It's probably worth clarifying why that matters. Trucks don't start out at the first delivery location; they have to be driven there. If you can reduce the number of trucks, then there are fewer people wasting fuel driving out to the first delivery location. The larger the geographical area covered by a single depot, the greater the savings.
Yeah, this is a hideous workaround. Apple's design is much saner, providing a default probe score based on how many properties were matched, then calling a probe method in the driver to give it an opportunity to dynamically change its probe scores for even more control. So with that scheme, the generic Windows driver would match based on something generic like vendor and device subclass, the NVidia reference driver would match with vendor + product (and optionally add bcdDevice), and the custom driver from the OEM would presumably return a higher probe score dynamically so that it always wins.
Version numbers and release dates have no legitimate place in driver matching behavior, IMO.
That's not a coverage issue, that's a congestion issue.
I would argue that congestion issues are a form of coverage issues. Full coverage requires that cell sites' coverage area must overlap by a large enough margin that you are able to start talking to a new cell site before you lose the old one. If the towers' fringe reception area doesn't overlap enough out to allow a seamless switch under typical load, you need more towers to have full coverage in any meaningful sense of the word. But that's just me.
I suspect this is some combination of the towers not shedding load aggressively enough (forcing a handoff for cell handsets that are within range of a less busy tower) and cell phones dropping the connection to the old tower too soon after a handoff then losing the new cell site and being unable to reconnect to the old one quickly, but that's just a gut feeling. I've noticed that rain fade seems to make the problem a lot worse in many of the places that I drive, which tells me that low single-digit dB make the difference between towers overlapping and not. So there's definitely a general lack of towers involved, at least in the greater SF Bay Area.
Probably 98% of the people trying to do these hacks are trying to use self-signed certs for testing purposes and then accidentally leave it in there. Adding letsencrypt support as a UI configuration setting in OS X Server would go a long way towards eliminating this problem, at least of iOS developers.
For the remaining 2%—the people who need to do something particularly unusual involving devices generating their own certs—the problem is caused by iOS lacking the same UI that OS X has for allowing the user to trust a cert. OS X made it pretty easy to add trust. Unfortunately, iOS lacks that mechanism almost entirely, and requires cumbersome workarounds at the app level to get the functionality back.
If that counts as "staying away from PKI", then yeah.
I'm sure they existed in the disposable products market (e.g. the crap that you get free with audio guide rental), but they certainly weren't the norm pre-iPod (2001). Either way, I have good reasons for blaming Apple for the reduction in quality:
In 2001, Apple's iPods used an undersized 1/8" connector back that made it impossible to build headphones without moving to a smaller wire size.
In 2007, the original iPhone's bad case design had an indentation for the headphone jack that was just big enough for Apple's undersized plug back, preventing use of most standard headphones without an adapter (dongle).
Headphone makers rapidly switched to use the same undersized connector, which in turn forced them to use smaller wires.
Because most new headphones used the smaller connectors, third-party case manufacturers designed the openings in their cases to meet Apple's substandard headphone connector specs; normal headphone plugs wouldn't fit, so headphone manufacturers couldn't move back to bigger wire sizes even if they started failing in volume (which some did).
So Apple drove the market towards lower-quality headphone cables, and thus is directly responsible for the poor quality of wired headphones that we see today. I'm not saying they're the only responsible party, but they played a pivotal role, without which the market would look very different.
Just to clarify, I've never had the jack fail. I've had plenty of wires fail. I mostly blame Apple for that. At some point, they decided to make headphones with really thin wires under the premise that more flexibility is somehow better. Unfortunately, that makes them more fragile. And then everybody copied their design, and now you almost can't buy headphones with wires that aren't thin and flimsy.
I use 1/8" plugs on an almost daily basis, and the last time I had a mini jack failure was on a PowerBook 145, back in the mid-1990s. By contrast, I've gone through three micro-USB-to-Lightning adapters (from different manufacturers) in less than a year. They get to the point where they only make proper contact if you shove them in at a certain exact angle, and the slightest touch causes them to disconnect.
I've given up on micro-USB. The failure rate is just plain staggering compared with anything else I've worked with. It is a terrible, horrible, flaky, unreliable nightmare connector suitable for use only in disposable products. If you haven't had problems with it, then you obviously aren't being nearly as hard on your charge connectors as you are on your headphones.
That isn't necessarily true. A FOIA request can potentially return information about people not associated with the collection of that data. On the whole, such personal information is supposed to be scrubbed in response to a FOIA request, but need not be scrubbed if the person requesting it is the person that the information is about.
Additionally, the response to the FOIA request could potentially contain information that makes people look bad when presented out of context. Partial disclosure of that information could cause irreparable harm, and the person harmed arguably has a right to know who disclosed that information so that they can take legal action. That isn't possible if nobody knows who filed the FOIA request.
I don't think so. Rather, I think the blame lies with developers using self signed certs on their test servers, and tweaking the SSL options in their app so that works for them.
I'm fairly certain of it. Just a couple of days ago, I was helping somebody out on Stack Overflow who couldn't get self-signed certs to work. I posted a full code snippet that did proper validation before accepting the provided cert, and the person asking the question then asked if a different snippet would work that didn't check anything from the cert other than ensuring that the hostname matched. And I see that same defective code snippet over and over in people's comments, so I can only assume that somewhere on the Internet, somebody has posted that horrible travesty as an example of how to handle self-signed certificates.
That's why I wrote documentation to explain how to do it correctly back when I worked at Apple. Unfortunately, Apple has not bothered to keep it up-to-date since then with relevant NSURLSession-based examples, but the concepts are still valid, and it should be straightforward enough for anyone who bothers to read the chapter in question to convert it to use NSURLSession. Instead, developers search the Internet for pre-written code snippets that are almost invariably wrong—usually in dangerous ways—and use them, resulting in this mess.
So even though the developers should take some of the blame, Apple deserves a lot of the blame for not updating the NSURLSession Programming Guide with new Objective-C code based on the current API, not providing Swift versions of those snippets, etc. And Apple's efforts to make programming easier, along with the efforts of sites like Stack Overflow, also deserve some of the blame, because at some point, there needs to be a bar that says, "You must be this tall to ride the ride." Encouraging developers to not take the time to learn how to write software properly can only lead to software quality that gets progressively worse until the wheels fall off the bus.
You can develop a resistance to H2 antagonists. My gut says you're probably better off with an antacid that directly neutralizes the root cause. That way, you can reduce the acid in your stomach only at night when you're most likely to experience reflux, while leaving it fully acidic during the day when that acid is actually playing a crucial role in digestion.
In spite of the horribly rude tone, there's a hint of truth to that. A big reason why people get snarky is because so many people don't even bother to try to figure things out before they ask for help. A sizable percentage of people seem to be completely helpless when anything goes wrong. They don't know how to do a Google search, they don't know how to read for comprehension, and they don't know how to figure out what things to look for when skimming/searching documentation for solutions to their problems. This lack of critical thinking skills is quite alarming.
As a result, even those of us who still try to help tend to point people to the right piece of documentation first, waiting to re-explain things until after they come back and say that they still don't understand something. And after a few rounds, even I have to say, "Read the doc and figure it out." After all, my job is not to write your code for you. I'll try to help, and I'll try to steer you in the right direction, but there are limits.
Cynically, I place the blame for these problems squarely at the feet of Apple for trying to dumb down programming, technical documentation, computer use, etc. to the point where people don't have to think to code, rather than saying, "You must be this tall to ride the ride." The result is a bunch of people who don't bother to think and who expect others to do the thinking for them. They've bred a whole class of "duh-velopers" who literally can't do much more than piece together code snippets and tweak them slightly. Heaven help them if a snippet contains something like "insert your customization here", because they go slack-jawed. And this results in everybody who actually understands what's going on having to waste a lot of time explaining things that should have been obvious.
IMO, you can't fix one problem without fixing both. People are jerks because the newbies have driven all the nice people away by incessantly asking questions whose answers should be obvious to anybody who actually read and comprehended the docs, and most of the people who didn't comprehend the docs are still not going to understand it no matter how many times you explain it. Fix the clueless question problem, and people who are able to actually figure out what they're doing will stick around and will continue to be helpful. Short of that, nothing will help in the long run.
To some degree, that is probably best solved by reputation-bssed segregation. Anybody should be able to answer any questions, but until you get rep, your questions should be initially seen only by other newbies (and if no newbie can answer them, they would then bubble up to folks with more rep). Rep should be awarded for asking good questions or giving good answers. Clueless people who are incapable of asking good questions and giving good answers should thus remain stuck in the newbie question cesspool while the adults discuss real issues.
That "more-or-less" is the main problem. The lottery system guarantees abuse by ensuring that companies that employ vast quantities of H1-B employees apply for a ridiculous number of H1-B visas and thus get most of the visas issued; other companies largely get screwed. The only way a lottery system could possibly work would be if there were a limit of five applications per company per year.
The secondary problem is that there aren't sufficient rules for what constitutes a "specialty" position. Low-level computer programmers shouldn't qualify for H1-Bs, whether they are being paid $60k or $130k. The problem is not the dollar amount, but rather the lack of proper vetting of the allowed job categories. Although setting a high fixed dollar amount means that companies like Tata will cease to be viable, it will also effectively set a maximum base salary for software engineers beyond which it will make more sense to hire H1-Bs, and thus will effectively deflate wages for everyone.
So it is a very temporary solution that has seriously bad side effects. The only problem that really needed to be solved was the idiotic lottery system. Moving to a "priority based on wage relative to median pay in a given field in a given geographic area" model by itself would have fixed the problem without causing any of the undesirable side effects. By making two changes instead of just one, they caused as many problems as they solved.
And how will folks in less expensive areas be "punished"? Top talent has already moved to the coast, so you're already hiring local.
Companies in less expensive areas will be punished, because A. there's a shortage of programmers everywhere but on the coast because most of the jobs are on the coast, B. those areas are places where H-1B visas would most be useful, and C. a $130,000 salary is 2–2.5x what a programming position would realistically pay in those areas, thus making H1-B employees completely out of reach and causing them to resort to less desirable strategies (e.g. outsourcing).
Or how about we just stop rolling out physical lines to individual houses in rural areas and migrate them all over to wireless technologies. Saves tons of money and is far easier to repair and upgrade.
I would argue that replacing the transceivers on either end of a fiber is a lot easier than replacing antennas on a tower, though that is arguably easier than pulling additional fibers for tens of miles, so which one is easier depends on the nature of the upgrade.
Well, I'm a Republican, and I'm fine with local governments, maybe even state governments deciding to create universal fiber infrastructure. I think that, going forward, we'll want to consider this as critical infrastructure, just like power, water, sewer, and street access. My only caveat would be to let people decide regionally how they want to handle this, rather than making some mess of a Federal bureaucracy to decide these things for everyone, and probably do it badly and expensively, just like the giant telcos.
The important part is that the feds should fund it so that it is available nationwide instead of only in areas whose leadership understands the importance. I'm okay with it being set up and run regionally or locally, so long as there are some nationwide minimum standards, e.g.
Mandatory 100% fiber coverage in areas with population density above some reasonably low limit
Wireless access as an optional alternative only in areas with extremely low density, such as farms
A requirement that the lines be owned by either a nonprofit or government entity and leased to providers in a nondiscriminatory fashion
Of course, that would require that Congress be willing to relinquish some control, rather than over-regulating it into dysfunction. I'm not entirely convinced they're capable of that (either the Democrats or the Republicans), but in theory, it could work, anyway.:-)
At the moment, there are municipalities who want to do this and are blocked. That really needs fixing.
A thousand times this. We need to pass a law that invalidates all of those state laws at the federal level. Either that or get the courts to rule that it violates the equal protection clause or interstate commerce clause somehow. Either way.
P.S. Thanks for at least engaging me rationally instead of modding me a Troll for... hell, I don't even know why. For having a contrary opinion to someone else, I guess? Apparently, that's the definition of a "troll" these days.
What do you think this is, Facebook?:-D
But seriously, yeah, there's nothing worse than people just turning you off for having a different opinion. That never moves the world forward. Better to disagree, figure out why you disagree, look for common ground, and when necessary, scrap everything and look for entirely different approaches that satisfy both sides of the political spectrum.
The phone grid has always been owned by for-profit companies. A better comparison is the electrical grid, which at least at the distribution level is typically owned by a nonprofit ISO. That's why many cities are able to easily offer multiple choices in power provider; the wires aren't owned by a company.
Well, we said the same thing about Wheeler, who had similar credentials, and he ended up being a pretty decent consumer advocate. Pai is not interested in net neutrality, but in removing regulation and barriers to actual competition - or so he says. That could work as well as FCC regulation in theory, or maybe even better.
Only if you're in a major city, at best. Everywhere else (and even in many parts of major cities), the biggest barrier to actual competition is the cost of actually running the lines. In rural areas, the cost to run fiber to a single customer could easily be $50k. If an ISP can only make $600 per year, a second ISP would have to be utterly insane to try to compete.
What we really need—and what I suspect no Republican would ever even consider doing, unfortunately—is for the government to build out the infrastructure and create a permanently government-owned nonprofit a la TVA to maintain it, then lease access to that fiber to any ISP that wants to provide service. Once you eliminate the need for competitors to provide independent, expensive infrastructures, suddenly the barriers to competition in the ISP space become almost nonexistent.
Do Indian companies make any components that could go into an iPhone? Last I heard, there weren't any fabs in India, which probably means there aren't any SMT parts makers, either. I mean, I suppose they could mill the enclosures, fabricate the PCBs, and mould the Gorilla Glass, and maybe say that it is 30% by weight, but I don't think that's what they mean. The best they could do is "source" the parts from an Indian company that then imports them from China/Japan/Taiwan, but I fail to see how that's going to encourage manufacturing of electronics in India in any meaningful way. Time will tell, I suppose.
False comparison. If there were a hundred idiots standing around the crosswalk making noise with their kazoos, then the signal would pretty much be useless for the blind, and could be turned off.
Don't know about kazoos, but I've noticed mockingbirds getting pretty good at imitating the cuckoo sounds from them. I'm starting to wonder if they're part of some vast animal conspiracy along with the cats to rid the world of humans once and for all, but maybe I'm just paranoid.
The current House bill to raise the H1-B minimum wage to $130k, and to allocate all H1-B slots based on salary rather than lottery - this is a great fix.
No, it's a temporary fix. In ten or fifteen years, $130k will fall below the median income for programmers in the Bay Area, and it will once again become cheaper to mass-import H1-B engineers. It also results in an unreasonable salary for programmers in other parts of the country, where $130k is more than a 50% premium above the prevailing wage.
A truly great fix would be to raise the H1-B minimum wage to 120% of the median income for jobs with a given job title within your particular geographic region. Want a programmer in Atlanta? It'll cost you about $96k. Want a networking engineer in the Bay Area? Probably more like $160k. And priority should be based on percentage of the regional median, so folks in less expensive areas won't get punished relative to their Bay Area counterparts.
A high-deductible health insurance plan without subsidies costs only about $4,000 per year. Employers aren't required to pay for insurance for anyone other than the employee, so that $25,000 number is pure fiction.
Slashdot Mirror
It's probably worth clarifying why that matters. Trucks don't start out at the first delivery location; they have to be driven there. If you can reduce the number of trucks, then there are fewer people wasting fuel driving out to the first delivery location. The larger the geographical area covered by a single depot, the greater the savings.
Um... no, there are lots of third-party drivers available in OS X. Maybe you're thinking of its mutant halfling spawn, iOS?
Yeah, this is a hideous workaround. Apple's design is much saner, providing a default probe score based on how many properties were matched, then calling a probe method in the driver to give it an opportunity to dynamically change its probe scores for even more control. So with that scheme, the generic Windows driver would match based on something generic like vendor and device subclass, the NVidia reference driver would match with vendor + product (and optionally add bcdDevice), and the custom driver from the OEM would presumably return a higher probe score dynamically so that it always wins.
Version numbers and release dates have no legitimate place in driver matching behavior, IMO.
I would argue that congestion issues are a form of coverage issues. Full coverage requires that cell sites' coverage area must overlap by a large enough margin that you are able to start talking to a new cell site before you lose the old one. If the towers' fringe reception area doesn't overlap enough out to allow a seamless switch under typical load, you need more towers to have full coverage in any meaningful sense of the word. But that's just me.
I suspect this is some combination of the towers not shedding load aggressively enough (forcing a handoff for cell handsets that are within range of a less busy tower) and cell phones dropping the connection to the old tower too soon after a handoff then losing the new cell site and being unable to reconnect to the old one quickly, but that's just a gut feeling. I've noticed that rain fade seems to make the problem a lot worse in many of the places that I drive, which tells me that low single-digit dB make the difference between towers overlapping and not. So there's definitely a general lack of towers involved, at least in the greater SF Bay Area.
Probably 98% of the people trying to do these hacks are trying to use self-signed certs for testing purposes and then accidentally leave it in there. Adding letsencrypt support as a UI configuration setting in OS X Server would go a long way towards eliminating this problem, at least of iOS developers.
For the remaining 2%—the people who need to do something particularly unusual involving devices generating their own certs—the problem is caused by iOS lacking the same UI that OS X has for allowing the user to trust a cert. OS X made it pretty easy to add trust. Unfortunately, iOS lacks that mechanism almost entirely, and requires cumbersome workarounds at the app level to get the functionality back.
If that counts as "staying away from PKI", then yeah.
And neither does spelling, apparently.
I'm sure they existed in the disposable products market (e.g. the crap that you get free with audio guide rental), but they certainly weren't the norm pre-iPod (2001). Either way, I have good reasons for blaming Apple for the reduction in quality:
So Apple drove the market towards lower-quality headphone cables, and thus is directly responsible for the poor quality of wired headphones that we see today. I'm not saying they're the only responsible party, but they played a pivotal role, without which the market would look very different.
Just to clarify, I've never had the jack fail. I've had plenty of wires fail. I mostly blame Apple for that. At some point, they decided to make headphones with really thin wires under the premise that more flexibility is somehow better. Unfortunately, that makes them more fragile. And then everybody copied their design, and now you almost can't buy headphones with wires that aren't thin and flimsy.
I use 1/8" plugs on an almost daily basis, and the last time I had a mini jack failure was on a PowerBook 145, back in the mid-1990s. By contrast, I've gone through three micro-USB-to-Lightning adapters (from different manufacturers) in less than a year. They get to the point where they only make proper contact if you shove them in at a certain exact angle, and the slightest touch causes them to disconnect.
I've given up on micro-USB. The failure rate is just plain staggering compared with anything else I've worked with. It is a terrible, horrible, flaky, unreliable nightmare connector suitable for use only in disposable products. If you haven't had problems with it, then you obviously aren't being nearly as hard on your charge connectors as you are on your headphones.
That isn't necessarily true. A FOIA request can potentially return information about people not associated with the collection of that data. On the whole, such personal information is supposed to be scrubbed in response to a FOIA request, but need not be scrubbed if the person requesting it is the person that the information is about.
Additionally, the response to the FOIA request could potentially contain information that makes people look bad when presented out of context. Partial disclosure of that information could cause irreparable harm, and the person harmed arguably has a right to know who disclosed that information so that they can take legal action. That isn't possible if nobody knows who filed the FOIA request.
I'm fairly certain of it. Just a couple of days ago, I was helping somebody out on Stack Overflow who couldn't get self-signed certs to work. I posted a full code snippet that did proper validation before accepting the provided cert, and the person asking the question then asked if a different snippet would work that didn't check anything from the cert other than ensuring that the hostname matched. And I see that same defective code snippet over and over in people's comments, so I can only assume that somewhere on the Internet, somebody has posted that horrible travesty as an example of how to handle self-signed certificates.
That's why I wrote documentation to explain how to do it correctly back when I worked at Apple. Unfortunately, Apple has not bothered to keep it up-to-date since then with relevant NSURLSession-based examples, but the concepts are still valid, and it should be straightforward enough for anyone who bothers to read the chapter in question to convert it to use NSURLSession. Instead, developers search the Internet for pre-written code snippets that are almost invariably wrong—usually in dangerous ways—and use them, resulting in this mess.
So even though the developers should take some of the blame, Apple deserves a lot of the blame for not updating the NSURLSession Programming Guide with new Objective-C code based on the current API, not providing Swift versions of those snippets, etc. And Apple's efforts to make programming easier, along with the efforts of sites like Stack Overflow, also deserve some of the blame, because at some point, there needs to be a bar that says, "You must be this tall to ride the ride." Encouraging developers to not take the time to learn how to write software properly can only lead to software quality that gets progressively worse until the wheels fall off the bus.
Worstern Digital?
Don't worry. Uber will make it happen in three years by ignoring all of those pesky regulations.
You can develop a resistance to H2 antagonists. My gut says you're probably better off with an antacid that directly neutralizes the root cause. That way, you can reduce the acid in your stomach only at night when you're most likely to experience reflux, while leaving it fully acidic during the day when that acid is actually playing a crucial role in digestion.
In spite of the horribly rude tone, there's a hint of truth to that. A big reason why people get snarky is because so many people don't even bother to try to figure things out before they ask for help. A sizable percentage of people seem to be completely helpless when anything goes wrong. They don't know how to do a Google search, they don't know how to read for comprehension, and they don't know how to figure out what things to look for when skimming/searching documentation for solutions to their problems. This lack of critical thinking skills is quite alarming.
As a result, even those of us who still try to help tend to point people to the right piece of documentation first, waiting to re-explain things until after they come back and say that they still don't understand something. And after a few rounds, even I have to say, "Read the doc and figure it out." After all, my job is not to write your code for you. I'll try to help, and I'll try to steer you in the right direction, but there are limits.
Cynically, I place the blame for these problems squarely at the feet of Apple for trying to dumb down programming, technical documentation, computer use, etc. to the point where people don't have to think to code, rather than saying, "You must be this tall to ride the ride." The result is a bunch of people who don't bother to think and who expect others to do the thinking for them. They've bred a whole class of "duh-velopers" who literally can't do much more than piece together code snippets and tweak them slightly. Heaven help them if a snippet contains something like "insert your customization here", because they go slack-jawed. And this results in everybody who actually understands what's going on having to waste a lot of time explaining things that should have been obvious.
IMO, you can't fix one problem without fixing both. People are jerks because the newbies have driven all the nice people away by incessantly asking questions whose answers should be obvious to anybody who actually read and comprehended the docs, and most of the people who didn't comprehend the docs are still not going to understand it no matter how many times you explain it. Fix the clueless question problem, and people who are able to actually figure out what they're doing will stick around and will continue to be helpful. Short of that, nothing will help in the long run.
To some degree, that is probably best solved by reputation-bssed segregation. Anybody should be able to answer any questions, but until you get rep, your questions should be initially seen only by other newbies (and if no newbie can answer them, they would then bubble up to folks with more rep). Rep should be awarded for asking good questions or giving good answers. Clueless people who are incapable of asking good questions and giving good answers should thus remain stuck in the newbie question cesspool while the adults discuss real issues.
That "more-or-less" is the main problem. The lottery system guarantees abuse by ensuring that companies that employ vast quantities of H1-B employees apply for a ridiculous number of H1-B visas and thus get most of the visas issued; other companies largely get screwed. The only way a lottery system could possibly work would be if there were a limit of five applications per company per year.
The secondary problem is that there aren't sufficient rules for what constitutes a "specialty" position. Low-level computer programmers shouldn't qualify for H1-Bs, whether they are being paid $60k or $130k. The problem is not the dollar amount, but rather the lack of proper vetting of the allowed job categories. Although setting a high fixed dollar amount means that companies like Tata will cease to be viable, it will also effectively set a maximum base salary for software engineers beyond which it will make more sense to hire H1-Bs, and thus will effectively deflate wages for everyone.
So it is a very temporary solution that has seriously bad side effects. The only problem that really needed to be solved was the idiotic lottery system. Moving to a "priority based on wage relative to median pay in a given field in a given geographic area" model by itself would have fixed the problem without causing any of the undesirable side effects. By making two changes instead of just one, they caused as many problems as they solved.
Companies in less expensive areas will be punished, because A. there's a shortage of programmers everywhere but on the coast because most of the jobs are on the coast, B. those areas are places where H-1B visas would most be useful, and C. a $130,000 salary is 2–2.5x what a programming position would realistically pay in those areas, thus making H1-B employees completely out of reach and causing them to resort to less desirable strategies (e.g. outsourcing).
I would argue that replacing the transceivers on either end of a fiber is a lot easier than replacing antennas on a tower, though that is arguably easier than pulling additional fibers for tens of miles, so which one is easier depends on the nature of the upgrade.
The important part is that the feds should fund it so that it is available nationwide instead of only in areas whose leadership understands the importance. I'm okay with it being set up and run regionally or locally, so long as there are some nationwide minimum standards, e.g.
Of course, that would require that Congress be willing to relinquish some control, rather than over-regulating it into dysfunction. I'm not entirely convinced they're capable of that (either the Democrats or the Republicans), but in theory, it could work, anyway. :-)
A thousand times this. We need to pass a law that invalidates all of those state laws at the federal level. Either that or get the courts to rule that it violates the equal protection clause or interstate commerce clause somehow. Either way.
What do you think this is, Facebook? :-D
But seriously, yeah, there's nothing worse than people just turning you off for having a different opinion. That never moves the world forward. Better to disagree, figure out why you disagree, look for common ground, and when necessary, scrap everything and look for entirely different approaches that satisfy both sides of the political spectrum.
The phone grid has always been owned by for-profit companies. A better comparison is the electrical grid, which at least at the distribution level is typically owned by a nonprofit ISO. That's why many cities are able to easily offer multiple choices in power provider; the wires aren't owned by a company.
Only if you're in a major city, at best. Everywhere else (and even in many parts of major cities), the biggest barrier to actual competition is the cost of actually running the lines. In rural areas, the cost to run fiber to a single customer could easily be $50k. If an ISP can only make $600 per year, a second ISP would have to be utterly insane to try to compete.
What we really need—and what I suspect no Republican would ever even consider doing, unfortunately—is for the government to build out the infrastructure and create a permanently government-owned nonprofit a la TVA to maintain it, then lease access to that fiber to any ISP that wants to provide service. Once you eliminate the need for competitors to provide independent, expensive infrastructures, suddenly the barriers to competition in the ISP space become almost nonexistent.
Do Indian companies make any components that could go into an iPhone? Last I heard, there weren't any fabs in India, which probably means there aren't any SMT parts makers, either. I mean, I suppose they could mill the enclosures, fabricate the PCBs, and mould the Gorilla Glass, and maybe say that it is 30% by weight, but I don't think that's what they mean. The best they could do is "source" the parts from an Indian company that then imports them from China/Japan/Taiwan, but I fail to see how that's going to encourage manufacturing of electronics in India in any meaningful way. Time will tell, I suppose.
Don't know about kazoos, but I've noticed mockingbirds getting pretty good at imitating the cuckoo sounds from them. I'm starting to wonder if they're part of some vast animal conspiracy along with the cats to rid the world of humans once and for all, but maybe I'm just paranoid.
To be pedantic, it isn't a change to the immigration program at all. H1-B visas are classified as non-immigrant visas.
No, it's a temporary fix. In ten or fifteen years, $130k will fall below the median income for programmers in the Bay Area, and it will once again become cheaper to mass-import H1-B engineers. It also results in an unreasonable salary for programmers in other parts of the country, where $130k is more than a 50% premium above the prevailing wage.
A truly great fix would be to raise the H1-B minimum wage to 120% of the median income for jobs with a given job title within your particular geographic region. Want a programmer in Atlanta? It'll cost you about $96k. Want a networking engineer in the Bay Area? Probably more like $160k. And priority should be based on percentage of the regional median, so folks in less expensive areas won't get punished relative to their Bay Area counterparts.
A high-deductible health insurance plan without subsidies costs only about $4,000 per year. Employers aren't required to pay for insurance for anyone other than the employee, so that $25,000 number is pure fiction.