobs the same way child molesters are attracted to schoolgrounds and bank robbers are attracted to banks.
And those who can't make the cut acting out their power-trip over adults will eventually find their way to being school board administrators and school principals where the targets are easier.
Hynix, Micron, and Nanya should sue the standards committee over this. Maybe that would force all standards committees to proactively get every participant to sign over all patent rights to participate in the standards process. Those companies that want to not do that would have to sit out.
For every analog channel they drop, they gain back 2 decent or 3 crappy HD channels. Or maybe they could do 2 half-way decent HD and 1 SD channel. And, yes, there is a requirement to provide analog until 2012. But they can meet that requirement by supplying a converter box that outputs analog (at no additional cost for basic customers). The question is, is the cost of providing that converter box greater than the benefit of the extra channels?
Even the mid range digital SLR cameras are moving up the number of bits per sample. For example Canon's new EOS 450D
has 14 bits per color for RAW (scroll down the the compression specs). And this is just the initial recording, not the editing steps where more is needed to avoid the errors. I wonder what their next high end camera will be like.
"Do not ascribe to malice that which can be explained by incompetence". But given the track record of sooooooooo many big corporations, I think this bit of wisdom just doesn't apply here. Think about the lawyers which had to approve this. Malice or incompetence? Both?
Slashdot Mirror
You're missing a unit. Is that per month?
OK, so if I get 12 mbps speed, which is what is needed for high definition video in real time, you're going to charge $360 per month? It seems like your network isn't ready for 2009. Oh wait, it's still 2008. I guess you have another year.
Just because all your customers already pay so much for so little now, because that's all you offer, does not justify keeping things at this level in the future.
You mention a measly 1.5 mbps. Someone wanting to download a 2 hour high definition movie is going to have to be doing the download for 16 hours to get it all. At 1.5 mbps, it will take a full solid month to download 486 gigabytes (not figuring in overhead and such). Are you going to cut off customers that use it?
$30 for a megabyte is a bit high.
White LEDs have the same problem as fluorescent. They are both based on fluorescent material driven by deep blue and/or UV light. Some LEDs are a combination of red, green, and blue LED dies. Having lots of distinct wavelengths might well be the solution. They could do that not only in LED (I've seen 22 different wavelengths available) but also in a new fluorescent formula. The catch is that it is expensive. But I suspect the LED field will be able to achieve this before the fluorescent field.
I was not aware of the lower density of cones. That would explain it, and the fact that blue "seems" darker.
... Canada has just begun to beef up the military defenses on its long southern border.
I wish there was a ROTFLMAO mod.
NDRs are evil if sent to the wrong address. You need to send them to THE SENDER ... not some victim the spammer has fingered. In the case of spam, the spammer is THE SENDER (regardless of what he puts on as the return address). So that means NDRs are not evil in cases of legitimate mail where the sender address is not a forgery. SMTP is flawed in the sense that if fails to provide a means to correctly identify the sender. Better do your SPF and other tests to be sure. Or just do the "safe harber" and do rejects during the SMTP session.
You can't be right because this is still creating an abusive situation. Keep in mind that the major issue with spam has nothing to do with the contents. So the fact that the backscatter does not include the spammer message just means the spammer can't make use of it to pass the message. But it still causes the same problems as spam, which is bogged down mail servers.
I have had one of my hosted email addresses used by a spammer in a large spam run of several million or so. My server once got hit by about 250,000 pieces of backscatter in just 3 days. About 6000 mail servers were the abuse vectors in this. They all got blocked. They will stay blocked until they can prove they fixed the issue. Google was NOT one of them (most were in Russia and the spam run was in Russian). But Google could have been one of them if they were doing then what they are reportedly doing today.
Sending back a bounce message after the fact is abuse ... if the bouncing server did not verify that the return email address it is using designates the sending server as appropriate for it. Google may well be applying that test (I see references to SPF in the headers on Gmail). So they could well be limiting backscatter to situations where the forged email address is one that matches the sending address (for example a botnet infected machine on Comcast forges a different Comcast user when sending mail through Comcast's outbound servers over to Gmail).
But any mail server not applying the appropriate test could be an abusive (and therefore evil) server. The correct logic is to never send a bounce message unless you have an unambiguous positive verification of the address via SPF, MX, etc. The RFC requires sending the delivery notification to the sender. The sender is the spammer. But you don't have the spammer's address; you just have some forged address. So you cannot send it correctly. So just don't send it. FYI: do not assume that a lack of negatives on that email address means it is a positive. It does not mean that at all. I just means you don't know if it is forged or not. If the domain administrator did not set up SPF to designate the outbound server AND that server is different than the inbound server identified by the MX record, then it's their loss to not get a delivery failure notification for otherwise legitimate mail (they can fix it).
BTW, it is not well known, but mail servers testing for email address validity could do an "MX test". An "MX test" never gives a negative because the domain involved may be using a separate outgoing server. But, if the mail is in fact coming from a server identified in the MX records, that can be treated as a positive for the purpose of bouncing. This is a usable test if SPF records are lacking for the email address domain.
You are on the right track but there is more. Yes, higher contrast is better than lower contrast. But how this works with color is complicated.
One big issue is that the eye is not perfect optically. It cannot focus all colors at the same focal plane. Just how well it does varies by individual and the optical conditions of their eyes, and the quality of corrective lenses (which usually make it worse with respect to the ability to simultaneously focus all colors).
An important factor to consider here is which color or colors the difference is at the edge being focused on. For example in the "hot dog" pattern that has been mentioned in a reply here, the difference is actually in green. If the red level of the yellow part is exactly the same as the level of the pure red part, then all the difference is in green and this is an issue of green contrast. Yellow on red like this is essentially the same as green on black ... except that the extra red light with yellow on red causes the iris to close down more than the darker green on black would.
I find blue to be the worst to focus with. That may be because my sources of blue light are not sufficiently narrow band in the spectrum. Being spread out over the spectrum, it basically comes in fuzzy. Blue is also lower in contrast.
Green (be it green on black or yellow on red or even cyan on blue) is better.
Red seems to be the best in terms of focusing a sharp defining edge. You get red contrast with red on black or yellow on green or magenta on blue.
Unfortunately, effective contrast goes down when extra light is added in other colors. So you have to find a balance trading off the sharpness of the edge vs. the contrast. I've found a good compromise in orange on dark green (the level of green in the orange is the same value as the green background). Think of the orange in a neon sign on the green felt of a pool table. Then when I need to highlight something, I shift over to pink on cyan ... basically add the same level of some blue to both the orange and the dark green.
A related issue is light quality when reading a book or newspaper. Usually we are stuck with black letters on white paper. The consideration is then what type of light. I find that incandescent light, or sunlight, works nearly best for me for long term reading. Fluorescent lighting is worse. Ironically, I find high pressure sodium vapor light is about as good as, and sometimes somewhat better than, incandescent light.
To understand this, look at the spectrum. Incandescent light has a fairly even level through all light wavelengths. This makes those black on white edges a bit fuzzy. But fluorescent light has two narrowband peaks at a red and green wavelength (the blue is broader). This can make the text edge sharper ... twice. The eye ends up with two contrast edges. I believe this increases the eyestrain by causing the focus to be constantly jumping in and out to alternate the focus on the two different edges. It's a very small adjustment, but it is there at least for me. With incandescent light, it just settles in the middle of the fuzzy range and doesn't change much. And this is affected by how much light there is, which dictates how small the iris becomes. Higher light levels with a smaller iris won't change the effect from fluorescent as much as for incandescent, since with fluorescent the two contrast edges are already rather sharp due to the two narrowband spectral peaks. But for incandescent, the high light level helps (up to the point that intensity is too stressful).
This is why I believe we still need to keep some incandescent lighting around for reading and other close/fine work for long periods of time. I get a headache when working on things I need to look at closely when doing so under fluorescent light. The onset is about 25 to 45 minutes. I don't get the headaches under incandescent. And I have verified that the flicker is not the cause. White LEDs
Why do you need upconverting? Your HD display won't show SD video full screen?
If certain bits of the protocol are designed in different ways, the shenanigans are harder. For example use a key exchange for certain state changes (cheaper than encrypting everything). They would have to handle each of the UDP based protocols differently if each has its own session management logic.
The IPs doing this shit are the end user addresses for home and office computers that are no different than all the other end users that use Gmail. They could block an IP, but eventually that IP will be used by someone else who is a legitimate and secure Gmail user. They are better off closing accounts that send spam. But Google isn't doing that (based on having seen spam from the very same user I reported to them as a spammer 2 weeks prior). If they do decide to pursue the user of the IP, once they get past the legal roadblocks of getting the identity out of the ISP (while doing this for 100,000 such IPs at the same time), all they get is some stupid loser who has an infected Windows box being used as part of the botnet. They can get this machine cleaned up, but they aren't anywhere near the real culprits.
What Google needs to do is segregate all users that are new since the crack (they know when it was, because they can see a spike in new user signups from random end user IP addresses). In the mean time, close down direct signups and fallback to the invite system only allowing the old users to send invitations. Re-engineer the CAPTCHA system to at least temporarily thwart the signups before bringing that back online. For all new users, run all their outgoing mail through the same filters that are used for incoming mail. Mail that can't be sent, put it in a new folder type for "blocked outgoing". The user has to pass a new CAPTCHA per each message to do a re-deliver around the filtering (or just rephrase and send a new one). And limit the number of these to 3 per day (although this may not do much good since the botnet may only be doing this much or less over a million accounts).
Of course these providers have improved their services. The problem is they have not improved them quite as well as they could have. And a lot of the ways they are "improving" them focues on ways to extract more money out of the customers, rather than providing a service that increases the value to customers. Would you expect any less of a business motived exclusively by revenue growth?
One big problem is that these companies are sitting on "gold mines" that were established for them (or for the company they bought out) through exclusive monopolies on the infrastructure. Although they invested in this infrastructure, they benefitted from government guarantees of an exclusive regulated monopoly. Now, with most of the regulation lifted, they are using this infrastructure they "inherited" to gouge customers (as opposed to supplying a regulated service that would be sufficient to pay back the investment). At the same time, they know competitors are basically unable to overbuild, not because of any exclusivity, but merely because it doesn't make sense to invest in another infrastructure (because the new builder would know they could at best get 50% of the customer base).
IMHO, the people have a "lien" in that infrastructure because of having guaranteed the exclusivity in the past. That "lien" should be exercised in the form of maintaining a level of regulation on the infrastructure that permits fair, equal, and neutral use, as well as pricing that is fair and does not gouge consumers.
It's bad enough that we have such a poor service from companies like several cable companies and many telephone companies in terms of how the internet layer services are rendered over the infrastructure. If we had fair access to the infrastructure by other providers of internet layer service, then competition would at least allow someone that does a better job to offer services, if not encourage others to do better to keep customers happy.
Long ago, AT&T was broken up between local service and long distance service because at the time it was seen that long distance would be better provided through competition. This was in fact correct and it did improve long distance through better offerings, better pricing, etc. But the split wasn't quite right in terms of today's needs. What we need today for telephone and cable service is a split that separates the ownership and management of the infrastructure, and the companies that can offer services over that infrastructure. We are already seeing this point of split taking place in many areas for electrical power service. In many areas, people can contract to get their electric power from any of a number of power providers (some that actually generate power, and some that merely buy it on the generation market). This has opened up options we would not have otherwise even seen, such as greener power preferences.
What I propose is that governments in all areas support (even financially) the development of an all new fiber based infrustructure. Instead of this being a branched fiber structure like Verizon FiOS, this infrastructure install a minimum of 4 fibers from each home (maybe more for businesses) all the way to a central office connection facility. This infrastructure, including the central office facilities, will be owned by the local government (or liened or otherwise regulated by it), and operated in a fully fair and neutral way. The home owner/renter can then acquire services from any company prepared to connect service to them through one or more of these fiber circuits. Legacy/incumbent providers of information/entertainment service like Comcast, and telco service like Verizon, can make use of this by being one of these providers. They would be able to offer any services they want through that fiber connection (which is plenty sufficient for a huge amount of service on just 1 of the 4 fibers). They could even choose to subcontract
If we use UDP for all traffic (yes, this would require a lot of development), then it would be a lot harder to reset this traffic. Or we could go forward instead of backwards and switch to SCTP but that runs the risk of Comcast adapting to that and resetting SCTP sessions. The ultimate would be to use IPsec, along with a better algorithm to handle lost packets (instead of doubling the time delay between each resend, just increment it slightly, with a 1 minute ceiling).
Then you aren't opt-ing out, anymore. That's why the whole concept of "opt-out" is wrong. People should never be required to opt-out. They should have the option to opt-in by taking that step, or not opt-in by not taking a step at all. This is why you, if a voter in the USA, should for "anyone but (Mc)Cain".
I've experiemented with IPsec. One problem I find is that I have to choose a policy per IP (I can make that be for all IPs, or select IPs, of course). So far, I have been unable to make an IPsec configuration that merely allows the peers to decide if they want to use IPsec or not.
The idea is to make a transitional web server. Anyone who does not have IPsec install, enabled, configured, or determined for this web server IP, can still access the web sites therein in the usual unencrypted way. However, anyone who does have IPsec set up to force encryption, that will work, too, and all their traffic will then be encrypted.
Only after a sufficiently large number of people are doing IPsec voluntarily, then I can set the web server to require IPsec for all. And maybe this step isn't even needed.
Frightened ... maybe. If CNN is making reference to them so their ads show up in the CNN page, then at least one of my worries is gone (that being they would inject their ads on sites not making money from them, such as substituting an entire image for some image ad request that should go to the user requested site).
That they collect the info about me outside of the scope of web sites I choose to visit based on agreements between those sites and the advertisers that aggregate the data from among many web sites, is still the big concern. I don't have much concern about the web sites themselves doing it because I can always choose to avoid them if I think they are doing something bad (or, rather, they have to decide if too many people will avoid them in their decision to do it or not).
I'm behind on upgrading my Firefox version, so I don't know if they've added such security features. But one thing I think would be useful is a proximity test on objects (images, stylesheets, javascript, etc) referenced by a given page. The proximity test is a bit complex but works like this. Get the IP address of the main page HTML server. Get the IP address of the server delivering the referenced object. If they are the same, don't worry. If they are in the same /24 subnet, don't worry much. If they are not in the same subnet, then lookup up the NS record (not the PTR record) of the IN-ADDR.ARPA equivalent for each IP. If the NS records are the same, don't worry too much. If they are different, suppress the object. But this is more of an ad blocker.
Ultimately we need a spy blocker. That's where HTTPS and IPsec can come in handy. I just need to figure out how to enabled IPsec in such a way that anyone can communicate either way at their choice.
If these are the ISPs (as opposed to the visited web sites) doing the spying, then how are the advertising companies involved supposed to deliver the content? Are they going to use the same "deep packet" method to inject the advertising? If the advertising delivery is away from that deep packet inspection, then how do they identify which user was interested in penis enlargement products vs. which user was interested in replica watches? Or are the ISPs going to lock-in the IP address, now?
Yes, services! Services really do provide real cash flow. In fact, business like service so much they often prefer to convert to that model when they can. Service is an incremental cash flow that keeps on coming. Selling software is a one time sale.
Sure, you can sell upgrades. But you can also sell maintenance, management, and consulting service. You can even sell installation service (unless you make software that installs itself).
The risk of service work is not this lack of ramp up that you claim. Instead, the real risk is a higher level of competition. That is, you'll have a lot of others who can provide the same kind of service, including support service for open source software. Another risk is that if you identify a need to make improvements, you won't invest money in that effort since you can't use it as market leverage. By contrast, a service can be to sell the work of customizing the software to meet individual client needs.
And those who can't make the cut acting out their power-trip over adults will eventually find their way to being school board administrators and school principals where the targets are easier.
Hynix, Micron, and Nanya should sue the standards committee over this. Maybe that would force all standards committees to proactively get every participant to sign over all patent rights to participate in the standards process. Those companies that want to not do that would have to sit out.
For every analog channel they drop, they gain back 2 decent or 3 crappy HD channels. Or maybe they could do 2 half-way decent HD and 1 SD channel. And, yes, there is a requirement to provide analog until 2012. But they can meet that requirement by supplying a converter box that outputs analog (at no additional cost for basic customers). The question is, is the cost of providing that converter box greater than the benefit of the extra channels?
Express means you more quickly get to the end of the list of things it can do.
Even the mid range digital SLR cameras are moving up the number of bits per sample. For example Canon's new EOS 450D
has 14 bits per color for RAW (scroll down the the compression specs). And this is just the initial recording, not the editing steps where more is needed to avoid the errors. I wonder what their next high end camera will be like."Do not ascribe to malice that which can be explained by incompetence". But given the track record of sooooooooo many big corporations, I think this bit of wisdom just doesn't apply here. Think about the lawyers which had to approve this. Malice or incompetence? Both?