First, that's an argument for mandatory dummy loads or steam bypasses to put proportionally less steam through the turbines. Go read page 506 of "Probabilistic safety assessment in the chemical and nuclear industries" (1999). Also, you can run power plants at less than 100% capacity. The amount of power (in watts) produced depends on how far those control rods are pulled out. In short, this is a readily solved problem that is well understood.
Second, regarding the bicycle lights, no it won't. That would only happen if the lights were wired in series instead of in parallel and if the front light had a failover device that caused it to short in the event that the bulb blows (in which case your effective operating voltage went up). The electrical lights in your house don't blow because you turn off all the switches but one, do they? No difference. With few exceptions, a load draws exactly as much power as it needs and no more.
No, those scans are terrible. They barely fill a 3" x 4" area on my laptop monitor. If I zoom in far enough to actually see stuff, they're very, very fuzzy. It looks like they were shot with a digital camera, then cropped. Stick it on a flatbed scanner and provide scans with at least 600 DPI resolution. Then we'll talk. Until then, it hurts my head just looking at it.
Right. I was thinking the same. I notice, for example, TFRNE appears twice, and then there's the pairing of RLSE and CRLSE that are likely indicative of two words differing only by the addition of one letter, though not necessarily at the beginning.
I think that P1 or that blob of not-quite-readable text in the upper right (ALSM?) is probably some sort of crypto key.
I wish the pictures were large enough and crisp enough to read more easily. Just looking at it for a minute gave me a headache.
So lower the control rods to the minimum level that allows for power generation to continue. Nuclear power is not an all or nothing thing. Sure, a partial scram isn't great for the reactor, but neither is a full scram.
To give a car analogy, the earthquake is like somebody shooting out the sidewall in your tire. A damaged run-flat tire could eventually blow out if you drive it at 65 MPH down the highway, particularly with a bullet hole in the sidewall. However, that doesn't mean you should turn off your car and get out of the vehicle rather than driving at a slower speed to the nearest tire shop.
Not sure how a runaway critical reaction is a better outcome than the current situation.
Why are you assuming that it would have been a runaway reaction? Last I heard, they still had no idea about the condition of the pipes. I'm a little concerned about that, too, because that means that the plant has woefully inadequate sensors... but I digress.
What happens if the quake knocks the control rods out of alignment
That's why you lower the control rods far enough to make that a nonissue, but not far enough for a complete scram. I'm not sure about this particular design, but AFAIK most reactors do support partial scrams and running the plant at a reduced output.
Even IF the tsunami didnt wipe out the generators AND POWERLINES (remember, it took them over a week to run a new powerline to the plant) between the plants, and one stayed operational, running pumps with nothing in them to pump does little good (cracked cooling line/evaporated coolant/steam releases), as does powering broken pumps or pumping coolant through broken lines (they still arent sure the pumps or lines are operational in some of the buildings).
First, the power lines that were damaged by the tsunami, as far as I am aware, are all lines outside the plant, not lines between parts of the plant itself, so to the extent that the plant could produce a reduced amount of power, there's nothing inherently preventing it from using that power to cool itself. They had to run lines to the plant only because there was no way to produce power at the plant. More on this later.
Second, according to the stories I've read, the cooling systems for the reactors and ponds were successfully keeping the scrammed reactors cool until the emergency batteries ran down. However, the diesel generators failed to kick in due to their missing fuel tanks, and when the batteries finally gave out, all hell broke loose. So if even a single reactor had stayed up and running enough to power those pumps (which definitely does not mean running full tilt, particularly given the lack of adequate load, just not in a full scram state), then there's no reason to believe that the meltdown would have occurred.
So no, a full scram wasn't "the 100% correct thing to do". A 98% partial scram would have done the same thing as far as ensuring that a full scram could be safely performed if necessary, but without causing the complete loss of all local power generation capacity. A full scram is by its very nature something of a last resort anyway, as it precludes a speedy restart.
What went wrong, besides under-designing the seismic and tsunami resistance of the plant, was placing the generators in a position where a tsunami could wipe them out.
What went wrong was placing the plant in a position where a tsunami could wipe it out. Having batteries that were insufficient to get through the critical first few days after a scram was also a pretty serious flaw. Placing the fuel tanks for the diesel generators above ground outdoors was just the straw that broke the camel's back. Note that the generators were fine. The fuel tanks, however, were swept away.
Even more disturbing is that even without the seismic scram hardware (recently added, IIRC), this type of reactor would have automatically scrammed in the event of loss of offsite power (a LOOP scram) anyway because the cooling system requires outside power in order to operate. That's a design flaw in my book (both that there is insufficient passive cooling and that the generation system is incapable of running in standalone mode without external power to regulate the turbine speed, thus necessitating the use of either external power or an emergency generator for basic cooling).
More to the point, this is a design flaw that was identified way back in 1979, and the NRC pointed to this very scenario as a potential cause for a meltdown a
But the level of hysteria and misinformation of public is so high with nuclear, I think Fukushima will retard nuclear reactor development and construction by at least a decade.
There's some irony for you. A bunch of reactors that were scheduled to be dismantled whose lives had to be extended beyond their design lifetime are causing a bunch of frightened people who lack sufficient technical knowledge to make informed decisions to protest against newer, safer nuclear plant construction, thus resulting in more old nuclear plants with outdated designs having to stay in operation well beyond their design lifetime, thus increasing the risk to everyone.
What do you think how "modern" the currently built ones (the few ones that are... finland one is the only one which is built right now) are in 40 years? Just as outdated....
Not necessarily. At some point, technology becomes "good enough" and stops evolving, at least until new serious flaws are discovered.
Ironically, this whole crisis was caused because they did precisely that—the reactors shut down automatically for safety reasons, and then they had no power with which to keep the pumps running because the diesel generators were underwater. Had pretty much any one those reactors not automatically scrammed, it is likely that things would be in better shape than they are now.
And what folks should take away from all this is that reactors should auto-scram only when they detect a coolant leak, not because of an earthquake that merely might cause a coolant leak. Or at least that's what should happen for older reactors like these that require active cooling in a scrammed state.
No, scratch that. The takeaway should be that reactors that require active cooling in a scrammed state are fundamentally unsafe in a seismic zone and should be replaced with newer reactors as soon as possible.
I exclusively use case-sensitive HFS+ partition on all my Macs. In all the years of doing so, I've only encountered three apps that don't work out of the box:
Steam
Adobe's Creative Suite
MakeMusic Finale
Of those, Finale can be made to work with a couple of symlinks, and the rest can be hacked to work by sticking them on a case-insensitive HFS+ disk image. (I don't know what the Steam folks were thinking. Unless they have fixed it recently, it looks to me like they have code in their app that deliberately converts every path to lowercase.... It's like something a first-year CS student would do....)
The good news is that iOS is case sensitive, which means that any of those Mac developers who want to port any part of their apps to iOS will be forced to fix such bugs. Hopefully, they will take what they learn and apply it to their Mac products, too. It's really not that difficult to fix the bugs once they have been discovered. The hard part is testing every part of large apps in the first place.
Mod parent up. Control rods do not prevent reactions. They merely control the reactions. That's why they're called control rods. They limit reactions such that once the fission byproducts stop breaking down, the reaction rate falls to almost nothing because you don't have enough mass in one place.
When the control rods are first inserted, there is still a lot of heat being produced. For the first few days after a scram, the reactor requires significant cooling to get rid of that heat. When you don't have enough cooling, you get what's happening at the Fukushima plants.
And there's good reason to be concerned about the low frequency (217 Hz) pulses produced by GSM phones (and other TDMA devices) as they turn on and off their transmitters rapidly. As frequency decreases, skin depth increases, which means that I would expect the changing field to reach deeper into your body than with a high frequency signal. It is the low frequency nature of this signal that makes it hard to shield against, resulting in interference in nearby devices.
The GSM cell phones produce signals whose effective frequency is relatively close in frequency to the 50Hz/60Hz associated with high tension lines, etc., and there's much debate about the safety of those for the same reason. Although I'm pretty sure that higher frequencies have minimal effect on the human body, there has been inadequate study on low frequency EMI to make the same claim.
I begin to have serious doubts about the "science" behind some of these studies when people disclaim a factor of 2 difference in childhood leukemia rates as not statistically significant, and claim that you need a factor of 6 to be statistically significant. A factor of two not being statistically significant means your sample size was too small, and that you should do the test over with a larger sample size to see if you can reproduce the effect.... It certainly does not mean that you should claim that there is no effect.
Not really, if you worked in multiple countries you could do the same thing.
No, you are expected to pay U.S. income tax even while working overseas. There's an exemption for middle class levels of income and below, but everything above that dollar limit is taxed. Only corporations can get away with paying no tax on arbitrarily large profits made overseas.
And that's okay. I would be fine if any ISP said that they would throttle my speed if I used too much bandwidth in a month, so long as those limits were spelled out clearly. What I'm not fine with is them spanking me with a huge penalty for something that I may or may not have any control over.
Given that I run servers out of my home, I was considering switching to U-Verse because of the higher upstream bandwidth I could get, but I have since scrapped that plan. I can't afford to find out that one of my websites got slashdotted or was being attacked by a DDOS, and suddenly I have a $500 bandwidth bill. Nor is the alternative—having to disconnect my server for the rest of the month—an acceptable alternative, either. Remember that you can hit the AT&T bandwidth limit in just a couple of days of continuous use at full throttle.
So if I went with U-Verse, I would have no choice but to have a second network connection to switch my server over to if my U-Verse connection was at risk of hitting me with overage charges. The only way to do that is to keep my DSL service, since there is no Cable internet available here. That's a problem; I can't keep DSL if I add U-Verse service. Thus, I can't be a U-Verse customer.
For anybody not on U-Verse, you should do what I did when the first story broke: call, write, or email AT&T. Tell them that you were going to switch to their service, but then you found out about the bandwidth overage charges and you changed your mind. If enough people did this, AT&T would pull their heads out of their asses, but it would take at least a few tens of thousands of people complaining. While you're at it, tell your friends to do the same. This is even better if you can honestly tell them that you have a business-class connection as I do. AT&T needs businesses who run servers to help balance out their bandwidth with other backbone providers. Without servers, they're screwed.
BTW, to the AT&T local wire service tech who put the indoor DSL splitter in my outdoor box and caused me to lose both phone service and DSL for most of the day today due to rain (this makes two spring rainy seasons in a row with loss of local phone service), thanks. I just rewired my house back to the way I had it before your meddling, with the splitter safely indoors. It's no wonder AT&T is begging for every penny they have when they can't even get something as simple as basic phone wiring right.
And I swear if I get a cold from having to redo your incompetent rewiring job outside in the rain today....
So in the places where HTTPS is most needed to protect people's lives, Microsoft kowtows to pressure from a bunch of soon-to-be-ex Pol Pot dictators to trick people into using unencrypted traffic so that they can be snooped upon?
To everyone in the Middle East, when the revolution is through, remember who your friends were, and remember which large company tried to sell you out, then choose your purchases accordingly. Remember, developing nations have more influence on corporations through their buying power than any nation that is already locked into a particular vendor's products. Just a helpful tip.
To Microsoft, you should be ashamed. No, wait, the other thing. Tried and executed for crimes against humanity. Not to mention treason if Libya is being handled similarly. For shame.
Not without a lot of collateral damage, you can't. There's nothing stopping a single hosting server from serving Whitehouse.gov and Whitehouse.com on the same IP.
I don't know of any CA that carefully checks all the source code of the requestor to ensure it is not malicious before issuing a certificate.
To at least some extent, for app store apps, Apple does, albeit with user testing and static analysis on the binaries rather than examining the source code itself. Not before issuing the cert, mind you, but before allowing it on the store. And until it goes on the store, an app can only be run by devices that the developer owns. So in effect, yes, one CA does do this.
But how many years would you make the cert for? Eventually, all certs expire. That's a real problem with certs.
There's also the problem of how the user would get the certs. Assuming it's electronic, what's stopping somebody from guessing serial numbers and getting access keys for random people's phones? Not sure how practical that would be to exploit, but I'm sure somebody could come up with a way.:-)
As a side note, expiration of certs is a problem that really ought to be fixed. As far as I can tell, the only purposes for cert expiration seem to be A. people who are too lazy to revoke their certs, B. the insane overhead that CRLs represent, and C. so that the cert sales people can make more money, of which B. is irrelevant as long as all your clients support OCSP, and A. could be made irrelevant with a little bit of basic management infrastructure on the part of the CA, leaving C. (greed) as the only reason to continue this stupidity. The key to my door doesn't expire after a year. Why should the SSL key for my server?
And cert expiration is particularly problematic for application signing:
Do you continue to trust an app that was signed while the cert was valid after the cert has expired?
How can you tell that the cert was still valid when the app was signed? Can you trust those time stamps? So now you need a time stamping server.
If you don't care when it was signed, how can you prove that the cert wasn't revoked? (Once the cert has expired, it falls off of any CRLs.)
It starts to get more and more complex after that, and most of the problems stem from the inability to create a permanent SSL cert.
So yeah, it's a good idea in principle, but making it work in practice turns out to be more of a pain than it should be, mainly because certs have some serious design flaws.:-)
The thing is, it's not exactly possible to censor a TLD. About all you can do is mandate that ISPs block it on their DNS servers. You can't realistically prevent an entire country from running their own DNS servers, nor from using DNS servers outside the country, nor from using a VPN to tunnel DNS requests, etc. And if there's a market for it, somebody will sell access using just such a VPN.
Slashdot Mirror
First, that's an argument for mandatory dummy loads or steam bypasses to put proportionally less steam through the turbines. Go read page 506 of "Probabilistic safety assessment in the chemical and nuclear industries" (1999). Also, you can run power plants at less than 100% capacity. The amount of power (in watts) produced depends on how far those control rods are pulled out. In short, this is a readily solved problem that is well understood.
Second, regarding the bicycle lights, no it won't. That would only happen if the lights were wired in series instead of in parallel and if the front light had a failover device that caused it to short in the event that the bulb blows (in which case your effective operating voltage went up). The electrical lights in your house don't blow because you turn off all the switches but one, do they? No difference. With few exceptions, a load draws exactly as much power as it needs and no more.
No, those scans are terrible. They barely fill a 3" x 4" area on my laptop monitor. If I zoom in far enough to actually see stuff, they're very, very fuzzy. It looks like they were shot with a digital camera, then cropped. Stick it on a flatbed scanner and provide scans with at least 600 DPI resolution. Then we'll talk. Until then, it hurts my head just looking at it.
Right. I was thinking the same. I notice, for example, TFRNE appears twice, and then there's the pairing of RLSE and CRLSE that are likely indicative of two words differing only by the addition of one letter, though not necessarily at the beginning.
I think that P1 or that blob of not-quite-readable text in the upper right (ALSM?) is probably some sort of crypto key.
I wish the pictures were large enough and crisp enough to read more easily. Just looking at it for a minute gave me a headache.
So lower the control rods to the minimum level that allows for power generation to continue. Nuclear power is not an all or nothing thing. Sure, a partial scram isn't great for the reactor, but neither is a full scram.
To give a car analogy, the earthquake is like somebody shooting out the sidewall in your tire. A damaged run-flat tire could eventually blow out if you drive it at 65 MPH down the highway, particularly with a bullet hole in the sidewall. However, that doesn't mean you should turn off your car and get out of the vehicle rather than driving at a slower speed to the nearest tire shop.
Why are you assuming that it would have been a runaway reaction? Last I heard, they still had no idea about the condition of the pipes. I'm a little concerned about that, too, because that means that the plant has woefully inadequate sensors... but I digress.
That's why you lower the control rods far enough to make that a nonissue, but not far enough for a complete scram. I'm not sure about this particular design, but AFAIK most reactors do support partial scrams and running the plant at a reduced output.
First, the power lines that were damaged by the tsunami, as far as I am aware, are all lines outside the plant, not lines between parts of the plant itself, so to the extent that the plant could produce a reduced amount of power, there's nothing inherently preventing it from using that power to cool itself. They had to run lines to the plant only because there was no way to produce power at the plant. More on this later.
Second, according to the stories I've read, the cooling systems for the reactors and ponds were successfully keeping the scrammed reactors cool until the emergency batteries ran down. However, the diesel generators failed to kick in due to their missing fuel tanks, and when the batteries finally gave out, all hell broke loose. So if even a single reactor had stayed up and running enough to power those pumps (which definitely does not mean running full tilt, particularly given the lack of adequate load, just not in a full scram state), then there's no reason to believe that the meltdown would have occurred.
So no, a full scram wasn't "the 100% correct thing to do". A 98% partial scram would have done the same thing as far as ensuring that a full scram could be safely performed if necessary, but without causing the complete loss of all local power generation capacity. A full scram is by its very nature something of a last resort anyway, as it precludes a speedy restart.
What went wrong was placing the plant in a position where a tsunami could wipe it out. Having batteries that were insufficient to get through the critical first few days after a scram was also a pretty serious flaw. Placing the fuel tanks for the diesel generators above ground outdoors was just the straw that broke the camel's back. Note that the generators were fine. The fuel tanks, however, were swept away.
Even more disturbing is that even without the seismic scram hardware (recently added, IIRC), this type of reactor would have automatically scrammed in the event of loss of offsite power (a LOOP scram) anyway because the cooling system requires outside power in order to operate. That's a design flaw in my book (both that there is insufficient passive cooling and that the generation system is incapable of running in standalone mode without external power to regulate the turbine speed, thus necessitating the use of either external power or an emergency generator for basic cooling).
More to the point, this is a design flaw that was identified way back in 1979, and the NRC pointed to this very scenario as a potential cause for a meltdown a
You're new here, aren't you? Do you think somebody would post a comment like that if he/she had actual child porn?
What we have here is a troll. Nothing more.
There's some irony for you. A bunch of reactors that were scheduled to be dismantled whose lives had to be extended beyond their design lifetime are causing a bunch of frightened people who lack sufficient technical knowledge to make informed decisions to protest against newer, safer nuclear plant construction, thus resulting in more old nuclear plants with outdated designs having to stay in operation well beyond their design lifetime, thus increasing the risk to everyone.
Not necessarily. At some point, technology becomes "good enough" and stops evolving, at least until new serious flaws are discovered.
Ironically, this whole crisis was caused because they did precisely that—the reactors shut down automatically for safety reasons, and then they had no power with which to keep the pumps running because the diesel generators were underwater. Had pretty much any one those reactors not automatically scrammed, it is likely that things would be in better shape than they are now.
And what folks should take away from all this is that reactors should auto-scram only when they detect a coolant leak, not because of an earthquake that merely might cause a coolant leak. Or at least that's what should happen for older reactors like these that require active cooling in a scrammed state.
No, scratch that. The takeaway should be that reactors that require active cooling in a scrammed state are fundamentally unsafe in a seismic zone and should be replaced with newer reactors as soon as possible.
Probably building stuff from source. And that would also explain why the Mini took so much longer. It has a laptop hard drive.
I exclusively use case-sensitive HFS+ partition on all my Macs. In all the years of doing so, I've only encountered three apps that don't work out of the box:
Of those, Finale can be made to work with a couple of symlinks, and the rest can be hacked to work by sticking them on a case-insensitive HFS+ disk image. (I don't know what the Steam folks were thinking. Unless they have fixed it recently, it looks to me like they have code in their app that deliberately converts every path to lowercase.... It's like something a first-year CS student would do....)
The good news is that iOS is case sensitive, which means that any of those Mac developers who want to port any part of their apps to iOS will be forced to fix such bugs. Hopefully, they will take what they learn and apply it to their Mac products, too. It's really not that difficult to fix the bugs once they have been discovered. The hard part is testing every part of large apps in the first place.
See also IRS form 2555-EZ: Foreign Earned Income Exclusion.
Mod parent up. Control rods do not prevent reactions. They merely control the reactions. That's why they're called control rods. They limit reactions such that once the fission byproducts stop breaking down, the reaction rate falls to almost nothing because you don't have enough mass in one place.
When the control rods are first inserted, there is still a lot of heat being produced. For the first few days after a scram, the reactor requires significant cooling to get rid of that heat. When you don't have enough cooling, you get what's happening at the Fukushima plants.
Agreed.
And there's good reason to be concerned about the low frequency (217 Hz) pulses produced by GSM phones (and other TDMA devices) as they turn on and off their transmitters rapidly. As frequency decreases, skin depth increases, which means that I would expect the changing field to reach deeper into your body than with a high frequency signal. It is the low frequency nature of this signal that makes it hard to shield against, resulting in interference in nearby devices.
The GSM cell phones produce signals whose effective frequency is relatively close in frequency to the 50Hz/60Hz associated with high tension lines, etc., and there's much debate about the safety of those for the same reason. Although I'm pretty sure that higher frequencies have minimal effect on the human body, there has been inadequate study on low frequency EMI to make the same claim.
I begin to have serious doubts about the "science" behind some of these studies when people disclaim a factor of 2 difference in childhood leukemia rates as not statistically significant, and claim that you need a factor of 6 to be statistically significant. A factor of two not being statistically significant means your sample size was too small, and that you should do the test over with a larger sample size to see if you can reproduce the effect.... It certainly does not mean that you should claim that there is no effect.
No, you are expected to pay U.S. income tax even while working overseas. There's an exemption for middle class levels of income and below, but everything above that dollar limit is taxed. Only corporations can get away with paying no tax on arbitrarily large profits made overseas.
Could be worse. This is valid C:
And that's okay. I would be fine if any ISP said that they would throttle my speed if I used too much bandwidth in a month, so long as those limits were spelled out clearly. What I'm not fine with is them spanking me with a huge penalty for something that I may or may not have any control over.
Given that I run servers out of my home, I was considering switching to U-Verse because of the higher upstream bandwidth I could get, but I have since scrapped that plan. I can't afford to find out that one of my websites got slashdotted or was being attacked by a DDOS, and suddenly I have a $500 bandwidth bill. Nor is the alternative—having to disconnect my server for the rest of the month—an acceptable alternative, either. Remember that you can hit the AT&T bandwidth limit in just a couple of days of continuous use at full throttle.
So if I went with U-Verse, I would have no choice but to have a second network connection to switch my server over to if my U-Verse connection was at risk of hitting me with overage charges. The only way to do that is to keep my DSL service, since there is no Cable internet available here. That's a problem; I can't keep DSL if I add U-Verse service. Thus, I can't be a U-Verse customer.
For anybody not on U-Verse, you should do what I did when the first story broke: call, write, or email AT&T. Tell them that you were going to switch to their service, but then you found out about the bandwidth overage charges and you changed your mind. If enough people did this, AT&T would pull their heads out of their asses, but it would take at least a few tens of thousands of people complaining. While you're at it, tell your friends to do the same. This is even better if you can honestly tell them that you have a business-class connection as I do. AT&T needs businesses who run servers to help balance out their bandwidth with other backbone providers. Without servers, they're screwed.
BTW, to the AT&T local wire service tech who put the indoor DSL splitter in my outdoor box and caused me to lose both phone service and DSL for most of the day today due to rain (this makes two spring rainy seasons in a row with loss of local phone service), thanks. I just rewired my house back to the way I had it before your meddling, with the splitter safely indoors. It's no wonder AT&T is begging for every penny they have when they can't even get something as simple as basic phone wiring right.
And I swear if I get a cold from having to redo your incompetent rewiring job outside in the rain today....
So in the places where HTTPS is most needed to protect people's lives, Microsoft kowtows to pressure from a bunch of soon-to-be-ex Pol Pot dictators to trick people into using unencrypted traffic so that they can be snooped upon?
To everyone in the Middle East, when the revolution is through, remember who your friends were, and remember which large company tried to sell you out, then choose your purchases accordingly. Remember, developing nations have more influence on corporations through their buying power than any nation that is already locked into a particular vendor's products. Just a helpful tip.
To Microsoft, you should be ashamed. No, wait, the other thing. Tried and executed for crimes against humanity. Not to mention treason if Libya is being handled similarly. For shame.
Not without a lot of collateral damage, you can't. There's nothing stopping a single hosting server from serving Whitehouse.gov and Whitehouse.com on the same IP.
And anybody who couldn't figure that out from the comment deserves to have to watch that video.... Just saying.
To at least some extent, for app store apps, Apple does, albeit with user testing and static analysis on the binaries rather than examining the source code itself. Not before issuing the cert, mind you, but before allowing it on the store. And until it goes on the store, an app can only be run by devices that the developer owns. So in effect, yes, one CA does do this.
But how many years would you make the cert for? Eventually, all certs expire. That's a real problem with certs.
There's also the problem of how the user would get the certs. Assuming it's electronic, what's stopping somebody from guessing serial numbers and getting access keys for random people's phones? Not sure how practical that would be to exploit, but I'm sure somebody could come up with a way. :-)
As a side note, expiration of certs is a problem that really ought to be fixed. As far as I can tell, the only purposes for cert expiration seem to be A. people who are too lazy to revoke their certs, B. the insane overhead that CRLs represent, and C. so that the cert sales people can make more money, of which B. is irrelevant as long as all your clients support OCSP, and A. could be made irrelevant with a little bit of basic management infrastructure on the part of the CA, leaving C. (greed) as the only reason to continue this stupidity. The key to my door doesn't expire after a year. Why should the SSL key for my server?
And cert expiration is particularly problematic for application signing:
It starts to get more and more complex after that, and most of the problems stem from the inability to create a permanent SSL cert.
So yeah, it's a good idea in principle, but making it work in practice turns out to be more of a pain than it should be, mainly because certs have some serious design flaws. :-)
The thing is, it's not exactly possible to censor a TLD. About all you can do is mandate that ISPs block it on their DNS servers. You can't realistically prevent an entire country from running their own DNS servers, nor from using DNS servers outside the country, nor from using a VPN to tunnel DNS requests, etc. And if there's a market for it, somebody will sell access using just such a VPN.
And even more problematically, if the SMS is garbled, you'll leave the third person wondering "Which seat do I take?"
And those of us who are musicians need that air gap. Or, I suppose, more accurately, our neighbors need that air gap. Either way....
Put another way, it's cool when your neighbors go outside so they can hear you jam, just so long as they can go inside and avoid hearing you jam.