Small niche ISPs like these don't get a fair shake in the current market because spectrum is tightly controlled by pandering bureaucrats and local franchise rights are impossible to obtain unless you sell your soul to Satan (buildout rules, 5% gross revenue taxes).
And shouldn't Comcast and Verizon offer the same services as Christian ISPs as an option for subscribers? They haven't, probably because most people actually don't really want centralized filtering--even parents themselves.
Unlike roads and electricity grids, the Internet isn't a public utility. Right now about 4 in 5 Americans can select from three residential-grade broadband services: cable, DSL and 3G, all of which offer enough throughput and usage for plenty of everyday Web browsing, emailing, and streaming YouTube vids. And once LTE and Wi-Max are built out, 3mbps symmetric for $30 a month (which you can now get in much of Baltimore) will be the norm in urban and suburban areas.
Besides, public utilities are not that great. They don't really advance much over time, mainly because they're insulated from competition and have a guaranteed profit margin. Water, roads, electricity utilities have all provided us pretty much the same service at the same price for decades. Don't we want something more from our ISPs? Say, companies taking big gambles on next-gen services that might fail or might become the next big thing?
I certainly won't argue that public utility-style regulation of the Internet might cut out some of the occasional asshat tactics of some big ISPs, but is turning Internet access into a stagnant-but-consistent service really worth it?
Except, of course, when you and your people are appealing to irrational fears for personal political gain and not actually representing the very consumers you're supposed to look out for. Which clearly was the case with the censored free wireless plan.
If consumers desire a filtered wireless network, then shouldn't it emerge even without the federal government forcing it upon us?
Even without rigid spectrum rules, there is nothing stopping a company from buying up spectrum rights and using it for family-friendly wireless broadband. But what if users with alternative preferences--say, parents who are capable of protecting their kids online without centralized censorship and nipple slip fines--outnumber those who want government nanny-state rules? The FCC's proposed conditions that would have mandated filtering mean that companies couldn't compete to deliver what people really want.
Any way you look at this, the plan was a perfect example of the FCC pretending it knows best what the People of the United States deserve.
It's good to hear this plan is dead. Kevin Martin backed this network so he'd look like a "family-values man" and score some points with cultural conservatives in North Carolina, where Martin has long been planning a bid for Congress.
This 25mhz of spectrum in the AWS3 band could go toward a lot of very cool services--LTE, for instance. Martin's plan--to earmark the 25mhz for 768kbps of censored wireless broadband that wouldn't even be widely deployed for a decade--is clearly not the smartest way to put these frequencies to use.
The FCC should do one of two things with this spectrum--a)auction it off with no strings attached and allow the winning firm to sell or rent the spectrum as if it were property, or b)set the band free as unlicensed flexible use spectrum subject only to basic EIRP and non-interference requirements and nothing more.
In theory, if the banking system were known to be compromised in such a huge way, and there were no way of knowing if your own bank account was compromised or not, shouldn't there be a massive bank run? Because everyone wants to withdraw their money right away to minimize the chance that this ridiculous security leak negatively affects them, right? Such a massive erosion of confidence can completely destroy a banking system.
a. Ubiquity doesn't require massive taxpayer spending. The Internet, Windows, Google, ISO standards, all emerged without much government involvement.
b. Sure, computers can play a useful role in schools, but anybody who attended public schools knows that they mostly just sit around and aren't often used for educational purposes. Money on education can be better spent on incentivizing good teachers, for one thing.
c. My stock retort to tax-and-spend leftists: So we wasted 2.5T on an unnecessary foreign war in Iraq, that somehow justifies wasting even more taxpayer money on infrastructure boondoggles that largely amount to ditch-digging for ditch-digging's sake?
d. Wow. This isn't 1939. The global economy is intrinsically intertwined, and the prospect of an all-out war that would seriously impair U.S. ability to acquire manufactured goods is incredibly remote. Besides, if other countries are busy building our goods in factories, that leaves American workers free to, you know, do the thinking that's required to make these products. There's a reason we have Intel, Microsoft, Google, Apple, etc all based right here in the U.S. And it isn't because of silly trade barriers in the name of "national security"
Obama's on the wrong track on infrastructure spending. Yes, his plans will create jobs, but at the expense of other, more economically efficient jobs. Like FDR, Obama falsely assumes that government can create wealth, when in fact government can only redistribute wealth from productive uses to unproductive ones. Besides, the real gems of U.S. infrastructure--our telecommunications system, our electricity grid--are largely in private hands, anyway.
The competitive ISP market you desire will arise on its own in the next five years because of smarter, more effective broadband technologies. At the heart of this is spectrum, which erases the rules of the game as they existed up until now and replaces them with conditions far more conducive to entry (and exit).
Laying a wire to each home is really, really expensive on a per-household basis. Verizon has spent upwards of $1,000 per home to add FiOS. Building completely from scratch is even more costly.
But wireless is a whole different story. Sure, you've got to have base stations, which require backhaul, but the costs involved there are much more bearable than those involved with old-fashioned overbuilding.
To see why wireless broadband is such an important vehicle for broadband competition, consider Wi-Max. It's already giving DSL a run for its money in Baltimore, because it offers reliable throughput of 3Mbps down/2Mbps up for $25 a month, with no strict usage limits. Of course, Wi-Max isn't the end-all of wireless, but merely a stopgap until Long-Term Evolution really takes off.
LTE will completely change the game when it starts getting widely deployed in a couple years. After the 700mhz band is vacated in a couple months, T-Mobile, Sprint, Verizon, and AT&T will all work on building out LTE networks, which will offer at least 10Mbps for the same price as a typical home broadband connection.
Six to eight ISP choices may not mean as many choices as we have for burgers, but it's still a lot of competition, especially compared to what we have now. And it is surely enough to ensure that ISPs can't get away with the heavy-handed tactics that have angered so many of us recently.
It wouldn't surprise me if by the time Congress finally gets around to regulating ISPs, the whole justification for regulating them in the first place is no longer logically valid.
I haven't noticed any hangs, but SMART reports 41165316 hardware read errors. ECC is recovering all of them, but it's still troubling, as I haven't seen nearly that many errors on any other healthy disk I've ever owned. But the drive is performing fine, so perhaps I'm worried about nothing.
Reliability can be achieved if you just don't get the biggest drives out there. I've purchased 14 hard disks from three different manufacturers over the past 2 years, ranging from 250 to 640GB, and I have yet to have a single failure with 24/7 use.
Right now, the biggest drives available that offer real reliability are in the 500-640GB range. I'd suggest the WD6400AAKS, which is truly an amazing drive (and under 12 cents a Gigabyte, too)
You make a very valid point. As sites like Newegg start to get hundreds--even thousands--of reviews for products like hard disks, a prospective buyer can look at the percentage of negative reviews as a rough guide to the probability of failure.
Compared to the 640GB WD6400AAKS, pretty much all 1TB-plus drives on Newegg have a lot of 1-star reviews. That does suggest these huge drives aren't up to par in terms of reliability.
No, having a high market share does not make you a monopoly unless you have the power of a monopoly. Since dynamic markets (like Web advertising) are highly contestable, the ever-present threat of entry by new firms forces even the most dominant players to behave as if they weren't monopolies.
If Google were really a monopoly, it wouldn't have to work so damn hard. The folks in Mountain View could just cut prices temporarily whenever somebody got too close, and otherwise sit on their chairs and twiddle their thumbs.
But that's not what Google is doing. Google continues to develop new products and improve existing ones pretty rapidly, because it knows that it operates in a cutthroat marketplace with intense competition.
How is Google not playing by the capitalist rules? Just ten years ago, Google was a grad student research project, and now it is the global leader in Web advertising. It's crucial to realize how Google came about, because the firm that ultimately dethrones Google will emerge unexpectedly from humble roots.
Look, nobody is forcing Internet users to rely on Google for search. The reason for Google's continued dominance isn't because it is an evil monopoly, but because Google managed to build a platform that a lot of people like. The success of Google has spurred competitors like Microsoft to invest serious cash to develop a superior set of online services, Isn't that exactly what capitalism is all about?
Cloud computing may be the next big thing, but it'll be quite some time before cloud computing can give computing on the ground a real run for its money.
Last-mile broadband, for most smaller businesses and homes, is synonymous with highly asymmetric connections. Since many applications don't function on networks with strained upstream bandwidth, only enterprise-grade data service sare able to offer the robust speeds needed for cloud computing to excel.
And web-based applications still can't approach locally hosted apps in terms of reliability and performance. This is why so many Gmail users connect via IMAP using their Outlook client instead of relying on Gmail's sometimes glitchy web interface.
One's body parts, if willingly defaced, are surely not immune from trademark infringement suits. Imagine a lowlife tatooing some valuable intellectual property on his arm and engaging in negative activities that might harm the value of the trademark. That'd surely be grounds for a hefty lawsuit.
For not much more money RAID 10 means significantly greater performance and a good deal more redundancy to boot.
With 1.5TB for $189 these days, it doesn't make a whole lot of sense to do RAID 5 unless you're penny pinching and performance is of no concern.
You'd be surprised how quickly 50GB Blu-Ray images add up. I have 6 TB in a RAID 10 currently, and within 18 months I am confident I'll be at twice that, at least. With 1.5TB disks for $189 these days, who doesn't gradually amass terabytes of data?
OK, let's say you can cut the rate of death by a factor of 10 by driving safely.
There are, on average, 40000 automobile deaths per year in the U.S., or roughly 1 per 7500 people. So a safe driver's risk of dying in a car accident would be 1 in 75000.
ISPs have always demonstrated preference for content. Inbound port 80 traffic and excessive SMTP traffic are just two examples of commonly restricted types of traffic.
Besides, even websites like YouTube actively remove videos with explicit scenes. Does that mean YouTube is liable every time somebody posts an infringing video, even if YouTube isn't actually aware of it?
Actually, there almost always is an alternative. 78 % of homes with a phone line can get DSL. 9 out of 10 people live in a home that is in a 3G coverage area. Anybody in the lower 48 with a clear view of the southern sky can get satellite.
Are these options often slower and pricier than cable? Sure. But if you hate cable so much, why aren't you willing to put up with slower speeds or higher latency in exchange for an ISP that doesn't forge packets?
Life is full of goods and services for which there's no perfect substitute, but that doesn't mean that there aren't substitutes at all. Being able to browse the web at 768kbps isn't lightning fast, but it's surely broadband and a hell of a lot better than dial-up.
Here's the thing: Under the Administrative Procedure Act, federal agencies have to follow specific rules when drafting legally binding principles. The Internet Policy Statement from 2005 was a non-binding statement and it didn't go through the notice and comment procedure that is required for a rule to become enforceable.
Safe Harbor, not common carrier, is what protects Comcast as per the DMCA and the CDA.
Common carrier is a completely different concept that affects telcos, not cable companies.
Modifying TCP streams--however repugnant--does not automatically mean the ISP is liable for the content that traverses its network. That's the law, like it or not.
Where have you been for the last decade? Broadband has always cost around $40 to $50 a month, but speeds keep moving upward on average as cost per byte decreases. Back when I first got broadband in 1998, it was 768k per second. Now I get 16/2 for $52.95 per month.
Monopolists may not behave like perfect competitors, but they DO pass some savings on to consumers, albeit less than an optimal amount.
In the case of employee rights, what's at stake aren't property rights, but the right to engage in voluntary private transactions.
Smoking isn't just something that occurs in bars for no reason. It confers utility on people--believe it or not, there are rational adults out there who knowingly choose to smoke cigarettes even though long-term smoking is known to reduce lifespan by a few years. Are people who engage in this activity crazy, or stupid, or irrational? Hardly. Some people really like smoking, enough so that they are willing to shed years of life. Just as many people choose to eat bacon cheeseburgers for lunch every day and drink a six pack of Bud every night, some people prefer a shorter life that involves smoking to a longer one that doesn't.
Furthermore, the magnitude of exposure to secondhand smoke is hardly comparable to, say, the risk of losing an eye in an industrial accident that can be mitigated with goggles. I gladly concede that prolonged employment in a smoke-filled bar reduces life expectancy and increases the risk of lung cancer and emphysema, but working in a bar 20 hours a week for a couple years while an undergrad is not likely to have a measurable impact on life expectancy. It may cause some short-term health problems, of course, but such ailments can usually be alleviated by simply switching to one of the many places of employment where ambient secondhand smoke is non-existent.
People want to be able to smoke in bars. Bar owners want to be able to offer patrons a place where they can smoke while drinking. People willingly enter in to employment in establishments where smoking takes place, despite the fact that it's well established that prolonged exposure to secondhand smoke is detrimental to human health. These are all voluntary decisions made by adults. If we believe people should be responsible for making decisions for themselves about what risks to take in life, then what justification is there for government to intervene?
Relying on government to protect people from themselves is is wrong when it comes to the War on Drugs, and it is wrong when it comes to bans on smoking in pubs to protect employees who voluntarily decide to work in a bar.
Slashdot Mirror
Small niche ISPs like these don't get a fair shake in the current market because spectrum is tightly controlled by pandering bureaucrats and local franchise rights are impossible to obtain unless you sell your soul to Satan (buildout rules, 5% gross revenue taxes).
And shouldn't Comcast and Verizon offer the same services as Christian ISPs as an option for subscribers? They haven't, probably because most people actually don't really want centralized filtering--even parents themselves.
Unlike roads and electricity grids, the Internet isn't a public utility. Right now about 4 in 5 Americans can select from three residential-grade broadband services: cable, DSL and 3G, all of which offer enough throughput and usage for plenty of everyday Web browsing, emailing, and streaming YouTube vids. And once LTE and Wi-Max are built out, 3mbps symmetric for $30 a month (which you can now get in much of Baltimore) will be the norm in urban and suburban areas.
Besides, public utilities are not that great. They don't really advance much over time, mainly because they're insulated from competition and have a guaranteed profit margin. Water, roads, electricity utilities have all provided us pretty much the same service at the same price for decades. Don't we want something more from our ISPs? Say, companies taking big gambles on next-gen services that might fail or might become the next big thing?
I certainly won't argue that public utility-style regulation of the Internet might cut out some of the occasional asshat tactics of some big ISPs, but is turning Internet access into a stagnant-but-consistent service really worth it?
Except, of course, when you and your people are appealing to irrational fears for personal political gain and not actually representing the very consumers you're supposed to look out for. Which clearly was the case with the censored free wireless plan.
If consumers desire a filtered wireless network, then shouldn't it emerge even without the federal government forcing it upon us?
Even without rigid spectrum rules, there is nothing stopping a company from buying up spectrum rights and using it for family-friendly wireless broadband. But what if users with alternative preferences--say, parents who are capable of protecting their kids online without centralized censorship and nipple slip fines--outnumber those who want government nanny-state rules? The FCC's proposed conditions that would have mandated filtering mean that companies couldn't compete to deliver what people really want.
Any way you look at this, the plan was a perfect example of the FCC pretending it knows best what the People of the United States deserve.
It's good to hear this plan is dead. Kevin Martin backed this network so he'd look like a "family-values man" and score some points with cultural conservatives in North Carolina, where Martin has long been planning a bid for Congress.
This 25mhz of spectrum in the AWS3 band could go toward a lot of very cool services--LTE, for instance. Martin's plan--to earmark the 25mhz for 768kbps of censored wireless broadband that wouldn't even be widely deployed for a decade--is clearly not the smartest way to put these frequencies to use.
The FCC should do one of two things with this spectrum--a)auction it off with no strings attached and allow the winning firm to sell or rent the spectrum as if it were property, or b)set the band free as unlicensed flexible use spectrum subject only to basic EIRP and non-interference requirements and nothing more.
In theory, if the banking system were known to be compromised in such a huge way, and there were no way of knowing if your own bank account was compromised or not, shouldn't there be a massive bank run? Because everyone wants to withdraw their money right away to minimize the chance that this ridiculous security leak negatively affects them, right? Such a massive erosion of confidence can completely destroy a banking system.
"In the Jews' defense, a fool and his money are soon parted." -Oscar Wilde
a. Ubiquity doesn't require massive taxpayer spending. The Internet, Windows, Google, ISO standards, all emerged without much government involvement.
b. Sure, computers can play a useful role in schools, but anybody who attended public schools knows that they mostly just sit around and aren't often used for educational purposes. Money on education can be better spent on incentivizing good teachers, for one thing.
c. My stock retort to tax-and-spend leftists: So we wasted 2.5T on an unnecessary foreign war in Iraq, that somehow justifies wasting even more taxpayer money on infrastructure boondoggles that largely amount to ditch-digging for ditch-digging's sake?
d. Wow. This isn't 1939. The global economy is intrinsically intertwined, and the prospect of an all-out war that would seriously impair U.S. ability to acquire manufactured goods is incredibly remote. Besides, if other countries are busy building our goods in factories, that leaves American workers free to, you know, do the thinking that's required to make these products. There's a reason we have Intel, Microsoft, Google, Apple, etc all based right here in the U.S. And it isn't because of silly trade barriers in the name of "national security"
Obama's on the wrong track on infrastructure spending. Yes, his plans will create jobs, but at the expense of other, more economically efficient jobs. Like FDR, Obama falsely assumes that government can create wealth, when in fact government can only redistribute wealth from productive uses to unproductive ones. Besides, the real gems of U.S. infrastructure--our telecommunications system, our electricity grid--are largely in private hands, anyway.
The competitive ISP market you desire will arise on its own in the next five years because of smarter, more effective broadband technologies. At the heart of this is spectrum, which erases the rules of the game as they existed up until now and replaces them with conditions far more conducive to entry (and exit).
Laying a wire to each home is really, really expensive on a per-household basis. Verizon has spent upwards of $1,000 per home to add FiOS. Building completely from scratch is even more costly.
But wireless is a whole different story. Sure, you've got to have base stations, which require backhaul, but the costs involved there are much more bearable than those involved with old-fashioned overbuilding.
To see why wireless broadband is such an important vehicle for broadband competition, consider Wi-Max. It's already giving DSL a run for its money in Baltimore, because it offers reliable throughput of 3Mbps down/2Mbps up for $25 a month, with no strict usage limits. Of course, Wi-Max isn't the end-all of wireless, but merely a stopgap until Long-Term Evolution really takes off.
LTE will completely change the game when it starts getting widely deployed in a couple years. After the 700mhz band is vacated in a couple months, T-Mobile, Sprint, Verizon, and AT&T will all work on building out LTE networks, which will offer at least 10Mbps for the same price as a typical home broadband connection.
Six to eight ISP choices may not mean as many choices as we have for burgers, but it's still a lot of competition, especially compared to what we have now. And it is surely enough to ensure that ISPs can't get away with the heavy-handed tactics that have angered so many of us recently.
It wouldn't surprise me if by the time Congress finally gets around to regulating ISPs, the whole justification for regulating them in the first place is no longer logically valid.
I haven't noticed any hangs, but SMART reports 41165316 hardware read errors. ECC is recovering all of them, but it's still troubling, as I haven't seen nearly that many errors on any other healthy disk I've ever owned. But the drive is performing fine, so perhaps I'm worried about nothing.
Right now, the biggest drives available that offer real reliability are in the 500-640GB range. I'd suggest the WD6400AAKS, which is truly an amazing drive (and under 12 cents a Gigabyte, too)
You make a very valid point. As sites like Newegg start to get hundreds--even thousands--of reviews for products like hard disks, a prospective buyer can look at the percentage of negative reviews as a rough guide to the probability of failure. Compared to the 640GB WD6400AAKS, pretty much all 1TB-plus drives on Newegg have a lot of 1-star reviews. That does suggest these huge drives aren't up to par in terms of reliability.
No, having a high market share does not make you a monopoly unless you have the power of a monopoly. Since dynamic markets (like Web advertising) are highly contestable, the ever-present threat of entry by new firms forces even the most dominant players to behave as if they weren't monopolies.
If Google were really a monopoly, it wouldn't have to work so damn hard. The folks in Mountain View could just cut prices temporarily whenever somebody got too close, and otherwise sit on their chairs and twiddle their thumbs.
But that's not what Google is doing. Google continues to develop new products and improve existing ones pretty rapidly, because it knows that it operates in a cutthroat marketplace with intense competition.
How is Google not playing by the capitalist rules? Just ten years ago, Google was a grad student research project, and now it is the global leader in Web advertising. It's crucial to realize how Google came about, because the firm that ultimately dethrones Google will emerge unexpectedly from humble roots.
Look, nobody is forcing Internet users to rely on Google for search. The reason for Google's continued dominance isn't because it is an evil monopoly, but because Google managed to build a platform that a lot of people like. The success of Google has spurred competitors like Microsoft to invest serious cash to develop a superior set of online services, Isn't that exactly what capitalism is all about?
Cloud computing may be the next big thing, but it'll be quite some time before cloud computing can give computing on the ground a real run for its money.
Last-mile broadband, for most smaller businesses and homes, is synonymous with highly asymmetric connections. Since many applications don't function on networks with strained upstream bandwidth, only enterprise-grade data service sare able to offer the robust speeds needed for cloud computing to excel.
And web-based applications still can't approach locally hosted apps in terms of reliability and performance. This is why so many Gmail users connect via IMAP using their Outlook client instead of relying on Gmail's sometimes glitchy web interface.
One's body parts, if willingly defaced, are surely not immune from trademark infringement suits. Imagine a lowlife tatooing some valuable intellectual property on his arm and engaging in negative activities that might harm the value of the trademark. That'd surely be grounds for a hefty lawsuit.
For not much more money RAID 10 means significantly greater performance and a good deal more redundancy to boot. With 1.5TB for $189 these days, it doesn't make a whole lot of sense to do RAID 5 unless you're penny pinching and performance is of no concern.
You'd be surprised how quickly 50GB Blu-Ray images add up. I have 6 TB in a RAID 10 currently, and within 18 months I am confident I'll be at twice that, at least. With 1.5TB disks for $189 these days, who doesn't gradually amass terabytes of data?
OK, let's say you can cut the rate of death by a factor of 10 by driving safely.
There are, on average, 40000 automobile deaths per year in the U.S., or roughly 1 per 7500 people. So a safe driver's risk of dying in a car accident would be 1 in 75000.
http://www.ntsb.gov/aviation/Paxfatal.htm
But there are only 16 deaths per year, on average for the past six years, due to airline crashes in the United States. That's 1 per 18,750,000 people.
So even if you could reduce your chances of dying in a car crash by a factor of 100, airline travel would still be significantly safer than driving.
ISPs have always demonstrated preference for content. Inbound port 80 traffic and excessive SMTP traffic are just two examples of commonly restricted types of traffic. Besides, even websites like YouTube actively remove videos with explicit scenes. Does that mean YouTube is liable every time somebody posts an infringing video, even if YouTube isn't actually aware of it?
78 % of homes with a phone line can get DSL.
9 out of 10 people live in a home that is in a 3G coverage area.
Anybody in the lower 48 with a clear view of the southern sky can get satellite.
Are these options often slower and pricier than cable? Sure. But if you hate cable so much, why aren't you willing to put up with slower speeds or higher latency in exchange for an ISP that doesn't forge packets?
Life is full of goods and services for which there's no perfect substitute, but that doesn't mean that there aren't substitutes at all. Being able to browse the web at 768kbps isn't lightning fast, but it's surely broadband and a hell of a lot better than dial-up.
Here's the thing: Under the Administrative Procedure Act, federal agencies have to follow specific rules when drafting legally binding principles. The Internet Policy Statement from 2005 was a non-binding statement and it didn't go through the notice and comment procedure that is required for a rule to become enforceable.
Safe Harbor, not common carrier, is what protects Comcast as per the DMCA and the CDA. Common carrier is a completely different concept that affects telcos, not cable companies. Modifying TCP streams--however repugnant--does not automatically mean the ISP is liable for the content that traverses its network. That's the law, like it or not.
Where have you been for the last decade? Broadband has always cost around $40 to $50 a month, but speeds keep moving upward on average as cost per byte decreases. Back when I first got broadband in 1998, it was 768k per second. Now I get 16/2 for $52.95 per month.
Monopolists may not behave like perfect competitors, but they DO pass some savings on to consumers, albeit less than an optimal amount.
In the case of employee rights, what's at stake aren't property rights, but the right to engage in voluntary private transactions.
Smoking isn't just something that occurs in bars for no reason. It confers utility on people--believe it or not, there are rational adults out there who knowingly choose to smoke cigarettes even though long-term smoking is known to reduce lifespan by a few years. Are people who engage in this activity crazy, or stupid, or irrational? Hardly. Some people really like smoking, enough so that they are willing to shed years of life. Just as many people choose to eat bacon cheeseburgers for lunch every day and drink a six pack of Bud every night, some people prefer a shorter life that involves smoking to a longer one that doesn't.
Furthermore, the magnitude of exposure to secondhand smoke is hardly comparable to, say, the risk of losing an eye in an industrial accident that can be mitigated with goggles. I gladly concede that prolonged employment in a smoke-filled bar reduces life expectancy and increases the risk of lung cancer and emphysema, but working in a bar 20 hours a week for a couple years while an undergrad is not likely to have a measurable impact on life expectancy. It may cause some short-term health problems, of course, but such ailments can usually be alleviated by simply switching to one of the many places of employment where ambient secondhand smoke is non-existent.
People want to be able to smoke in bars. Bar owners want to be able to offer patrons a place where they can smoke while drinking. People willingly enter in to employment in establishments where smoking takes place, despite the fact that it's well established that prolonged exposure to secondhand smoke is detrimental to human health. These are all voluntary decisions made by adults. If we believe people should be responsible for making decisions for themselves about what risks to take in life, then what justification is there for government to intervene?
Relying on government to protect people from themselves is is wrong when it comes to the War on Drugs, and it is wrong when it comes to bans on smoking in pubs to protect employees who voluntarily decide to work in a bar.