It was in a different post, but I did point out that looking at 1-2Q numbers was misleading. The year-to-year numbers are still in favor of Android, but not to the same extent. I was just pointing out that lots of people do in fact like the S3. It isn't bad, but I'm holding out to see what the Nexus offerings are like this year...
My employer used to offer 1x salary for free, up to 6x for a very modest premium, and in addition to that they also offered a spouse 50% of income for the rest of their life or until they remarried.
The latter has gone away in favor of a one-time distribution - I think now it goes up to 8x salary, but again only 1x is free.
Well, I can't vouch for how much of this happens with adults, but it certainly applies to teenagers. I remember having a discussion with my stepdaughter during her early teen years in which she would occasionally refer to creepy people, and what I managed to put together was that creepy meant "not cute/hot/etc." I'd like to think that to some degree she's grown past that - certainly I've done what I've can to try to educate her that the disarming ones are the ones you have to watch out for, but life has probably taught her that much more effectively than I could.
I wouldn't be surprised if a fair bit of the college date rape rate is made up of those who don't learn at an early age.
They're only best-selling when you look at them against individual vendors. The problem is that they've tended to take an Apple-vs-the-world approach to everything and that means that you have to add up the other vendors.
While the term "command line" or "WIMP" may not be perfect analogies, the fact is that Apple ties their hardware to software, and if you can find a better way to differentiate the Apple II from a Mac feel free to suggest it. On simple terminal-like interfaces they lost out to DOS, and on the windowed operating systems they lost out to Win95. Now iOS is losing badly to Android.
Sure, they might sell more iPhones than HTC sells One X's or whatever it is they're making this month, but that is because you can only run iOS on about two products sold in any given year, and you can run Android on about 2000. Those 2000 might sell individually fewer than Apple, but from an OS standpoint they greatly outsell Apple. The 68% thing is probably misleading since that is based on two quarters that Apple undersells in, but even looking at year-round sales Android has majority share.
Sure, Apple can continue to make lots of money with only 10% market share, as they already do in the laptop business. However, that doesn't change the fact that they don't drive the industry the way they did when they first started out. This is the point of that slide.
My only reason for not gravitating towards the current Nexus: There's no removable SD storage available. Not a major issue, just a preference.
Yes, but your choices are basically lousy upgrade support, or no removable SD storage. The Nexus phones are decent, but generally only superior in the sense that they don't suck as much. I didn't get one last time because it was inferior in hardware to the phone I did get, but it has been a battle getting the later OS versions to work since nobody releases source for hardware drivers (INCLUDING the Nexus phones).
Carrier subsidies are often a problem as well - you can get a Nexus device that will work on AT&T right now, but you'll pay hundreds of dollars for it and you won't get any discount for AT&T for not getting a phone from them. Your only real option is to get a discounted phone from AT&T, eBay it, and then use some of that to defray the cost. Talk about a crazy market.
I'll probably get a Nexus phone next time, but it would be nice to have a better selection of phones that actually get support. Hopefully the rumors are true and there will be more than one Nexus device next time around...
In the US a typical phone costs $199 with a commitment to an overpriced data plan - usually something like $30-40/mo.
Retail values of the phones without that commitment (assuming you can get them) tend to be more like $600 or so, for the nicer phones.
T-Mobile is about the only major carrier I'm aware of that will let you divorce the one from the other, and then you can get a data plan for $5-10/mo (the latter with 2GB of 4G data, and unlimited 2G data). Phone subsidies and exclusivity deals are really something that needs to go away.
Ugh - there is room in the market for phones that AREN'T cookie-cutter copies of each other. What if I want a smartphone with a small screen - nobody would call that a flagship phone so nobody would make that a single-product focus. How about a phone with a keyboard - most people don't want that, so nobody would make that their single product.
The whole point of Android is that you actually get a choice. I don't want that to be a choice of 3 vendors who all make phones designed to look just like an iPhone...
I tend to agree. However, UI optimization has to be the key.
My wife is still using a 4 year old feature phone because the newer models are all inferior. Tasks she does often are buried in the menus (things like pull up contacts, send/read SMS, etc.). Stuff she doesn't care about is front-and-center, like browse the web, look for ringtones, and all that.
It seems like the newer devices are just designed to get people who don't have data plans to accidentally pull a kilobyte of data here or there so they can be charged through the noses.
Sure, feature phones are becoming a niche, but if you're going to make one at least make it user-centric, and not carrier-centric.
This doesn't really counter the argument against vendor lock-in. Also the vast majority of ARM devices are actually NOT bootloader locked. Just go and check the compatibility list for Cyanogenmod.
Uh, virtually all the devices on the Cyanogenmod compatibility list HAVE locked bootloaders. They just have poorly implemented locked bootloaders/etc, which means that you can defeat them and install Cyanogenmod anyway. The only Android device I'm aware of that have unlocked bootloaders are the Nexus line, which probably have
Anyway the main point of my post was that there's a significant number of people who are interested in doing with their devices whatever the hell they want, and this even includes installing Android on an iPhone. People want to do this and given it's their hardware we should not be promoting systems to prevent this.
I'm all for legislation that requires device owners to be given all the stuff required by GPLv3/etc. However, I've got nothing against the technology itself. I'd prefer a device with UEFI to what I have on my PC now. Right now anybody who wants to can just install a keylogger on my hard drive just by booting from a CD, and there is little I could do to stop it. Rootkits installed over the internet can make themselves almost undetectable unless you boot off of clean media. UEFI has the power to change that. My only objection to the proposed implementation is the inability to change the OEM-supplied machine key (that isn't the one used to sign windows) - I want to be able to change it all.
Where did I say that this article had anything to do with detectors? I said that people were spending lots of money on THz detectors, which is true. The argument was that working with THz radiation wasn't anything unusual, and my point was that the reason people were spending all kinds of money on the technology were doing so because it was.
Honestly, anyone loading Linux themselves should be more than capable of going into the UEFI setup and flipping it to off
What a crap argument. So you say that currently Linux isn't for Joe Sixpack and therefore it is OK to keep it that way? No. We want the Linux to be available for everyone!
Personally, I'd rather my bootloader be secure against malware than for Linux to be available for everyone. The former is quite achievable with UEFI, and the latter is nebulous at best. Plus attempts to do the latter have steadily tended to make Linux less usable for me. It isn't like I'm the one getting paid for every PC that Ubuntu or whatever gets installed on - my only possible benefit is if some of those users contribute back, and I doubt I'll get much of that from folks who can't reconfigure their firmware.
That pretty much guarantees some percentage of users will brick their systems.
Uh, how exactly is changing a firmware setting going to brick a system? You don't need to reflash the firmware to change the boot key, and per the spec the firmware has to have a factory reset function that restores all settings including keys to the shipped settings.
The sort of user who couldn't turn off secure boot in their firmware probably couldn't reinstall Windows if they got frustrated with Unity or whatever anyway...:) If you want to target these users you need to ship the PCs pre-configured.
Most big software projects I've seen fail hard, like millions and 10's of wasted dollars hard. By comparison you just dont see that very often in big electrical/mechanical/civil projects, which can be equally complex (eg refineries, cruise liners etc).
It really depends on the engineering project. When you're talking about designing a cruise liner that has 50 more rooms than the last one, that might cost millions of dollars, but the complexity is more analogous to a software point release. Building the thing is like running make. Neither of these two operations tend to have massive cost overruns in the software world.
Engineering projects have a greater tendency to be about making incremental improvements on work that has already been done, if you're even making improvements at all. The number of interacting components tends to be much smaller as well compared to software. The typical software project tends to involve things that nobody on the team has ever done in quite that way before. When you do get engineering projects like building a new manned spacecraft or fighter jet or aircraft carrier they tend to be loaded with overruns.
Engineering also involves human beings in the construction, vs software which uses computer programs to do all the compilation. If your ship contains a wire that runs along and then just stops and doesn't connect to anything, chances are that somebody is going to report that when they're actually running the wire. It is also unlikely that the ship just dynamically "allocates" wires and other interacting parts while it is at sea.
I don't think competence is really the fundamental problem so much as the problem is just that much more difficult.
Software engineering is still playing catch up, in the sense that most developers and development companies I've seen still dont follow a formal enough process for it really to be called engineering. Usually it's a bunch of computer science graduates having a wild stab at it, and the good ones are closer to artisans than engineers...Until the entire software industry gets off it's high horse and admits this to itself - and more importantly admits this to the customers, we are going to continue to be dissapointed with the quality.
People only state that they're disappointed in software quality in a vacuum. When their free smartphone app crashes they say, "man, we should have better quality software." They don't say, "man, we should have better quality software - I'd have gladly paid $100 for what was a free app." Quality costs money, and completely formalizing all aspects of software development the way they formalize space shuttle design would cost a LOT of money (and as I recall you're still much more likely to be killed by a space shuttle ride than driving a car). At work somebody was griping about having to fix a cosmetic bug from the last round of development, just after complaining about the fact that we can't make all their desired changes due to budget. Well, if the team were to increase quality to cut down on even cosmetic bugs then FAR fewer changes would be made.
Software needs to be good enough. If my rain gutters had a slow drip somewhere along the line, chances are I wouldn't spend hundreds of dollars to have it repaired.
Are you suggesting that if I sell somebody a phone it shouldn't be able to boot unless they insert an install SD card or such? Or does this just pertain to PCs? Most people buy PCs with pre-installed OSes. Is there really any value to making it so that those PCs can't be booted without sticking in a CD as the first step?
And if that happened, how would that help? Anybody with a Windows PC will have stuck in the Windows CD, which will install the MS key and now it won't boot linux when they want to switch.
If the firmware has to accept any key it finds on any later boot, well then secure boot gets you nothing anyway since the malware can just supply a new key.
While I'm all for getting rid of lock-in, the fact is that almost all arm-based system in consumer use have locked bootloaders already. Just about every android phone in use falls into this category (and yes, I know the sliver of market share held by Nexus devices are an exception).
I guess you meant fool-proof. And it is. It is fool-proof against all those fools who want to decide to run their own code on the computer without having to ask permission beforehand.
You don't have to ask for permission - you just have to configure your computer to boot it. If you stick an Ubuntu CD in a PC that isn't configured to boot off of CD, it won't run that either unless you "ask for permission" by telling it to boot from CD.
On amd64 at least you'll be able to disable it if you want, or configure it with your own keys, so that MS won't be able to install something on your PC without asking YOU for permission.
Software doesn't have to be perfect. It simply needs to be economical due to its qualities (again, qualities well known for decades.)
This last bit needs to be emphasized - perfect software that you can't afford is useless. As is often attributed to Stalin, sometimes quantity has a quality all its own. If some other start-up's hacked together social media site that isn't scalable lets them get to 10k users before yours is written, then the network effect will prevent you from ever taking advantage of the fact that yours can handle a billion, and they'll have plenty of money to spend doing a complete rewrite.
Usually the way these sorts of things get looked at in the real world involves some kind of FMEA approach (failure modes effect analysis). You look at what can go wrong, how often it will go wrong, what happens when it goes wrong, and whether any of that can be mitigated and if so how likely it is that the problem will be spotted and mitigated.
There are lots of ways to do it, but the bottom line is that not all problems are created equal, as you've stated. If I've got some system that "just works" placing that stability at risk to fix some UI glitch that isn't causing problems is a dumb decision, especially if it costs a lot of money.
I think a better question would be "Why does it NEED to be better?" and the answer in most cases is "it doesn't". Remember perfect is the enemy of good and often the difference between "good enough" and great can be truly insane levels of expense.
Yup, just the other day at work we had somebody complain about a cosmetic regression in a piece of highly specialized software. They were basically on a crusade for quality, and pushing hard to have it fixed. I basically went to the project managers and pointed to the pile of much more critical needs and fixed budget, and it is very unlikely we'll ever get to the cosmetic issue as a result.
If this were something displayed on the company homepage I'd look at it differently, but for an internal piece of software it just isn't worth the trouble. Some might argue that a good quality system would not result in such regressions in the first place, but again I have to point out to the pile of high-importance needs and fixed budget. If I have to choose between a system improvement that has documented ROI and creates a business opportunity, and spending money so that the stuff we deliver is less likely to have very minor bugs, I'll pick the feature. If we were talking about regressions with a significant functional impact then that would be a different matter.
Scope*Quality, Cost, Time - pick two. Pick the wrong balance and your company might go out of business...
Well, the appropriate solution varies considerably by industry.
In the case of telecom I'd probably treat them as has been done with electricity - get rid of vertical integration. regulate the living daylights out of the last mile and anything that uses spectrum as natural monopolies, and then make the rest free market.
So, if you have phone service or DSL you'd pay a company $8/month to run a twisted pair from a CO to your home and keep it running to FCC standards. Then you'd get to buy telephone/internet/etc service from any of a multitude of companies who pay the telco a standard rate per 1U of rackspace and per termination.
If you have a cell phone then it would connect to a tower operated by a tower owner. Towers would use standard protocols on standard frequencies under FCC regulation. Towers would charge by the packet, but would not bill customers directly. Within any area no company could own more than a third of the spectrum for any piece of land, so that there would always be competition between redundant operators for any area. Tower operators could set their rates, but would have to charge the same rate per packet to all of their customers, who would be national carriers or wholesalers. The national carriers would be forbidden from owning towers - they could buy coverage from operators in whatever manner makes sense, and put together plans in whatever way makes sense to them, and would sell service to the customer. The national carriers would not be regulated.
The goal would be to separate the physical layer that is expensive to maintain and tends to have high barriers to entry from the content layer which has much lower barriers to entry. This is similar to how many areas are breaking up electrical generation from transmission/distribution/etc.
Sure, it isn't perfect, but it cuts down on the number of big pockets who can bribe regulators, and limits the scope of regulation further than it is now while making the regulation of those areas more effective.
Sure, but in China you only get punished for accepting bribes if you embarrass the party.
Bribery is rampant in China. I hear they're trying to reduce it, but it is a big culture issue.
Take the recent cases of melamine in infants formula that led to a few executed regulatory officials. While officially they were executed for accepting bribes, what they were really punished for is letting botched formula through. The correct way to accept bribes in China is to accept them only to approve things that you should have approved anyway, not to approve things that shouldn't have been approved.
If you want a car analogy in the USA, imagine your kid takes a driver's license exam and does well. The officer administering the test tells them that they do well, but they need to process the paperwork before passing them and that could take a long time. The kid hands the officer $100, and the officer says hey, I can do the paperwork right now while you wait and they check three boxes and sign a form and hand it back to them and they can drive. Or they can not hand them the $100 and check back in a few weeks and find out it just isn't getting done. The officer isn't passing somebody who should fail - they're just adding a "tax" of sorts to the process.
By Western standards it is horribly corrupt. My intent is just to illustrate how the culture works. I remember reading in C&EN about 15 years ago that it was routine for huge companies to pay bribes to offload supertankers full of chemicals and such. Everything they were doing (aside from the bribes) was legal and would be allowed in any country, but if they didn't pay the bribes their expensive tanker would just rot in port for a few weeks while the paperwork was sorted out. I knew a guy who once was involved in setting up a plant in South Korea and they had a budget item for bribes - often for things like importing components made in Japan (this was many years ago, and the purpose of the law/bribes was intended more as an insult to Japan than anything). Business culture in Asia is often different, though many of these examples were from a long time ago.
At that frequency what would be the range of such a radar, except in more exotic conditions like transmitters at very high elevation or in deserts? I would think that atmospheric absorption would be a big problem.
Maybe for air-to-air radar it might be practical - once the transmitter is at altitude the absorption of moisture would be less of an issue.
Slashdot Mirror
It was in a different post, but I did point out that looking at 1-2Q numbers was misleading. The year-to-year numbers are still in favor of Android, but not to the same extent. I was just pointing out that lots of people do in fact like the S3. It isn't bad, but I'm holding out to see what the Nexus offerings are like this year...
My employer used to offer 1x salary for free, up to 6x for a very modest premium, and in addition to that they also offered a spouse 50% of income for the rest of their life or until they remarried.
The latter has gone away in favor of a one-time distribution - I think now it goes up to 8x salary, but again only 1x is free.
Well, I can't vouch for how much of this happens with adults, but it certainly applies to teenagers. I remember having a discussion with my stepdaughter during her early teen years in which she would occasionally refer to creepy people, and what I managed to put together was that creepy meant "not cute/hot/etc." I'd like to think that to some degree she's grown past that - certainly I've done what I've can to try to educate her that the disarming ones are the ones you have to watch out for, but life has probably taught her that much more effectively than I could.
I wouldn't be surprised if a fair bit of the college date rape rate is made up of those who don't learn at an early age.
They're only best-selling when you look at them against individual vendors. The problem is that they've tended to take an Apple-vs-the-world approach to everything and that means that you have to add up the other vendors.
While the term "command line" or "WIMP" may not be perfect analogies, the fact is that Apple ties their hardware to software, and if you can find a better way to differentiate the Apple II from a Mac feel free to suggest it. On simple terminal-like interfaces they lost out to DOS, and on the windowed operating systems they lost out to Win95. Now iOS is losing badly to Android.
Sure, they might sell more iPhones than HTC sells One X's or whatever it is they're making this month, but that is because you can only run iOS on about two products sold in any given year, and you can run Android on about 2000. Those 2000 might sell individually fewer than Apple, but from an OS standpoint they greatly outsell Apple. The 68% thing is probably misleading since that is based on two quarters that Apple undersells in, but even looking at year-round sales Android has majority share.
Sure, Apple can continue to make lots of money with only 10% market share, as they already do in the laptop business. However, that doesn't change the fact that they don't drive the industry the way they did when they first started out. This is the point of that slide.
Uh, the last time I was in a T-Mobile store they had S3s on the shelf.
As far as ratings go - last time I checked they were outselling the iPhone...
My only reason for not gravitating towards the current Nexus: There's no removable SD storage available. Not a major issue, just a preference.
Yes, but your choices are basically lousy upgrade support, or no removable SD storage. The Nexus phones are decent, but generally only superior in the sense that they don't suck as much. I didn't get one last time because it was inferior in hardware to the phone I did get, but it has been a battle getting the later OS versions to work since nobody releases source for hardware drivers (INCLUDING the Nexus phones).
Carrier subsidies are often a problem as well - you can get a Nexus device that will work on AT&T right now, but you'll pay hundreds of dollars for it and you won't get any discount for AT&T for not getting a phone from them. Your only real option is to get a discounted phone from AT&T, eBay it, and then use some of that to defray the cost. Talk about a crazy market.
I'll probably get a Nexus phone next time, but it would be nice to have a better selection of phones that actually get support. Hopefully the rumors are true and there will be more than one Nexus device next time around...
In the US a typical phone costs $199 with a commitment to an overpriced data plan - usually something like $30-40/mo.
Retail values of the phones without that commitment (assuming you can get them) tend to be more like $600 or so, for the nicer phones.
T-Mobile is about the only major carrier I'm aware of that will let you divorce the one from the other, and then you can get a data plan for $5-10/mo (the latter with 2GB of 4G data, and unlimited 2G data). Phone subsidies and exclusivity deals are really something that needs to go away.
They work at 4G speeds as well - just not 3G. I can't imagine T-Mobile's 3G network is much larger than their 4G network.
Ugh - there is room in the market for phones that AREN'T cookie-cutter copies of each other. What if I want a smartphone with a small screen - nobody would call that a flagship phone so nobody would make that a single-product focus. How about a phone with a keyboard - most people don't want that, so nobody would make that their single product.
The whole point of Android is that you actually get a choice. I don't want that to be a choice of 3 vendors who all make phones designed to look just like an iPhone...
I tend to agree. However, UI optimization has to be the key.
My wife is still using a 4 year old feature phone because the newer models are all inferior. Tasks she does often are buried in the menus (things like pull up contacts, send/read SMS, etc.). Stuff she doesn't care about is front-and-center, like browse the web, look for ringtones, and all that.
It seems like the newer devices are just designed to get people who don't have data plans to accidentally pull a kilobyte of data here or there so they can be charged through the noses.
Sure, feature phones are becoming a niche, but if you're going to make one at least make it user-centric, and not carrier-centric.
This doesn't really counter the argument against vendor lock-in. Also the vast majority of ARM devices are actually NOT bootloader locked. Just go and check the compatibility list for Cyanogenmod.
Uh, virtually all the devices on the Cyanogenmod compatibility list HAVE locked bootloaders. They just have poorly implemented locked bootloaders/etc, which means that you can defeat them and install Cyanogenmod anyway. The only Android device I'm aware of that have unlocked bootloaders are the Nexus line, which probably have
Anyway the main point of my post was that there's a significant number of people who are interested in doing with their devices whatever the hell they want, and this even includes installing Android on an iPhone. People want to do this and given it's their hardware we should not be promoting systems to prevent this.
I'm all for legislation that requires device owners to be given all the stuff required by GPLv3/etc. However, I've got nothing against the technology itself. I'd prefer a device with UEFI to what I have on my PC now. Right now anybody who wants to can just install a keylogger on my hard drive just by booting from a CD, and there is little I could do to stop it. Rootkits installed over the internet can make themselves almost undetectable unless you boot off of clean media. UEFI has the power to change that. My only objection to the proposed implementation is the inability to change the OEM-supplied machine key (that isn't the one used to sign windows) - I want to be able to change it all.
Where did I say that this article had anything to do with detectors? I said that people were spending lots of money on THz detectors, which is true. The argument was that working with THz radiation wasn't anything unusual, and my point was that the reason people were spending all kinds of money on the technology were doing so because it was.
What a crap argument. So you say that currently Linux isn't for Joe Sixpack and therefore it is OK to keep it that way? No. We want the Linux to be available for everyone!
Personally, I'd rather my bootloader be secure against malware than for Linux to be available for everyone. The former is quite achievable with UEFI, and the latter is nebulous at best. Plus attempts to do the latter have steadily tended to make Linux less usable for me. It isn't like I'm the one getting paid for every PC that Ubuntu or whatever gets installed on - my only possible benefit is if some of those users contribute back, and I doubt I'll get much of that from folks who can't reconfigure their firmware.
That pretty much guarantees some percentage of users will brick their systems.
Uh, how exactly is changing a firmware setting going to brick a system? You don't need to reflash the firmware to change the boot key, and per the spec the firmware has to have a factory reset function that restores all settings including keys to the shipped settings.
The sort of user who couldn't turn off secure boot in their firmware probably couldn't reinstall Windows if they got frustrated with Unity or whatever anyway... :) If you want to target these users you need to ship the PCs pre-configured.
I'm not trolling, it's a reality check.
Most big software projects I've seen fail hard, like millions and 10's of wasted dollars hard. By comparison you just dont see that very often in big electrical/mechanical/civil projects, which can be equally complex (eg refineries, cruise liners etc).
It really depends on the engineering project. When you're talking about designing a cruise liner that has 50 more rooms than the last one, that might cost millions of dollars, but the complexity is more analogous to a software point release. Building the thing is like running make. Neither of these two operations tend to have massive cost overruns in the software world.
Engineering projects have a greater tendency to be about making incremental improvements on work that has already been done, if you're even making improvements at all. The number of interacting components tends to be much smaller as well compared to software. The typical software project tends to involve things that nobody on the team has ever done in quite that way before. When you do get engineering projects like building a new manned spacecraft or fighter jet or aircraft carrier they tend to be loaded with overruns.
Engineering also involves human beings in the construction, vs software which uses computer programs to do all the compilation. If your ship contains a wire that runs along and then just stops and doesn't connect to anything, chances are that somebody is going to report that when they're actually running the wire. It is also unlikely that the ship just dynamically "allocates" wires and other interacting parts while it is at sea.
I don't think competence is really the fundamental problem so much as the problem is just that much more difficult.
Software engineering is still playing catch up, in the sense that most developers and development companies I've seen still dont follow a formal enough process for it really to be called engineering. Usually it's a bunch of computer science graduates having a wild stab at it, and the good ones are closer to artisans than engineers...Until the entire software industry gets off it's high horse and admits this to itself - and more importantly admits this to the customers, we are going to continue to be dissapointed with the quality.
People only state that they're disappointed in software quality in a vacuum. When their free smartphone app crashes they say, "man, we should have better quality software." They don't say, "man, we should have better quality software - I'd have gladly paid $100 for what was a free app." Quality costs money, and completely formalizing all aspects of software development the way they formalize space shuttle design would cost a LOT of money (and as I recall you're still much more likely to be killed by a space shuttle ride than driving a car). At work somebody was griping about having to fix a cosmetic bug from the last round of development, just after complaining about the fact that we can't make all their desired changes due to budget. Well, if the team were to increase quality to cut down on even cosmetic bugs then FAR fewer changes would be made.
Software needs to be good enough. If my rain gutters had a slow drip somewhere along the line, chances are I wouldn't spend hundreds of dollars to have it repaired.
Are you suggesting that if I sell somebody a phone it shouldn't be able to boot unless they insert an install SD card or such? Or does this just pertain to PCs? Most people buy PCs with pre-installed OSes. Is there really any value to making it so that those PCs can't be booted without sticking in a CD as the first step?
And if that happened, how would that help? Anybody with a Windows PC will have stuck in the Windows CD, which will install the MS key and now it won't boot linux when they want to switch.
If the firmware has to accept any key it finds on any later boot, well then secure boot gets you nothing anyway since the malware can just supply a new key.
While I'm all for getting rid of lock-in, the fact is that almost all arm-based system in consumer use have locked bootloaders already. Just about every android phone in use falls into this category (and yes, I know the sliver of market share held by Nexus devices are an exception).
I guess you meant fool-proof. And it is. It is fool-proof against all those fools who want to decide to run their own code on the computer without having to ask permission beforehand.
You don't have to ask for permission - you just have to configure your computer to boot it. If you stick an Ubuntu CD in a PC that isn't configured to boot off of CD, it won't run that either unless you "ask for permission" by telling it to boot from CD.
On amd64 at least you'll be able to disable it if you want, or configure it with your own keys, so that MS won't be able to install something on your PC without asking YOU for permission.
Software doesn't have to be perfect. It simply needs to be economical due to its qualities (again, qualities well known for decades.)
This last bit needs to be emphasized - perfect software that you can't afford is useless. As is often attributed to Stalin, sometimes quantity has a quality all its own. If some other start-up's hacked together social media site that isn't scalable lets them get to 10k users before yours is written, then the network effect will prevent you from ever taking advantage of the fact that yours can handle a billion, and they'll have plenty of money to spend doing a complete rewrite.
Usually the way these sorts of things get looked at in the real world involves some kind of FMEA approach (failure modes effect analysis). You look at what can go wrong, how often it will go wrong, what happens when it goes wrong, and whether any of that can be mitigated and if so how likely it is that the problem will be spotted and mitigated.
There are lots of ways to do it, but the bottom line is that not all problems are created equal, as you've stated. If I've got some system that "just works" placing that stability at risk to fix some UI glitch that isn't causing problems is a dumb decision, especially if it costs a lot of money.
I think a better question would be "Why does it NEED to be better?" and the answer in most cases is "it doesn't". Remember perfect is the enemy of good and often the difference between "good enough" and great can be truly insane levels of expense.
Yup, just the other day at work we had somebody complain about a cosmetic regression in a piece of highly specialized software. They were basically on a crusade for quality, and pushing hard to have it fixed. I basically went to the project managers and pointed to the pile of much more critical needs and fixed budget, and it is very unlikely we'll ever get to the cosmetic issue as a result.
If this were something displayed on the company homepage I'd look at it differently, but for an internal piece of software it just isn't worth the trouble. Some might argue that a good quality system would not result in such regressions in the first place, but again I have to point out to the pile of high-importance needs and fixed budget. If I have to choose between a system improvement that has documented ROI and creates a business opportunity, and spending money so that the stuff we deliver is less likely to have very minor bugs, I'll pick the feature. If we were talking about regressions with a significant functional impact then that would be a different matter.
Scope*Quality, Cost, Time - pick two. Pick the wrong balance and your company might go out of business...
The real version is on, but only on BBCA and with a considerable delay. I can understand how the folks across the pond feel about US series.
Well, the appropriate solution varies considerably by industry.
In the case of telecom I'd probably treat them as has been done with electricity - get rid of vertical integration. regulate the living daylights out of the last mile and anything that uses spectrum as natural monopolies, and then make the rest free market.
So, if you have phone service or DSL you'd pay a company $8/month to run a twisted pair from a CO to your home and keep it running to FCC standards. Then you'd get to buy telephone/internet/etc service from any of a multitude of companies who pay the telco a standard rate per 1U of rackspace and per termination.
If you have a cell phone then it would connect to a tower operated by a tower owner. Towers would use standard protocols on standard frequencies under FCC regulation. Towers would charge by the packet, but would not bill customers directly. Within any area no company could own more than a third of the spectrum for any piece of land, so that there would always be competition between redundant operators for any area. Tower operators could set their rates, but would have to charge the same rate per packet to all of their customers, who would be national carriers or wholesalers. The national carriers would be forbidden from owning towers - they could buy coverage from operators in whatever manner makes sense, and put together plans in whatever way makes sense to them, and would sell service to the customer. The national carriers would not be regulated.
The goal would be to separate the physical layer that is expensive to maintain and tends to have high barriers to entry from the content layer which has much lower barriers to entry. This is similar to how many areas are breaking up electrical generation from transmission/distribution/etc.
Sure, it isn't perfect, but it cuts down on the number of big pockets who can bribe regulators, and limits the scope of regulation further than it is now while making the regulation of those areas more effective.
Sure, but in China you only get punished for accepting bribes if you embarrass the party.
Bribery is rampant in China. I hear they're trying to reduce it, but it is a big culture issue.
Take the recent cases of melamine in infants formula that led to a few executed regulatory officials. While officially they were executed for accepting bribes, what they were really punished for is letting botched formula through. The correct way to accept bribes in China is to accept them only to approve things that you should have approved anyway, not to approve things that shouldn't have been approved.
If you want a car analogy in the USA, imagine your kid takes a driver's license exam and does well. The officer administering the test tells them that they do well, but they need to process the paperwork before passing them and that could take a long time. The kid hands the officer $100, and the officer says hey, I can do the paperwork right now while you wait and they check three boxes and sign a form and hand it back to them and they can drive. Or they can not hand them the $100 and check back in a few weeks and find out it just isn't getting done. The officer isn't passing somebody who should fail - they're just adding a "tax" of sorts to the process.
By Western standards it is horribly corrupt. My intent is just to illustrate how the culture works. I remember reading in C&EN about 15 years ago that it was routine for huge companies to pay bribes to offload supertankers full of chemicals and such. Everything they were doing (aside from the bribes) was legal and would be allowed in any country, but if they didn't pay the bribes their expensive tanker would just rot in port for a few weeks while the paperwork was sorted out. I knew a guy who once was involved in setting up a plant in South Korea and they had a budget item for bribes - often for things like importing components made in Japan (this was many years ago, and the purpose of the law/bribes was intended more as an insult to Japan than anything). Business culture in Asia is often different, though many of these examples were from a long time ago.
At that frequency what would be the range of such a radar, except in more exotic conditions like transmitters at very high elevation or in deserts? I would think that atmospheric absorption would be a big problem.
Maybe for air-to-air radar it might be practical - once the transmitter is at altitude the absorption of moisture would be less of an issue.