Familiar with the concept of a natural monopoly? File formats are onesuch; look it up. Converters aren't an adequate fix, because for a document of significant complexity they're inevitably lossy.
In any event, your post is based on a false premise. Most of the "pro-ODF" laws I've seen proposed are in fact endorsements of free and open formats, with a definition of "free and open" sufficiently exhaustive as to exclude OOXML. Once a law is in place requiring the use of free and open formats, deciding which specific free and open format (and software supporting it) to standardize on is a legitimate decision for an executive branch of government to make and apply -- the same way that a company can decide to standardize on a single piece of software across its many departments.
Here's my case: As you can see now, many OEMs have upgraded even their low-end computer specifications to meet Vista's demands. This means minimum 512mb ram, 1.x Ghz processor, etc. With their upgrade to Vista, their distributed-medicine computing calculations have also gotten a boost. Hence, the help to humanity!
What you're forgetting about is opportunity cost. Money being spent to buy higher-end hardware could instead be going to a myriad of other purposes -- or simply enabling people to buy more computers (or, turning that around, enabling more people to buy computers). Artificially increasing system requirements effectively creates a price floor, pricing the low-end consumer out of the market.
To go the reductio ad absurdum route, consider this claim: we should legally prevent anyone from buying anything less than a $20,000 32-processor parallel workstation, because humanity will benefit from the spare processing power.
Artificially raising the cost of computers (by law or by unnecessarily inflated system requirements) is harmful in the same way that raising taxes is harmful: Individuals are denied the opportunity to optimize for the most effective use of their funds.
Yes, burning coal emits a great deal of CO2 -- but technologies are available for carbon capture and storage rather than emitting it to the environment, and most other emissions are already routinely scrubbed in modern, first-world coal plants. Granted, carbon capture and storage is not widely in place -- but some of the first plants using it should be coming online inside the next four years.
As for transmission loss and such, last time I saw the numbers on that, they didn't outweigh the inefficient-little-engines problem.
Only if you're using corn to produce it. Sugarcane-based ethanol is vastly more efficient, and runs a clear net positive. I'm looking forward to algae-based biofuel production coming on-line; that set of technologies is even more promising, once the kinks get worked out.
I don't know the tone of your post. Is this something you see as a problem?
One big coal plant (with scrubbers and such) isn't necessarily any worse than hundreds of thousands of small, inefficient gasoline engines -- and infrastructure upgrades to reduce the pollution from that plant (and otherwise mitigate its effects) can be done at one time, in one place, rather than needing to upgrade hundreds of thousands of small, separately owned vehicles. (If the folks working on fusion power get that worked out, every EV is suddenly fusion-powered -- while folks with gasoline vehicles are still releasing the carbon from long-dead forests).
Coal is dirty, sure, but lots of little inefficient gasoline engines isn't necessarily any better. (Also, not everyone gets their power from coal).
A modern (scrubber-enabled) coal-fired power plant doesn't pollute as much as the thousands of individual vehicles with less-efficient combustion engines it could provide power to. Further, even if you were only moving the pollution into one place with no net decrease (which isn't the case), having it all in one place makes it easier to mitigate; retrofitting a single power plant with new technology to reduce its pollution output is easier than retrofitting every car which is charged off that plant.
Oh. I misread your post, so at the risk of repeating myself, let me clarify a little:
They don't charge more for books via Kindle than on paper. The $9.99 is for expensive books (just-released, hardcover-only, bestsellers, etc); cheaper books are... well, cheaper.
Not that $9.99 is the absolute price ceiling (I did a bit of spot-checking and technical books were significantly more expensive, though still below the price of hardcopy). Frankly, I'd be happier if I could get technical books for the Kindle via something like Safari, while using the existing model for other content.
Just-released, hardcover-only novels for $9.99 isn't as unreasonable as you suggest, given that mass market novels are generally much cheaper. As for the blogs... looked at how much EVDO costs lately? Something needs to foot that bill.
As for the suggestion that Amazon should be obligated to give their hardware away for free, I think that the CueCat is ample demonstration of where that business model leads.
Look, when the feel of the device is most important is when you're reading books on it, not the 1% of the interaction time that you spend loading books on it. The "free wireless" is restricted to book purchasing and a handful of specific features, so it's not really that useful.
When I was going to school, my roommate subscribed to the Wall Street Journal. I loved that paper -- but I don't get it at home today; I don't have enough time at home to get much reading in, the physical paper isn't really portable enough to keep with me through the day (much less portable than a paperback, though even those aren't ideal), and having that much ink and wood pulp need to go into my recycling bin every day just isn't something I can justify. For $10/month, on the other hand, I'd jump at having a small, portable device with an ever-current copy of the Journal at hand. And the purchasing experience might not matter much in the common case, but when I'm visiting the in-laws out in the boonies with only long-distance dial-up Internet and no bookstores for miles (but tolerable cell phone coverage), the EVDO has the potential to make a very, very big difference.
Also, it hasn't been established that the feel of the device when reading books is something that Amazon got wrong; the folks asserting as much are people who haven't actually held the device.
The "two year contract" is for using the iPhone as a phone. A book reader can't place phone calls, so it doesn't exactly compete with the iPhone in that regard.
It may not place calls, but it does download content via a phone provider's cellular network. With the iPhone, doing that requires a contract.
What I'd like to know is if the Kindle does the same "white to black to white flash" refresh as the Sony Reader. I have checked the Reader out a couple times, but that screen refresh on every page turn is absolutely a functionality killer. Who'd read a book that flashed on every page turn?
The demo videos for the Kindle show its page turn process. I want to say it flashes to white and then to the new text, but don't remember that with certainty -- either way, it's much less annoying than the Sony reader.
Remember that Tablet PCs sound like a good idea and deliver "features," but nobody buys them because they don't really do enough of anything you can't do with a regular laptop, and can't match a laptop's features, but they still cost a lot.
Not true. I'm in medical software. Lots of people buy Tablet PCs in this field. Also, most of the newer models can be used as laptops -- they have keyboards that fold into the back, or that are included in a case which acts as a cover for the tablet when it's not in use, or somesuch like that.
I haven't looked at the Kindle in depth yet, but if it's a great product that would make reading more accessible, I'm all for it. I think $400 might be too high of an entry price, as it doesn't offer to replace a similarly priced phone (as does the iPhone), but rather the free nature of begin able to read books. Kindle isn't "$400 vs the $400 iPhone plus subscription fees," it's "$400 investment vs $0 reading physical books."
To some extent, yes. On the other hand, it has a niche: I can take a Kindle to Thanksgiving at my grandmother-in-law's ranch in southeastern Oklahoma and get new reading material even though the only 'net access available is long-distance dialup and the nearest place I can buy books is a Wal-Mart 40 miles away (and they're not likely to have anything I'd find interesting anyhow). Sure, I could do the same thing with my laptop if I owned an EVDO card, but EVDO plans are bloody expensive; I'm much more comfortable paying a one-time cost for a device and having the access built in. Also, there's arguably something to be said for reading the WSJ on paper (electronic or otherwise); I spend all day looking at an LCD, and would rather not keep doing so in my leisure time.
Or I could take a Kindle to Thanksgiving in southeastern Oaklahoma, if they had started shipping a few days earlier. That they didn't has taken me out of the early-adopter pool for now.
A reasonable position, but it outlines another flaw with the Kindle. Amazon isn't putting them into stores, so we can't test them out. We have to spend $400 basically blind in order to see whether we like the feel of the device.
That's what early adopters are for. Just like that guy across the hall who bought an iPod on the first day they were available, there's going to be someone you know who buys a Kindle. If not, perhaps there's a media source whose reviews you trust?
And if not, keep in mind that you're risking not $400, but rather the cost of return shipping.
And needing to hook up a USB cable to add content is more like a real book? I don't think the wireless connectivity (free EVDO! What's not to like?) is something to complain about.
Indeed, the proof is in the pudding -- or, in this case, the reading. I'm not inclined to give much credence to the complaints of folks who haven't actually held one of these in their hands.
You do realize that slashdot is a pretty lousy place to be challenging the crowd to show their kernel hacking creds, right?
(Me, I've fixed PS2 keyboard support on some obscure MIPS subarchitecture, and ported the MPPE driver to Linux 2.4 [think I was actually the first person to do that, though it's someone else's port that made it upstream], and did a little tooling around the input core, and fixed a DSDT bug that was causing the PCI bus on some Hitachi prototype hardware to be initialized wrong... but then, I'm mostly a userspace type).
Stop free support, you mean. If you want a bug in a 10-year-old kernel fixed, you can pay me to do a backport. Sure, I'll charge completely insane rates (my primary job keeps me busy as it is), but there are enough C developers with kernelspace experience that if you have a reason to use a 10-year-old kernel (and in embedded space, that's not a completely unreasonable thing to do), you can find someone who'll maintain it for you.
Open source may not guarantee you profit in your core business, but it does guarantee that you're not held hostage to a single developer who owns exclusive rights to the infrastructure you built on.
Meaning something that works on your desktop PC will also work on the Solaris server you intend to deploy it to.
Only until your code depends on filesystem locking semantics which differ between your development and target platforms.
Yes, I've had to clean up after someone who relied on Microsoft's locking semantics for their Java application (or rather, their component of a much, much larger application). Knowing the abstractions and not the underlying details does not a useful developer make.
Thank you for your reply. I can't respond point-by-point, but let me say that there's a great deal of legitimacy in your position. I still choose to think of the movement by its better-natured members -- but perhaps that's the eternal optimist in me. And speaking of my personal nature...
I have no doubt that you are among thousands of people who would have no moral issue with personally waterboarding Dick Cheney.
I'm not so sure of that. Causing anyone physical pain makes me uncomfortable in the extreme -- and bringing such things into the formerly abstract political realm raises the same concern which the US military has with regard to allowing torture in interrogations: What goes around, comes around.
One suggestion, though: When you argue (what can be loosely interpreted as) "you progressives have turned into an evil bunch", it groups the "evil bunch" in with the "progressives", increasing the cross-identification of the two groups, and leading all who consider themselves "progressives" to reflexively defend the group as a whole. Would it not be better to take an approach which encourages progressives' fellows to excise those who are mistaken? "You have no right to call yourself a progressive" has the effect not only of providing a public reprimand, additionally as emphasizing to other people who consider themselves progressives that they can maintain that self-identification while ostracizing those who spread hate.
Actually, let me follow up a little more, because I think it will allow a bit of clarification.
If you disapprove of my actions, then I suggest you take your own advice and snub me.
I don't snub people simply on account of disagreement -- that way lies groupthink. I snub people who act so egregiously as to make it clear that any attempt to reason with them would be futile. Rape and razor wire fall into that category -- but so long as you're willing to respond to thoughtful posts in such a manner as to make it clear that you consider the reasoning and positions behind those who believe and act differently than you do (even should you ultimately find their arguments unpersuasive), social interaction has the potential to be worthwhile.
I don't disapprove of your actions, inasmuch as they address the guilty parties; I simply think you're painting with a fairly wide brush. A good chunk of your argument is "where's the condemnation?"; my intent, then, was to answer that. Certainly, forcing individuals to recognize their own hate speech for what it is is necessary for those individuals to reform themselves -- but if the effort to do so is itself divisive, it's unlikely to have the intended effect.
I was about to go as far as to argue that the core of the movement (as with most social movements) is made up of reasonable people working not out of anger but towards a reasonable and considered (if slanted by differences of perspective) view of "the greater good" -- but after some consideration, I'm willing to back away from that ever so slightly. Consider one difference between Clinton and Obama. Both of these candidates are in favor of abortion. Obama, however, is careful in his stance to acknowledge that those who are pro-life have a strong and legitimate moral argument, and that his position is a nuanced one taking moral considerations into account; after a staffer put the typical Democratic boilerplate on his web site, he ended up making a personal phone call to apologize to a pro-life member of the public who objected (and removed the boilerplate). Clinton's position, on the other hand, is effectively just that boilerplate -- Side A is Good, Side B is Wrong And Evil. One is thoughtful; the other is divisive mantra.
Now, your argument that "progressives" have become largely a hateful bunch doesn't hold much water with me -- I live in Austin, TX; being Austin, we have a lot of (sometimes fairly extreme) progressives, and being Texas, we have a lot of (sometimes fairly extreme) conservatives, and I don't see a lot of hate. What I do see is individuals unwilling to acknowledge that their opponents might just maybe have some kind of a point and that what those opponents say should at least be given enough consideration to come up with a thoughtful refutation that accounts for the reasoning behind the position they hold in the first place. The hate? That's the immature kiddies ranting on the Internet, not the liberal-leaning guy down the street. The closed-mindedness? That's a real and serious problem -- but not one limited to the progressive movement, but rather a blight on all modern American politics today.
I ignore assholes on the Internet. Someone goes off about rape and razor wire and such? They go onto my mental blacklist. To condemn them would be to acknowledge them -- and that's more than they're worth. Social ostracism works better than open condemnation sometimes, because the latter allows those who flaunt social conditioning to decide the topic of discussion; further, this policy means that these most offensive of trolls are denied the response they crave.
So -- you have "progressives" tagged with any number of negative correlations. That doesn't tell me, in your mind, what defensible or positive positions you also have associated -- so I can't debate them. It's obvious that rape and murder are unacceptable -- so obvious that there's no point in jumping into the conversation to say as much; individuals who would argue to the contrary are inherently unreasonable, and there's no point in talking to unreasonable people. If you want to talk about socialized healthcare, or intellectual property law, or states rights, or taxation... well, those are all good topics, and I'm happy to discuss them. But to discuss "progressivism" with someone whose view of what that represents is so tainted with baggage unassociated with the philosophy and agenda would be simply a waste of time.
I enjoy talking religion with an individual who knows the subject and the arguments -- who is willing to discuss the anthropic principal and offer considered rebuttals. I enjoy talking politics or intellectual property law or computers with someone who has something interesting to say and is willing to challenge my beliefs in the saying of it. Those who mark themselves "nutter", on the other hand, are not worth my time and attention. Do I thus overlook their behavior? I consider my behavior to be snubbing it -- and them.
If the risk were significant enough for me to care about the chances of being on the short end of the stick, the insurance companies would disallow it. They don't care, thus I don't either: folks tend to work pretty hard at getting their risk analysis right when there's money on the line, and statistics is a well-developed field. (That said, as I've mentioned elsewhere in this thread, I'm ready to acknowledge that a filtering repeater that relays text messages and pages inbound-only is probably a pretty good idea -- would cover several "emergency" scenarios discussed, but require recipients of messages to step outside before being able to respond).
WRT loss of power and such -- wtf do you think UPSs are for? Telephony equipment typically has large batteries attached for just that sort of reason. Also, emergency bands for radio are well-established; adding support for those frequencies to a repeater already capable of several cell phone bands can't be very difficult at all, if the cost/benefit analysis (taking into account the probability of a scenario requiring them) supports it.
Taking most of this discussion into consideration, I'm leaning towards going the copper-mesh approach, and then relaying numeric pages inside (for on-call doctors and the like... though incoming-only text messages for conventional cell phones might make sense as well).
That approach would be legal and allow high-priority alerts to be passed in, but no two-way communications of any sort without stepping outside.
No, it's not illegal to block mobile phones; it's illegal to jam mobile phones. The relevant law is the same thing that applies to every other device which emits RF, not anything passed by Congress with mobile phones specifically in mind.
That said, something like a cell phone extender which forwards numeric pages but not text messages or regular phone calls may be appropriate; that would handle both the doctor-on-call case and the urgent-family-emergency one (by relaying only a numeric message, and thus requiring the target of that message to step outside before calling back).
Slashdot Mirror
Familiar with the concept of a natural monopoly? File formats are onesuch; look it up. Converters aren't an adequate fix, because for a document of significant complexity they're inevitably lossy.
In any event, your post is based on a false premise. Most of the "pro-ODF" laws I've seen proposed are in fact endorsements of free and open formats, with a definition of "free and open" sufficiently exhaustive as to exclude OOXML. Once a law is in place requiring the use of free and open formats, deciding which specific free and open format (and software supporting it) to standardize on is a legitimate decision for an executive branch of government to make and apply -- the same way that a company can decide to standardize on a single piece of software across its many departments.
Patriotism and nationalism are not the same thing. You might consider learning the difference.
To go the reductio ad absurdum route, consider this claim: we should legally prevent anyone from buying anything less than a $20,000 32-processor parallel workstation, because humanity will benefit from the spare processing power.
Artificially raising the cost of computers (by law or by unnecessarily inflated system requirements) is harmful in the same way that raising taxes is harmful: Individuals are denied the opportunity to optimize for the most effective use of their funds.
Yes, burning coal emits a great deal of CO2 -- but technologies are available for carbon capture and storage rather than emitting it to the environment, and most other emissions are already routinely scrubbed in modern, first-world coal plants. Granted, carbon capture and storage is not widely in place -- but some of the first plants using it should be coming online inside the next four years.
As for transmission loss and such, last time I saw the numbers on that, they didn't outweigh the inefficient-little-engines problem.
Only if you're using corn to produce it. Sugarcane-based ethanol is vastly more efficient, and runs a clear net positive. I'm looking forward to algae-based biofuel production coming on-line; that set of technologies is even more promising, once the kinks get worked out.
I don't know the tone of your post. Is this something you see as a problem?
One big coal plant (with scrubbers and such) isn't necessarily any worse than hundreds of thousands of small, inefficient gasoline engines -- and infrastructure upgrades to reduce the pollution from that plant (and otherwise mitigate its effects) can be done at one time, in one place, rather than needing to upgrade hundreds of thousands of small, separately owned vehicles. (If the folks working on fusion power get that worked out, every EV is suddenly fusion-powered -- while folks with gasoline vehicles are still releasing the carbon from long-dead forests).
Coal is dirty, sure, but lots of little inefficient gasoline engines isn't necessarily any better. (Also, not everyone gets their power from coal).
A modern (scrubber-enabled) coal-fired power plant doesn't pollute as much as the thousands of individual vehicles with less-efficient combustion engines it could provide power to. Further, even if you were only moving the pollution into one place with no net decrease (which isn't the case), having it all in one place makes it easier to mitigate; retrofitting a single power plant with new technology to reduce its pollution output is easier than retrofitting every car which is charged off that plant.
Problem? Sure. Huge problem? Not so much.
Oh. I misread your post, so at the risk of repeating myself, let me clarify a little:
They don't charge more for books via Kindle than on paper. The $9.99 is for expensive books (just-released, hardcover-only, bestsellers, etc); cheaper books are... well, cheaper.
Not that $9.99 is the absolute price ceiling (I did a bit of spot-checking and technical books were significantly more expensive, though still below the price of hardcopy). Frankly, I'd be happier if I could get technical books for the Kindle via something like Safari, while using the existing model for other content.
Just-released, hardcover-only novels for $9.99 isn't as unreasonable as you suggest, given that mass market novels are generally much cheaper. As for the blogs... looked at how much EVDO costs lately? Something needs to foot that bill.
As for the suggestion that Amazon should be obligated to give their hardware away for free, I think that the CueCat is ample demonstration of where that business model leads.
Also, it hasn't been established that the feel of the device when reading books is something that Amazon got wrong; the folks asserting as much are people who haven't actually held the device.
Or I could take a Kindle to Thanksgiving in southeastern Oaklahoma, if they had started shipping a few days earlier. That they didn't has taken me out of the early-adopter pool for now.
And if not, keep in mind that you're risking not $400, but rather the cost of return shipping.
And needing to hook up a USB cable to add content is more like a real book? I don't think the wireless connectivity (free EVDO! What's not to like?) is something to complain about.
Indeed, the proof is in the pudding -- or, in this case, the reading. I'm not inclined to give much credence to the complaints of folks who haven't actually held one of these in their hands.
One of them costs $400 plus a 2-year contract. That's a pretty big difference.
You do realize that slashdot is a pretty lousy place to be challenging the crowd to show their kernel hacking creds, right?
(Me, I've fixed PS2 keyboard support on some obscure MIPS subarchitecture, and ported the MPPE driver to Linux 2.4 [think I was actually the first person to do that, though it's someone else's port that made it upstream], and did a little tooling around the input core, and fixed a DSDT bug that was causing the PCI bus on some Hitachi prototype hardware to be initialized wrong... but then, I'm mostly a userspace type).
Stop free support, you mean. If you want a bug in a 10-year-old kernel fixed, you can pay me to do a backport. Sure, I'll charge completely insane rates (my primary job keeps me busy as it is), but there are enough C developers with kernelspace experience that if you have a reason to use a 10-year-old kernel (and in embedded space, that's not a completely unreasonable thing to do), you can find someone who'll maintain it for you.
Open source may not guarantee you profit in your core business, but it does guarantee that you're not held hostage to a single developer who owns exclusive rights to the infrastructure you built on.
Yes, I've had to clean up after someone who relied on Microsoft's locking semantics for their Java application (or rather, their component of a much, much larger application). Knowing the abstractions and not the underlying details does not a useful developer make.
And speaking of my personal nature...I'm not so sure of that. Causing anyone physical pain makes me uncomfortable in the extreme -- and bringing such things into the formerly abstract political realm raises the same concern which the US military has with regard to allowing torture in interrogations: What goes around, comes around.
One suggestion, though: When you argue (what can be loosely interpreted as) "you progressives have turned into an evil bunch", it groups the "evil bunch" in with the "progressives", increasing the cross-identification of the two groups, and leading all who consider themselves "progressives" to reflexively defend the group as a whole. Would it not be better to take an approach which encourages progressives' fellows to excise those who are mistaken? "You have no right to call yourself a progressive" has the effect not only of providing a public reprimand, additionally as emphasizing to other people who consider themselves progressives that they can maintain that self-identification while ostracizing those who spread hate.
I don't disapprove of your actions, inasmuch as they address the guilty parties; I simply think you're painting with a fairly wide brush. A good chunk of your argument is "where's the condemnation?"; my intent, then, was to answer that. Certainly, forcing individuals to recognize their own hate speech for what it is is necessary for those individuals to reform themselves -- but if the effort to do so is itself divisive, it's unlikely to have the intended effect.
I was about to go as far as to argue that the core of the movement (as with most social movements) is made up of reasonable people working not out of anger but towards a reasonable and considered (if slanted by differences of perspective) view of "the greater good" -- but after some consideration, I'm willing to back away from that ever so slightly. Consider one difference between Clinton and Obama. Both of these candidates are in favor of abortion. Obama, however, is careful in his stance to acknowledge that those who are pro-life have a strong and legitimate moral argument, and that his position is a nuanced one taking moral considerations into account; after a staffer put the typical Democratic boilerplate on his web site, he ended up making a personal phone call to apologize to a pro-life member of the public who objected (and removed the boilerplate). Clinton's position, on the other hand, is effectively just that boilerplate -- Side A is Good, Side B is Wrong And Evil. One is thoughtful; the other is divisive mantra.
Now, your argument that "progressives" have become largely a hateful bunch doesn't hold much water with me -- I live in Austin, TX; being Austin, we have a lot of (sometimes fairly extreme) progressives, and being Texas, we have a lot of (sometimes fairly extreme) conservatives, and I don't see a lot of hate. What I do see is individuals unwilling to acknowledge that their opponents might just maybe have some kind of a point and that what those opponents say should at least be given enough consideration to come up with a thoughtful refutation that accounts for the reasoning behind the position they hold in the first place. The hate? That's the immature kiddies ranting on the Internet, not the liberal-leaning guy down the street. The closed-mindedness? That's a real and serious problem -- but not one limited to the progressive movement, but rather a blight on all modern American politics today.
Here's another option for you:
I ignore assholes on the Internet. Someone goes off about rape and razor wire and such? They go onto my mental blacklist. To condemn them would be to acknowledge them -- and that's more than they're worth. Social ostracism works better than open condemnation sometimes, because the latter allows those who flaunt social conditioning to decide the topic of discussion; further, this policy means that these most offensive of trolls are denied the response they crave.
So -- you have "progressives" tagged with any number of negative correlations. That doesn't tell me, in your mind, what defensible or positive positions you also have associated -- so I can't debate them. It's obvious that rape and murder are unacceptable -- so obvious that there's no point in jumping into the conversation to say as much; individuals who would argue to the contrary are inherently unreasonable, and there's no point in talking to unreasonable people. If you want to talk about socialized healthcare, or intellectual property law, or states rights, or taxation... well, those are all good topics, and I'm happy to discuss them. But to discuss "progressivism" with someone whose view of what that represents is so tainted with baggage unassociated with the philosophy and agenda would be simply a waste of time.
I enjoy talking religion with an individual who knows the subject and the arguments -- who is willing to discuss the anthropic principal and offer considered rebuttals. I enjoy talking politics or intellectual property law or computers with someone who has something interesting to say and is willing to challenge my beliefs in the saying of it. Those who mark themselves "nutter", on the other hand, are not worth my time and attention. Do I thus overlook their behavior? I consider my behavior to be snubbing it -- and them.
If the risk were significant enough for me to care about the chances of being on the short end of the stick, the insurance companies would disallow it. They don't care, thus I don't either: folks tend to work pretty hard at getting their risk analysis right when there's money on the line, and statistics is a well-developed field. (That said, as I've mentioned elsewhere in this thread, I'm ready to acknowledge that a filtering repeater that relays text messages and pages inbound-only is probably a pretty good idea -- would cover several "emergency" scenarios discussed, but require recipients of messages to step outside before being able to respond).
WRT loss of power and such -- wtf do you think UPSs are for? Telephony equipment typically has large batteries attached for just that sort of reason. Also, emergency bands for radio are well-established; adding support for those frequencies to a repeater already capable of several cell phone bands can't be very difficult at all, if the cost/benefit analysis (taking into account the probability of a scenario requiring them) supports it.
Taking most of this discussion into consideration, I'm leaning towards going the copper-mesh approach, and then relaying numeric pages inside (for on-call doctors and the like... though incoming-only text messages for conventional cell phones might make sense as well).
That approach would be legal and allow high-priority alerts to be passed in, but no two-way communications of any sort without stepping outside.
No, it's not illegal to block mobile phones; it's illegal to jam mobile phones. The relevant law is the same thing that applies to every other device which emits RF, not anything passed by Congress with mobile phones specifically in mind.
That said, something like a cell phone extender which forwards numeric pages but not text messages or regular phone calls may be appropriate; that would handle both the doctor-on-call case and the urgent-family-emergency one (by relaying only a numeric message, and thus requiring the target of that message to step outside before calling back).