"you get your connectivity from Pedro, and something goes wrong."
That's between you and Pedro, not you and AT&T. Let Pedro duke it out with them. He's not going to let them screw him over or you'll take your business down the road to Shnycorp.
The reality is that Pedro provides substantially better service for a couple of bucks more than Shinycorp
Shinycorp do big nationwide advertising and low prices, by cutting deals with AT&T for lower quality of service and longer times to get onsite than AT&T sells as standard.
The result is that when your line goes down with Shinycorp, you argue for hours with their phone droids and go through the entire modem waggling routine before they admit there might be a problem and someone will be out next week to fix it.
When you ring Pedro, you speak to Pedro's tech, who looks at your circuit and goes "I'm on it. There'll be someone onsite by the end of the day"
Shinycorp competes with Bastardcorp, Screwupcorp and Cheatemcorp. They all have low prices with various gotchas and they all offer the same poor level of service. They have the highest rate of customer complaints in the industry but everyone goes with them because they all put saturation advertising in the media, so joe average has never heard of the others. They make millions each day and everyone hates them but doesn't look elsewhere.
Pedro makes a comfortable income. You're happy with your level of service. He lives or dies on this, so it's in his interest to keep his customers happy.
Fantasy? No, that's what happens in places with LLU and fairly closely describes the way I buy my connection
"My point was that points needed to be constantly adjusted."
Yup. I put a homemadeTAI system in my mother's car (A 1974 Datsun B210) back in the 80s and it went from needing point adjustments every 5000 miles to almost never (about 20k miles due to the rubbing block wearing out)
When I changed the points over to an optocoupler assembly the mechanic was horrified, but having eliminated the rubbing block the timing stayed solid for 60,000 miles.
As a nice side effect the engine could whiz through to 9000rpm if you weren't careful (6000rpm redline) but as we'd balanced everything this wasn't too much of an issue
The pickup runs off a notched tooth ring on the flywheel plus another on the camshaft if you're lucky. each plug has its own coil pack. Wasted spark is common (partly because it ensures unburned hydrocarbons don't go down the exhaust)
It's been a long time since cars had distributors.
"Sadly, even most shops now are useless as they just plug the computer in and do what it tells them"
Yup. A GF spent about $3000 getting her mercedes C220 fixed because of this - and it made no difference whatsoever.
The actual fault was a jammed open thermostat. Replacing that ( $50 because it comes in a special housing) solved every other problem.
Because the engine remained stone cold when it should have warmed up the computers got confused and reported alarm codes out the wazoo. A competent mechanic would have looked for a common cause. Competent mechanics at dealerships are a rarity.
You can use a paper clip on (E)ODB cars too, or with some there's a "secret" mode for getting the readout (My nissan has something like keyoff/keyon/keyoff/keyon/ 5 *fullthrottle/off then hold the pedal down for 10 seconds)
It's generally easier to just use an ODB dongle though.
"I prefer to buy an import from a country where the engineers can count and know what metric is, like Germany or Japan. Every fastener will be metric, except spark plugs of course."
And the bolts holding the seatbelt anchors. For some reason even in europe and japan these are SAE
Presumably this is to prevent someone kludging something unsuitable in.
"However, I've never seen a single reference to a study of the effects of windmills on regional wind patterns, massive areas of solar panels on regional temperature/wind/etc., let alone manufacturing of these things. Are they issues?"
In a nutshell: not really. The effects on wind pattern and temperature is negligable, even when compared to the downwind effects we already generate from cities.
In a larger nutshell: With appropriate nuclear technology, these things become totally superfluous. Solar and wind sources are notoriously "unreliable" which means you need backing plant to deal with extended downtimes or lose ~35% to storage systems (batteries).
If you can run the backing (aka peaking) plant from a nuclear source then you no longer need the Windmills / solar PV / tidal installations to keep your carbon emissions down and they revert to simply being a pain in the ass as far as power grid distribution designs go.
Even now, the economics of the things is shonky as all hell (if they were a viable source, power companies would be installing them all over and not sucking on subsidies to do so) and a good chunk of the seriously nasty pollution in china is a result of making solar PV systems.
I know I'm a LFTR cheerleader, but they have lots of upsides and few downs.
One of the major upsides over conventional nukes is that their power output can be changed rapidly and repeatedly without being crippled by the xenon poisoning that hits light water plants when you turn the output power down and results in an inability to turn LWRs back up for several hours. This is the critical factor which means that conventional "peaking plant" can be dispensed with.
LFTRs _will_ happen as large civil power plants within 20 years. Test ones will happen before the end of this decade (I'm expecting to see small testbeds the size of the USA's 1960s ones before the end of 2016)
They'll undoubtedly be chinese technology and that's a direct result of the USA cancelling all research 40 years ago because the test reactors were too good at making heat for driving turbines and effectively useless at producing weaponisable materials.
The chinese are pouring hundreds of millions of dollars into LFTR research, whilst USA and EU groups are left begging for enough money (a few hundred thousand dollars) to run a few computer simulations to validate their models. It would make more sense for the western researchers to go and contract for China.
Why is this happening? Because china's leadership are mostly drawn from engineering backgrounds and when they see a need for something, they make sure it happens. The pollution crisis has underscored that need, but China announced it was planning to eliminate coal plants back in 2010 so this shouldn't be a surprise to anyone.
"Uranium has to be mined (most likely using similar circumstances) as well. Most everything that we use and dispose of has an environmental impact."
Thorium is currently a noxious byproduct of rare earth mining and the biggest dry toxic component of the tailings by volume. Make it valuable and 90% of the problem would be cleaned up overnight.
The day thorium becomes a product that's in demand is the day that Moly Corp and others operate Thorium mines with rare earths as a sideline.
If you really want to reduce environmental damage from rare earth mining then you need to find a use for ALL the products. Thorium is a particular problem for RE mines as it outweighs the other products combined by between 5:1 and 10:1 depending on the mine.
Bear in mind that the sidestream thorium output from Moly Corp is about 5000 tons/year(*) and that's about enough to feed the USA's _current_ total annual electrical requirements if used in LFTRs. Now factor in the increase in demands to replace carbon-based energy sources such as fuel-oil/gas heating systems and most short-range transportation (even better would be to put the thorium systems close enough to inhabited areas to pipe waste heat around as district heating but I can't see people going for that in big numbers - having said that I'd happily live downwind of a LFTR)
(*) About 100kg of "waste" per year from using that in reactors, vs 4993 tons of waste for comparable volumes of raw uranium yellowcake.
"I'm sure the Google closed factory has lots of other reasons"
The primary reason: Even without cheaper labour(*), if you're mass-producing it's a hell of a lot cheaper to build your devices close to the component supply sources and ship finished product around the world than it is to do it the other way around - it also means you're not subject to weeks of lead time in development cycles.
(*) Chinese labour costs are a lot higher than a pure $/hour rate might indicate. Turnover on employees is ridiculously high, so there's a huge burden on training them and all the large manufacturers provide dormitory/canteen facilities, etc, because most workers in the coastal zones come from the interior of the country (many simply don't bother returning to work after the chinese new year holiday, if they have saved enough or are burned out)
Western economic thinking is that an employee costs around double their hourly wage when all ancilliaries are factored in. In china it's more like 4-6 times the wage.
As a result, chinese manufacturers are moving factories inland (practical now they have high speed rail transport networks which didn't exist 15 years ago along with much better power infrastructure) and has the highest rate of automation installation in the world (Foxconn has more than a million robots installed already).
Interestingly this has resulted in chinese workers starting to take industrial action because they feel they're not getting enough work (remember, most come from inland, They're at the factory to work hard, save money and buy some land back home, so they want to do this as quickly as possible).
This parallels the west and japan - where robots doing the dirty dangerous stuff were initially welcomed (most notably in carmaking plants for welding and painting) but moving into the "comfortable" areas resulted in strikes.
FWIW, a pick'n'place machine can easily replace 20 people on a line and turns out consistent work with the lights out, 24*7*52. With appropriate programming they can be as flexible as people too in terms of handling what comes down the line. (although a PnP programmer costs more than 1 or 2 solder jockeys). They're pretty much essential for SMD work, but people are still better (for the moment) when it comes to assembling boards into the finished product.
The chinese are wising up to the facts of environmental destruction - they've pretty much had awareness forced on them by the smogs of the last decade - and unlike the USA, which is corporatist, the chinese govt _does_ answer to the will of the people in the end.
On the bright side, those tailing ponds are likely to be enormous sources of thorium and other "nasties" which are actually beneficial resources when you have a use for them. Let's just hope that it's not leaking.
"I would assume that's because now their economy will be much better off because of the lifted sanctions, and they know it."
Lots of iranians I know in europe are dancing in the streets too - and it's mainly because they believe that with the lifted sanctions the religious govt will cease to exist in short order and they can go home without fear of persecution.
Maybe, just _maybe_, Iran will have its first really democratically elected govt since the 1950s before the end of this decade.
The difference between NK and Iran is that NK was a puppet state of the Soviet Union, setup by Stalin and inheriting its cult of personality from him. To this day NK still gets the vast majority of its support from Russia with only grudging support from China - to the point where the chinese cut off oil and electricity supplies into NK for several months last year and only resumed them when it became clear that the general population was suffering wildly as the leadership grabbed everything they could for themselves. (China would prefer that NK didn't exist, but until relatively recently they've been worried about having a rich Korea just over the river.)
Iran on the other hand may well be run by a bunch of "religious nutters" but that's primarily as a bounce against the extreme actions of the puppet dictator the USA installed and cemented in place by the sanctions imposed on Iran as a result of kicking him out. The reality is that the vast majority of the country is well educated and sick of the control freaks. What remains of the religious police are having to walk around in groups lest they get beaten up - and it happens regularly to them even with that precaution. Take away the bogeymen that the iranian establishment relies on and you take away that establishment's power (and on that note, the best response to posturing from these kinds of countries is "Yeah right, whatever". Anything else feeds the troll). The vast majority of the army are conscripts and they'd sooner drop their guns and walk if a real war broke out than risk getting killed (Many of them actually did so during the iran/iraq war).
Nothing's stopping Iran redirecting its nuclear research as long as it's not capable of being weaponised. That means there are a lot of researchers now sitting on their thumbs whose knowledge could be used to advance LFTR-style technology dramatically. I wonder if they're willing to make a deal with the chinese research labs already exploring this path.
Whose own statements have been repeatedly undermined by leaked MOSSAD reports stating that Iran has neither the capabilities or wish to build nuclear weapons. - and those same reports pointed out that had Iran wished to build any, it could have many years ago as they've had more than enough highly enriched uranium onhand for a couple of decades.
Iran needs to be brought in from the cold. They have good reason for distrusting the west, even before a certain coup against a democratically elected government back in the 1950s
The prime use of them being out there has been as a bogeyman the West could point to to justify its military activities in the middle east - and as it turned out, the most destabilising influence in the Middle East since the end of the Cold War has been the West's military activities (Apart from the puppet govts such as Saddam Hussain, the Syrian clusterfuck is a direct result of french bureaucracy and a deliberate plan to keep the area weak by dividing tribes up into different countries, post WW1, against the advice of T.E.Lawrence - who drew up plans which would have kept things stable and secure)
And of course the fact that Iran's been out in the cold means the establishment there could use the West as a bogeyman to keep their own people subjugated. Noone won anything except the arms suppliers.
If a Vista or XP machine shows up on _my_ network it's going to find it can't talk to much (yes, they can be identified and yes they can be diverted into quarantine zones. Gotta love Packetfence)
Once the hardware's "good enough" to run the software you want, the drive to keep "upgrading" goes away.
Phones and tablets have a limited physical lifespan, simply because of the abuse they take. It gets cheaper to replace than repair very quickly, especially when landfill android is cheap and has better performance than the flagships of 2-3 years ago.
At some point (and I suspect that it will be much sooner than it took in the desktop development cycle), people will be updating their devices because they're broken, not for software or style reasons.
"blind firing of people for using Linux is almost certainly not a good practice"
It's cropping up more and more in companies which have been handed their asses for violating the GPL(*), instead of coming to terms with the fact that complying is quite easy.
Ther windows kernel is fairly solid. As you say it's all the cruft on top which is an issue, but the average dev or enduser can't strip all that cruft away.
The problem is that you're comparing kernels to OSes and they are different things.
The OS is everything that wraps around the kernel. Scarily enough there's a Debian port with a BSD kernel. Anything is doble if you want it.
"you get your connectivity from Pedro, and something goes wrong."
That's between you and Pedro, not you and AT&T. Let Pedro duke it out with them. He's not going to let them screw him over or you'll take your business down the road to Shnycorp.
The reality is that Pedro provides substantially better service for a couple of bucks more than Shinycorp
Shinycorp do big nationwide advertising and low prices, by cutting deals with AT&T for lower quality of service and longer times to get onsite than AT&T sells as standard.
The result is that when your line goes down with Shinycorp, you argue for hours with their phone droids and go through the entire modem waggling routine before they admit there might be a problem and someone will be out next week to fix it.
When you ring Pedro, you speak to Pedro's tech, who looks at your circuit and goes "I'm on it. There'll be someone onsite by the end of the day"
Shinycorp competes with Bastardcorp, Screwupcorp and Cheatemcorp. They all have low prices with various gotchas and they all offer the same poor level of service. They have the highest rate of customer complaints in the industry but everyone goes with them because they all put saturation advertising in the media, so joe average has never heard of the others. They make millions each day and everyone hates them but doesn't look elsewhere.
Pedro makes a comfortable income. You're happy with your level of service. He lives or dies on this, so it's in his interest to keep his customers happy.
Fantasy? No, that's what happens in places with LLU and fairly closely describes the way I buy my connection
"As an aside, why the fuck are there drum brakes any more?"
For some applications, they work better. (and they're cheaper)
Why do you need to borrow a scantool when you can buy a bluetooth ODB one for $20-$60 (or $5 for the fake ELMs, which still work fine) ?
"My point was that points needed to be constantly adjusted."
Yup. I put a homemadeTAI system in my mother's car (A 1974 Datsun B210) back in the 80s and it went from needing point adjustments every 5000 miles to almost never (about 20k miles due to the rubbing block wearing out)
When I changed the points over to an optocoupler assembly the mechanic was horrified, but having eliminated the rubbing block the timing stayed solid for 60,000 miles.
As a nice side effect the engine could whiz through to 9000rpm if you weren't careful (6000rpm redline) but as we'd balanced everything this wasn't too much of an issue
" Just degree the pickup."
Uh yeah right.
The pickup runs off a notched tooth ring on the flywheel plus another on the camshaft if you're lucky.
each plug has its own coil pack. Wasted spark is common (partly because it ensures unburned hydrocarbons don't go down the exhaust)
It's been a long time since cars had distributors.
"Sadly, even most shops now are useless as they just plug the computer in and do what it tells them"
Yup. A GF spent about $3000 getting her mercedes C220 fixed because of this - and it made no difference whatsoever.
The actual fault was a jammed open thermostat. Replacing that ( $50 because it comes in a special housing) solved every other problem.
Because the engine remained stone cold when it should have warmed up the computers got confused and reported alarm codes out the wazoo. A competent mechanic would have looked for a common cause. Competent mechanics at dealerships are a rarity.
"The problem is that people like you who want to work on their car are becoming more and more rare"
I'll phrase it slightly differently: I want _my_ choice of mechanic to be able/allowed to work on it, not the stealership.
"But yeah, you can use a paper clip."
You can use a paper clip on (E)ODB cars too, or with some there's a "secret" mode for getting the readout (My nissan has something like keyoff/keyon/keyoff/keyon/ 5 *fullthrottle/off then hold the pedal down for 10 seconds)
It's generally easier to just use an ODB dongle though.
"I prefer to buy an import from a country where the engineers can count and know what metric is, like Germany or Japan. Every fastener will be metric, except spark plugs of course."
And the bolts holding the seatbelt anchors. For some reason even in europe and japan these are SAE
Presumably this is to prevent someone kludging something unsuitable in.
"Thus it must be branded as a waste product, and disposing of a radioactive waste product is insanely expensive if it is possible at all."
Which is silly, because thorium isn't particularly radioactive, even when turned into ingots.
Seriously. The stuff in the tailing ponds has a 14 billion year halflife - all the other isotopes are long-gone.
The main problem (as with many of the transuranics) is heavy metal poisoning if it gets into the biosphere.
"However, I've never seen a single reference to a study of the effects of windmills on regional wind patterns, massive areas of solar panels on regional temperature/wind/etc., let alone manufacturing of these things. Are they issues?"
In a nutshell: not really. The effects on wind pattern and temperature is negligable, even when compared to the downwind effects we already generate from cities.
In a larger nutshell: With appropriate nuclear technology, these things become totally superfluous. Solar and wind sources are notoriously "unreliable" which means you need backing plant to deal with extended downtimes or lose ~35% to storage systems (batteries).
If you can run the backing (aka peaking) plant from a nuclear source then you no longer need the Windmills / solar PV / tidal installations to keep your carbon emissions down and they revert to simply being a pain in the ass as far as power grid distribution designs go.
Even now, the economics of the things is shonky as all hell (if they were a viable source, power companies would be installing them all over and not sucking on subsidies to do so) and a good chunk of the seriously nasty pollution in china is a result of making solar PV systems.
I know I'm a LFTR cheerleader, but they have lots of upsides and few downs.
One of the major upsides over conventional nukes is that their power output can be changed rapidly and repeatedly without being crippled by the xenon poisoning that hits light water plants when you turn the output power down and results in an inability to turn LWRs back up for several hours. This is the critical factor which means that conventional "peaking plant" can be dispensed with.
LFTRs _will_ happen as large civil power plants within 20 years. Test ones will happen before the end of this decade (I'm expecting to see small testbeds the size of the USA's 1960s ones before the end of 2016)
They'll undoubtedly be chinese technology and that's a direct result of the USA cancelling all research 40 years ago because the test reactors were too good at making heat for driving turbines and effectively useless at producing weaponisable materials.
The chinese are pouring hundreds of millions of dollars into LFTR research, whilst USA and EU groups are left begging for enough money (a few hundred thousand dollars) to run a few computer simulations to validate their models. It would make more sense for the western researchers to go and contract for China.
Why is this happening? Because china's leadership are mostly drawn from engineering backgrounds and when they see a need for something, they make sure it happens. The pollution crisis has underscored that need, but China announced it was planning to eliminate coal plants back in 2010 so this shouldn't be a surprise to anyone.
"Atoms for Peace" - from china. Yes really.
"Uranium has to be mined (most likely using similar circumstances) as well. Most everything that we use and dispose of has an environmental impact."
Thorium is currently a noxious byproduct of rare earth mining and the biggest dry toxic component of the tailings by volume. Make it valuable and 90% of the problem would be cleaned up overnight.
The day thorium becomes a product that's in demand is the day that Moly Corp and others operate Thorium mines with rare earths as a sideline.
If you really want to reduce environmental damage from rare earth mining then you need to find a use for ALL the products. Thorium is a particular problem for RE mines as it outweighs the other products combined by between 5:1 and 10:1 depending on the mine.
Bear in mind that the sidestream thorium output from Moly Corp is about 5000 tons/year(*) and that's about enough to feed the USA's _current_ total annual electrical requirements if used in LFTRs. Now factor in the increase in demands to replace carbon-based energy sources such as fuel-oil/gas heating systems and most short-range transportation (even better would be to put the thorium systems close enough to inhabited areas to pipe waste heat around as district heating but I can't see people going for that in big numbers - having said that I'd happily live downwind of a LFTR)
(*) About 100kg of "waste" per year from using that in reactors, vs 4993 tons of waste for comparable volumes of raw uranium yellowcake.
"I'm sure the Google closed factory has lots of other reasons"
The primary reason: Even without cheaper labour(*), if you're mass-producing it's a hell of a lot cheaper to build your devices close to the component supply sources and ship finished product around the world than it is to do it the other way around - it also means you're not subject to weeks of lead time in development cycles.
(*) Chinese labour costs are a lot higher than a pure $/hour rate might indicate. Turnover on employees is ridiculously high, so there's a huge burden on training them and all the large manufacturers provide dormitory/canteen facilities, etc, because most workers in the coastal zones come from the interior of the country (many simply don't bother returning to work after the chinese new year holiday, if they have saved enough or are burned out)
Western economic thinking is that an employee costs around double their hourly wage when all ancilliaries are factored in. In china it's more like 4-6 times the wage.
As a result, chinese manufacturers are moving factories inland (practical now they have high speed rail transport networks which didn't exist 15 years ago along with much better power infrastructure) and has the highest rate of automation installation in the world (Foxconn has more than a million robots installed already).
Interestingly this has resulted in chinese workers starting to take industrial action because they feel they're not getting enough work (remember, most come from inland, They're at the factory to work hard, save money and buy some land back home, so they want to do this as quickly as possible).
This parallels the west and japan - where robots doing the dirty dangerous stuff were initially welcomed (most notably in carmaking plants for welding and painting) but moving into the "comfortable" areas resulted in strikes.
FWIW, a pick'n'place machine can easily replace 20 people on a line and turns out consistent work with the lights out, 24*7*52. With appropriate programming they can be as flexible as people too in terms of handling what comes down the line. (although a PnP programmer costs more than 1 or 2 solder jockeys). They're pretty much essential for SMD work, but people are still better (for the moment) when it comes to assembling boards into the finished product.
This is changing, rapidly.
The chinese are wising up to the facts of environmental destruction - they've pretty much had awareness forced on them by the smogs of the last decade - and unlike the USA, which is corporatist, the chinese govt _does_ answer to the will of the people in the end.
On the bright side, those tailing ponds are likely to be enormous sources of thorium and other "nasties" which are actually beneficial resources when you have a use for them. Let's just hope that it's not leaking.
"I would assume that's because now their economy will be much better off because of the lifted sanctions, and they know it."
Lots of iranians I know in europe are dancing in the streets too - and it's mainly because they believe that with the lifted sanctions the religious govt will cease to exist in short order and they can go home without fear of persecution.
Maybe, just _maybe_, Iran will have its first really democratically elected govt since the 1950s before the end of this decade.
The difference between NK and Iran is that NK was a puppet state of the Soviet Union, setup by Stalin and inheriting its cult of personality from him. To this day NK still gets the vast majority of its support from Russia with only grudging support from China - to the point where the chinese cut off oil and electricity supplies into NK for several months last year and only resumed them when it became clear that the general population was suffering wildly as the leadership grabbed everything they could for themselves. (China would prefer that NK didn't exist, but until relatively recently they've been worried about having a rich Korea just over the river.)
Iran on the other hand may well be run by a bunch of "religious nutters" but that's primarily as a bounce against the extreme actions of the puppet dictator the USA installed and cemented in place by the sanctions imposed on Iran as a result of kicking him out. The reality is that the vast majority of the country is well educated and sick of the control freaks. What remains of the religious police are having to walk around in groups lest they get beaten up - and it happens regularly to them even with that precaution. Take away the bogeymen that the iranian establishment relies on and you take away that establishment's power (and on that note, the best response to posturing from these kinds of countries is "Yeah right, whatever". Anything else feeds the troll). The vast majority of the army are conscripts and they'd sooner drop their guns and walk if a real war broke out than risk getting killed (Many of them actually did so during the iran/iraq war).
Nothing's stopping Iran redirecting its nuclear research as long as it's not capable of being weaponised. That means there are a lot of researchers now sitting on their thumbs whose knowledge could be used to advance LFTR-style technology dramatically. I wonder if they're willing to make a deal with the chinese research labs already exploring this path.
"according to Netanyahu"
Whose own statements have been repeatedly undermined by leaked MOSSAD reports stating that Iran has neither the capabilities or wish to build nuclear weapons. - and those same reports pointed out that had Iran wished to build any, it could have many years ago as they've had more than enough highly enriched uranium onhand for a couple of decades.
Iran needs to be brought in from the cold. They have good reason for distrusting the west, even before a certain coup against a democratically elected government back in the 1950s
The prime use of them being out there has been as a bogeyman the West could point to to justify its military activities in the middle east - and as it turned out, the most destabilising influence in the Middle East since the end of the Cold War has been the West's military activities (Apart from the puppet govts such as Saddam Hussain, the Syrian clusterfuck is a direct result of french bureaucracy and a deliberate plan to keep the area weak by dividing tribes up into different countries, post WW1, against the advice of T.E.Lawrence - who drew up plans which would have kept things stable and secure)
And of course the fact that Iran's been out in the cold means the establishment there could use the West as a bogeyman to keep their own people subjugated. Noone won anything except the arms suppliers.
"milking a dying cow is also not a profitable business model long term"
Someone needs to explain that to Cisco.
If a Vista or XP machine shows up on _my_ network it's going to find it can't talk to much (yes, they can be identified and yes they can be diverted into quarantine zones. Gotta love Packetfence)
Once the hardware's "good enough" to run the software you want, the drive to keep "upgrading" goes away.
Phones and tablets have a limited physical lifespan, simply because of the abuse they take. It gets cheaper to replace than repair very quickly, especially when landfill android is cheap and has better performance than the flagships of 2-3 years ago.
At some point (and I suspect that it will be much sooner than it took in the desktop development cycle), people will be updating their devices because they're broken, not for software or style reasons.
"The concept of making money by selling an operating system is a 1990's idea."
MS has never made much money out of Windows. It's mostly a means to an end.
The _real_ moneyspinner has always been the office family.
"blind firing of people for using Linux is almost certainly not a good practice"
It's cropping up more and more in companies which have been handed their asses for violating the GPL(*), instead of coming to terms with the fact that complying is quite easy.
(*) Or run by people who have been.
Ther windows kernel is fairly solid. As you say it's all the cruft on top which is an issue, but the average dev or enduser can't strip all that cruft away.
The problem is that you're comparing kernels to OSes and they are different things.
The OS is everything that wraps around the kernel. Scarily enough there's a Debian port with a BSD kernel. Anything is doble if you want it.
In a contemporary hijacking scenario, passengers won't sit still for hijackers anymore.
This is not conjecture. Passengers have taken matters into their own hands on more than one occasion.