To get downgrade rights it has to be the pro or enterprise edition and it must be from the OEM or volume license channels. Retail copies (whether full product or upgrade) don't come with downgrade rights.
So if you have a laptop that came with the basic edition of "windows 8" (for some silly reason it seems MS has decided not to give this edition a name) afaict your only options for dowgrading are to either buy a complete new copy of windows or to buy an upgrade through a volume license program:(
The problem comes when the people in charge think something is better and it gets pushed on others even though they disagree about it being better. We have seen this happen sufficiently often with software (both open and closed source) recently that whenever something like this is announced we fear it happening again.
They're still the same basic computer architecture as traditional PCs, with the primary difference being the interface.
tablets come into roughly two categories.
smartphone based tablets (ipad, nexus, kindle fire etc) are designed as consumption and communication devices. They can be programmed but doing so without a seperate computer is going to be painful to say the least.
tablet PCs on the other hand are PCs that have tablet functionality and you can easilly use them to do anything you can use a regular PC for but they have so-far been too bulky and heavy to really use as tablets.
And with Microsoft's recent release of the Surface Pro there are now tablets that will run the same software as PCs.
There have been tablet PCs for years, it's just they were generally too big and heavy to really use as tablets.
What makes the surface pro potentially interesting is it has the functionality of a PC based tablet while coming pretty close to being thin and light enough to really use as a tablet. They aren't quite there yet with the current version of the product but I suspect they will be in the not too distant future.
Often abbreviations take on meanings that differ from their constituent words for example
http doesn't reffer to any protocol designed for transferring hypertext, it reffers to a specific protocol for transffering hypertext. ftp doesn't reffer to any protocol for transffering files, it reffers to a specific protocol for transferring files. ssh doesn't reffer to any protocol for establishing a shell connection securely, it reffers to a specific protocol for establishing a shell connection securely
Similarly pc doesn't usually reffer to just a device that is personal and a computer. It's more than that, exactly how much more depends on who you talk to, often PC is taken to mean "IBM compatible personal computer", some use a wider definition with just how much wider depending on who you talk to.
It appears that canalys are excluding smartphones but including smatphone-like tablets. That makes even less sense than excluding both smartphones and smartphone-like tablets.
you could easily port any programming language to that platform.
Suppose you were locked in a room with a brand new ipad, an internet connection and a bluetooth keyboard and asked to write some software. You'd be pretty stuffed, apple don't allow programming languages on the appstore and jailbreaking or using the official developer tools requires you to connect your ipad to a "PC" or mac. You might be able to hack something together in javascript in the browser but that would be about the limit of what you could do.
Finally suppose you were locked in a room with a brand new nexus 10, an internet connection and a bluetooth keyboard. You could download some programming tools from the andriod market but still you would be frustrated by the fact that the device ships in a locked down state and the only way to unlock it is to connect it to another computer and the fact that all the developer kits are designed for use on a PC.
Finally suppose you were locked in a room with a brand new surface pro, the keyboard cover and an internet connection. You'd have no trouble downloading and installing the compiler for any language you like. It does ship with a locked down bootloader by default but you can disable that lockdown without needing another computer.
To me there is a big difference between these three cases.
Some ISPs certainly do this sort of thing already.
Whether it's a good idea really depends on the ratio of different types of traffic. Packing all the traffic from "heavy downloaders" into a free period in the middle of the night could end up moving the peak to the free period and making the peak worse than it was before.
There is a fundamental compromise in flying between size and efficiency.
Flying machines come into two basic categories, "lighter than air"* and "heavier than air". In both cases gravity is imparting a constant downward force on the flying machine, in the lighter than air case this is balanced by boyancy, in the heavier than air case it is balanced by pushing air downwards.
Lighter than air flying machines end up big because air has a much lower density than cargo or construction materials so to get the density down the vast majority of your craft has to be flotation gas.
Heavier than air flying machines also have a size compromise. Momentum is linearly proportional to speed while kinetic energy is proportional to the square of speed**. So for efficiency you want to push a large mass or air downwards slowly rather than a small mass downwards quickly but that means your craft needs to cover a large area. They aren't quite as big as lighter than air machines but even a small two seater helicopter is too wide (with the rotors running, if it's a two blade rotor then it can be aligned front-back to fit within a road for transport) to take off from a single lane of an ordinary road.
Thinking about it the only way I see a flying car*** could be possible would be massive advances in lightweight but strong materials. If you could reduce the weight of your flying machine to weigh barely more than the pilot and fuel you could make it a lot smaller. An engineer can tell you what properties your materials need to have but discovering how to produce materials with those properties is another matter.
I suspect a fly/drive hybrid will make it to market at some point but limited usability (it will need a dedicated takeoff area and will likely have far less internal space than a normal car) and high costs (getting to the limits of strength vs weight gets pricy) will keep it as a toy for the rich.
* Really lower average density than air. ** Until you get near light speed and relativity gets involved ***That is a vehicle in which I can load as much as a regular car and which I can drive out of a normal garage onto a normal road, then fly off without needing a dedicated launch area or needing to reconfigure the shape of the vehicle.
Not really to count it as more than one generation behind you'd have to count die shrinks with minor design tweaks as "generations"* and you would also have to consider ivy bridge as being "current" even though there are no dual socket capable ivy bridge parts on the market yet.
still based on the first-gen Core i7 Xeons
Apple brought in the 6 core gulftown (westmere microarchitecture) processors with the mid 2010 model and replaced the quad cores with gulftown versions with the early 2012 model.
It certainly feels like it's been a long time since there has been a significant update but that is partly because intel moved from a model of doing the high end desktop/dual socket server parts first and then the "laptop"/"mainstream desktop" parts later to a model of doing the laptop/mainstream desktop parts first and the "high end desktop"/"dual socket server parts"
*intel has been kind of inconsistent on that reffering to sandy bridge as "second generation i7" and it's shrink ivy bridge as "third generation i7" implying that they count both nahelm and its shrink westmere as "first generation". This may have been partly because you needed both nahelm and westmere to get a full product lineup.
There is a general rule in safety regulations that things that require a tool or excessive force to open are treated differently from things that can be easilly opened without tools.
Early xbox 360s had major reliability problems caused by a combination of poor thermal management and the newfangled lead free solder. 5 failures for one gamer is pretty high put not unbelivable given the failure rates I saw reported. Heck things were so bad that they made a special motherboard version (opus) for repairing failed boxes that didn't have HDMI ports.
Once they admitted the problem they did handle it pretty well extending the warranties for consoles affected by the issue (though the extension wasn't always enough:( ), making the RMA process as painless as possible and replacing the main chips with more power efficient versions that didn't get as hot but still it's the sort of fuckup that shows a company inexperianced in hardware design and testing.
And there were other hardware issues too like not supporting the disk adequately so a slight knock can scratch it especially if the console is vertical.
mmm, both the PS3 and xbox 360 were pretty noisy at the start of this generation, people still bought them and they gradually got quieter as the generation went on.
What gets me with my PS3 (a 40GB UK model) is it seems to have two noise levels, too quiet to notice while playing and sounding like someone is hoovering in the next room. It switches between these noise levels periodically while playing.
Button space on a controller is premium space and doesn't need to be wasted on "fluff".
Button space in the quickly accessible locations is premium space.
OTOH there is plenty of space for extra buttons that aren't used so often in the middle of the controller. Right now the dualshock 3 has three buttons in that area and there is plenty of room for a fourth.
Steam already claims that if you receive a gift that was bought using a stolen credit card they will revoke the copy of the game. They have also revoked copies that were sold in the wrong region. They also offer guest passes and free weekends. So the infrastucture to remove games from accounts is clearly already there.
The games has to be downloaded by more users
That much is true. Whether it will be "a lot" of stress or a negligable change is another matter.
"Steamworks is entirely free. There are no licensing fees and there’s no charge for bandwidth, retail copies, or OEM distribution."
In other words steam will distribute your software for free in exchange for bringing customers to the platform. I think in light of that it will be hard to argue that the bandwidth from resales is an undue burden.
Actually doing that would likely be legally problematic both because of agreements with publishers and because by the time the company is going out of buisness gabe may well no longer be in control or at least be under restrictions imposed by a bankrupcy court.
Unless that term is actually in the agrement with third party escrow to make sure it happens then IMO it's pretty worthless. Talk is cheap.
Kings cross serves the east cost main line to cambridge leeds, york etc. Services to cambridge also branch off from that line St pancras serves kent via high speed one and also serves brighton via thameslink Euston just down the road serves the west coast mainline to birmingham, manchester etc All three of these stations also have various local connections.
Between many tube services to different parts of london, eurostar services to the continent and domestic train services to pretty much everywhere except the southwest it wouldn't surprise me if it's the best connected location in the UK.
In the early 2000s, games (at least all the ones I played) could be installed and played without online activation. Many games did prevent multiple simultaneous uses of the same CD key online which made buying a used copy for online use a little risky but that was about the extent of problems with resale.
Steam was first made available to the public in 2002 but didn't really rise to prominance until 2004 when valve shut down it's WON servers and released half life 2. People bitched and moaned a lot but ultimately desire to continue playing half life/counter strike online and desire to play half life 2 at all outweighed dislike of steam and gamers sucked it up. As the performance issues were sorted out some even started to like steam. Having seen gamers put up with steam gradually everyone else started putting some form of online activation into their games too.
Unless you are using them day in day out regexes have a habbit of looking very much like line noise.
By putting regexes into the language at syntax level (rather than merely as a library function) they elevated it's status and encourage people to turn to it first. IMO the questions of "should regexes be used as the default way of solving matching/parsing problems" and "what format is best when you do need a regex" are orthogonal questions.
The problem with that approach is that the person with the raw name is very likely to get email intended for the other people with the same name.
Depending on where those people sit in the organisation that may be a major PITA for the person with the plain name and it may also be a security issue. IMO if introducing a new system and there are people with the same "default name" it's better not to use the "default name" at all.
In my opinion git provides better history guarantees than most other VCSes.
Lets say I have a main repository. With subversion provided I trust those operating the main repository I can easilly answer the following questions
What code was at the head of a given branch in the main repository at a given time. How did the code at the head of the branches in the main repository change over time. Who actually placed that code in the main repository (and in doing so took responsibility for it).
To me being able to answer these questions is the essence of "history" in a version control system.
Afaict git has no good system for answering these questions. Reflog can give you this information but afaict it can only be used if you have direct shell access to the main repository and only if the relavent entries haven't been expired from the reflog.
I don't think that explains it, lets look at the minimum disk space requirements for various versions of windows
Conventional line:
windows for workgroups: 6.2MB (plus a bit for dos so lets say 15MB total) windows 95: 55MB windows 98: 175MB
NT line: windows NT3.51: 90MB windows NT4: 124MB windows 2000: 650MB windows XP: 1.5GB Windows vista: 15GB Windows 7: 16GB for 32-bit 20GB for 64-bit Windows 8: same as windows 7
So even if each version of windows really contained a seperate copy of the previous version (in reality it doesn't, while there are some compatibility layers the core components haven't changed much in a long time) in most cases the new crap would still be larger than the copy of the old crap.
What is your understanding of the word "jailbreaking"? My understanding is "using a bug/security hole to gain more access to a system you own but that was locked down by the manufacturer".
Assuming my understanding is correct then many (but not all) andriod devices can have their bootloader unlocked without resorting to security holes (though they often wipe user data during the unlock process for "security reasons").
And afaict tablet PCs have not been locked down historically either.
A quick search doesn't seem to reveal whether the surface pro is unlocked like conventional tablet PCs or locked like like the surface rt.
The problem is that since the whole point of a plea bargin is to bypass the process of determining guilt you have no way of measuring how many people decided it was a better idea to confess to a crime they did not commit than to risk a trial with the potential (however unlikely) to lock them up for life. In other words there is no way to determine what proportion of plea deals are bad.
I don't believe the court system would handle everything going to trial.
If that is true then that points to an even deeper problem. Either your law has too many crimes or your court system needs fixing.
However this does have a downside especially for small planes.
Ultimately it is simply not possible to reduce risk to zero and if we are going to do anything in life then we have to accept that risk (even risk to life) MUST be balanced against cost. Since red tape increases the cost of safety systems it further follows that to maximise safety for a given "safety budget" a balance is needed between ensuring that the safety systems are sufficiently safe and not driving up the cost so much that safety improvements (whether fitting a new safety system to an existing plane or moving to a newer safer plane entirely) simply get rejected completely
For airlines the system seems to work pretty well. Even with the high cost imposed by safety compliance the new planes are a sufficient improvement in fuel efficiency and reliability to make buying them economically worthwhile. But for other types of flying afaict it is quite common to see decades old planes still in service either because newer planes are too expensive or because there is simply noone making a suitable plane anymore. For non-commerical flying it's also common to see homebuilt planes registered as "experimental". Does this combination of old planes and homebuilt planes really make flying in small planes safer than it would be if lighter regulations made new small planes more affordable and allowed a better choice of new small planes to be offered?
On pretty much every 32-bit unix-like system int, long, void* and time_t are all 32 bit. The only exception i'm aware of is the new "x32 abi".
On 32-bit windows int, long and void* are 32 bit. time_t is not considered a core system type but part of the C libraries (which on windows unlike unix like systems is not a core system library) so whether it is true depends on what compiler you are using. In visual studio 2005 the default size of time_t was changed from 32-bit to 64-bit.
It should be possible to add 64-bit time_t support to a 32-bit unix like operating system by taking an approach similar to that used for large file support. I haven't heard of anyone actually trying to do it though and applications would still at best need to be recompiled with the define to enable the new "large time support" mode and at worst may require considerably more work.
Slashdot Mirror
To get downgrade rights it has to be the pro or enterprise edition and it must be from the OEM or volume license channels. Retail copies (whether full product or upgrade) don't come with downgrade rights.
So if you have a laptop that came with the basic edition of "windows 8" (for some silly reason it seems MS has decided not to give this edition a name) afaict your only options for dowgrading are to either buy a complete new copy of windows or to buy an upgrade through a volume license program:(
The problem comes when the people in charge think something is better and it gets pushed on others even though they disagree about it being better. We have seen this happen sufficiently often with software (both open and closed source) recently that whenever something like this is announced we fear it happening again.
Note that cooked line input is not part of the "console", it's part of the tty system.
They're still the same basic computer architecture as traditional PCs, with the primary difference being the interface.
tablets come into roughly two categories.
smartphone based tablets (ipad, nexus, kindle fire etc) are designed as consumption and communication devices. They can be programmed but doing so without a seperate computer is going to be painful to say the least.
tablet PCs on the other hand are PCs that have tablet functionality and you can easilly use them to do anything you can use a regular PC for but they have so-far been too bulky and heavy to really use as tablets.
And with Microsoft's recent release of the Surface Pro there are now tablets that will run the same software as PCs.
There have been tablet PCs for years, it's just they were generally too big and heavy to really use as tablets.
What makes the surface pro potentially interesting is it has the functionality of a PC based tablet while coming pretty close to being thin and light enough to really use as a tablet. They aren't quite there yet with the current version of the product but I suspect they will be in the not too distant future.
Often abbreviations take on meanings that differ from their constituent words for example
http doesn't reffer to any protocol designed for transferring hypertext, it reffers to a specific protocol for transffering hypertext.
ftp doesn't reffer to any protocol for transffering files, it reffers to a specific protocol for transferring files.
ssh doesn't reffer to any protocol for establishing a shell connection securely, it reffers to a specific protocol for establishing a shell connection securely
Similarly pc doesn't usually reffer to just a device that is personal and a computer. It's more than that, exactly how much more depends on who you talk to, often PC is taken to mean "IBM compatible personal computer", some use a wider definition with just how much wider depending on who you talk to.
It appears that canalys are excluding smartphones but including smatphone-like tablets. That makes even less sense than excluding both smartphones and smartphone-like tablets.
you could easily port any programming language to that platform.
Suppose you were locked in a room with a brand new ipad, an internet connection and a bluetooth keyboard and asked to write some software. You'd be pretty stuffed, apple don't allow programming languages on the appstore and jailbreaking or using the official developer tools requires you to connect your ipad to a "PC" or mac. You might be able to hack something together in javascript in the browser but that would be about the limit of what you could do.
Finally suppose you were locked in a room with a brand new nexus 10, an internet connection and a bluetooth keyboard. You could download some programming tools from the andriod market but still you would be frustrated by the fact that the device ships in a locked down state and the only way to unlock it is to connect it to another computer and the fact that all the developer kits are designed for use on a PC.
Finally suppose you were locked in a room with a brand new surface pro, the keyboard cover and an internet connection. You'd have no trouble downloading and installing the compiler for any language you like. It does ship with a locked down bootloader by default but you can disable that lockdown without needing another computer.
To me there is a big difference between these three cases.
Some ISPs certainly do this sort of thing already.
Whether it's a good idea really depends on the ratio of different types of traffic. Packing all the traffic from "heavy downloaders" into a free period in the middle of the night could end up moving the peak to the free period and making the peak worse than it was before.
There is a fundamental compromise in flying between size and efficiency.
Flying machines come into two basic categories, "lighter than air"* and "heavier than air". In both cases gravity is imparting a constant downward force on the flying machine, in the lighter than air case this is balanced by boyancy, in the heavier than air case it is balanced by pushing air downwards.
Lighter than air flying machines end up big because air has a much lower density than cargo or construction materials so to get the density down the vast majority of your craft has to be flotation gas.
Heavier than air flying machines also have a size compromise. Momentum is linearly proportional to speed while kinetic energy is proportional to the square of speed**. So for efficiency you want to push a large mass or air downwards slowly rather than a small mass downwards quickly but that means your craft needs to cover a large area. They aren't quite as big as lighter than air machines but even a small two seater helicopter is too wide (with the rotors running, if it's a two blade rotor then it can be aligned front-back to fit within a road for transport) to take off from a single lane of an ordinary road.
Thinking about it the only way I see a flying car*** could be possible would be massive advances in lightweight but strong materials. If you could reduce the weight of your flying machine to weigh barely more than the pilot and fuel you could make it a lot smaller. An engineer can tell you what properties your materials need to have but discovering how to produce materials with those properties is another matter.
I suspect a fly/drive hybrid will make it to market at some point but limited usability (it will need a dedicated takeoff area and will likely have far less internal space than a normal car) and high costs (getting to the limits of strength vs weight gets pricy) will keep it as a toy for the rich.
* Really lower average density than air.
** Until you get near light speed and relativity gets involved
***That is a vehicle in which I can load as much as a regular car and which I can drive out of a normal garage onto a normal road, then fly off without needing a dedicated launch area or needing to reconfigure the shape of the vehicle.
the CPU is a few generations behind
Not really to count it as more than one generation behind you'd have to count die shrinks with minor design tweaks as "generations"* and you would also have to consider ivy bridge as being "current" even though there are no dual socket capable ivy bridge parts on the market yet.
still based on the first-gen Core i7 Xeons
Apple brought in the 6 core gulftown (westmere microarchitecture) processors with the mid 2010 model and replaced the quad cores with gulftown versions with the early 2012 model.
It certainly feels like it's been a long time since there has been a significant update but that is partly because intel moved from a model of doing the high end desktop/dual socket server parts first and then the "laptop"/"mainstream desktop" parts later to a model of doing the laptop/mainstream desktop parts first and the "high end desktop"/"dual socket server parts"
*intel has been kind of inconsistent on that reffering to sandy bridge as "second generation i7" and it's shrink ivy bridge as "third generation i7" implying that they count both nahelm and its shrink westmere as "first generation". This may have been partly because you needed both nahelm and westmere to get a full product lineup.
It can also be easily LOCKED [mac-pro.com]
I wouldn't consider having to buy a special third party lock because apple didn't think to design in a simple padlock hole to be "easilly LOCKED".
to prevent theft AND OPENING. So NOW who's got the responsibility?
The company who supplied it in a state where it could be opened without a tool.
There is a general rule in safety regulations that things that require a tool or excessive force to open are treated differently from things that can be easilly opened without tools.
Early xbox 360s had major reliability problems caused by a combination of poor thermal management and the newfangled lead free solder. 5 failures for one gamer is pretty high put not unbelivable given the failure rates I saw reported. Heck things were so bad that they made a special motherboard version (opus) for repairing failed boxes that didn't have HDMI ports.
Once they admitted the problem they did handle it pretty well extending the warranties for consoles affected by the issue (though the extension wasn't always enough :( ), making the RMA process as painless as possible and replacing the main chips with more power efficient versions that didn't get as hot but still it's the sort of fuckup that shows a company inexperianced in hardware design and testing.
And there were other hardware issues too like not supporting the disk adequately so a slight knock can scratch it especially if the console is vertical.
mmm, both the PS3 and xbox 360 were pretty noisy at the start of this generation, people still bought them and they gradually got quieter as the generation went on.
What gets me with my PS3 (a 40GB UK model) is it seems to have two noise levels, too quiet to notice while playing and sounding like someone is hoovering in the next room. It switches between these noise levels periodically while playing.
Button space on a controller is premium space and doesn't need to be wasted on "fluff".
Button space in the quickly accessible locations is premium space.
OTOH there is plenty of space for extra buttons that aren't used so often in the middle of the controller. Right now the dualshock 3 has three buttons in that area and there is plenty of room for a fourth.
and extra added overhead to verify ownership
Steam already claims that if you receive a gift that was bought using a stolen credit card they will revoke the copy of the game. They have also revoked copies that were sold in the wrong region. They also offer guest passes and free weekends. So the infrastucture to remove games from accounts is clearly already there.
The games has to be downloaded by more users
That much is true. Whether it will be "a lot" of stress or a negligable change is another matter.
Note that http://www.valvesoftware.com/business/ claims
"Steamworks is entirely free. There are no licensing fees and there’s no charge for bandwidth, retail copies, or OEM distribution."
In other words steam will distribute your software for free in exchange for bringing customers to the platform. I think in light of that it will be hard to argue that the bandwidth from resales is an undue burden.
Actually doing that would likely be legally problematic both because of agreements with publishers and because by the time the company is going out of buisness gabe may well no longer be in control or at least be under restrictions imposed by a bankrupcy court.
Unless that term is actually in the agrement with third party escrow to make sure it happens then IMO it's pretty worthless. Talk is cheap.
And then there are the domestic rail services.
Kings cross serves the east cost main line to cambridge leeds, york etc. Services to cambridge also branch off from that line
St pancras serves kent via high speed one and also serves brighton via thameslink
Euston just down the road serves the west coast mainline to birmingham, manchester etc
All three of these stations also have various local connections.
Between many tube services to different parts of london, eurostar services to the continent and domestic train services to pretty much everywhere except the southwest it wouldn't surprise me if it's the best connected location in the UK.
BS,
In the early 2000s, games (at least all the ones I played) could be installed and played without online activation. Many games did prevent multiple simultaneous uses of the same CD key online which made buying a used copy for online use a little risky but that was about the extent of problems with resale.
Steam was first made available to the public in 2002 but didn't really rise to prominance until 2004 when valve shut down it's WON servers and released half life 2. People bitched and moaned a lot but ultimately desire to continue playing half life/counter strike online and desire to play half life 2 at all outweighed dislike of steam and gamers sucked it up. As the performance issues were sorted out some even started to like steam. Having seen gamers put up with steam gradually everyone else started putting some form of online activation into their games too.
Unless you are using them day in day out regexes have a habbit of looking very much like line noise.
By putting regexes into the language at syntax level (rather than merely as a library function) they elevated it's status and encourage people to turn to it first. IMO the questions of "should regexes be used as the default way of solving matching/parsing problems" and "what format is best when you do need a regex" are orthogonal questions.
The problem with that approach is that the person with the raw name is very likely to get email intended for the other people with the same name.
Depending on where those people sit in the organisation that may be a major PITA for the person with the plain name and it may also be a security issue. IMO if introducing a new system and there are people with the same "default name" it's better not to use the "default name" at all.
In my opinion git provides better history guarantees than most other VCSes.
Lets say I have a main repository. With subversion provided I trust those operating the main repository I can easilly answer the following questions
What code was at the head of a given branch in the main repository at a given time.
How did the code at the head of the branches in the main repository change over time.
Who actually placed that code in the main repository (and in doing so took responsibility for it).
To me being able to answer these questions is the essence of "history" in a version control system.
Afaict git has no good system for answering these questions. Reflog can give you this information but afaict it can only be used if you have direct shell access to the main repository and only if the relavent entries haven't been expired from the reflog.
I don't think that explains it, lets look at the minimum disk space requirements for various versions of windows
Conventional line:
windows for workgroups: 6.2MB (plus a bit for dos so lets say 15MB total)
windows 95: 55MB
windows 98: 175MB
NT line:
windows NT3.51: 90MB
windows NT4: 124MB
windows 2000: 650MB
windows XP: 1.5GB
Windows vista: 15GB
Windows 7: 16GB for 32-bit 20GB for 64-bit
Windows 8: same as windows 7
So even if each version of windows really contained a seperate copy of the previous version (in reality it doesn't, while there are some compatibility layers the core components haven't changed much in a long time) in most cases the new crap would still be larger than the copy of the old crap.
What is your understanding of the word "jailbreaking"? My understanding is "using a bug/security hole to gain more access to a system you own but that was locked down by the manufacturer".
Assuming my understanding is correct then many (but not all) andriod devices can have their bootloader unlocked without resorting to security holes (though they often wipe user data during the unlock process for "security reasons").
And afaict tablet PCs have not been locked down historically either.
A quick search doesn't seem to reveal whether the surface pro is unlocked like conventional tablet PCs or locked like like the surface rt.
Not all plea deals are bad
The problem is that since the whole point of a plea bargin is to bypass the process of determining guilt you have no way of measuring how many people decided it was a better idea to confess to a crime they did not commit than to risk a trial with the potential (however unlikely) to lock them up for life. In other words there is no way to determine what proportion of plea deals are bad.
I don't believe the court system would handle everything going to trial.
If that is true then that points to an even deeper problem. Either your law has too many crimes or your court system needs fixing.
However this does have a downside especially for small planes.
Ultimately it is simply not possible to reduce risk to zero and if we are going to do anything in life then we have to accept that risk (even risk to life) MUST be balanced against cost. Since red tape increases the cost of safety systems it further follows that to maximise safety for a given "safety budget" a balance is needed between ensuring that the safety systems are sufficiently safe and not driving up the cost so much that safety improvements (whether fitting a new safety system to an existing plane or moving to a newer safer plane entirely) simply get rejected completely
For airlines the system seems to work pretty well. Even with the high cost imposed by safety compliance the new planes are a sufficient improvement in fuel efficiency and reliability to make buying them economically worthwhile. But for other types of flying afaict it is quite common to see decades old planes still in service either because newer planes are too expensive or because there is simply noone making a suitable plane anymore. For non-commerical flying it's also common to see homebuilt planes registered as "experimental". Does this combination of old planes and homebuilt planes really make flying in small planes safer than it would be if lighter regulations made new small planes more affordable and allowed a better choice of new small planes to be offered?
On pretty much every 32-bit unix-like system int, long, void* and time_t are all 32 bit. The only exception i'm aware of is the new "x32 abi".
On 32-bit windows int, long and void* are 32 bit. time_t is not considered a core system type but part of the C libraries (which on windows unlike unix like systems is not a core system library) so whether it is true depends on what compiler you are using. In visual studio 2005 the default size of time_t was changed from 32-bit to 64-bit.
It should be possible to add 64-bit time_t support to a 32-bit unix like operating system by taking an approach similar to that used for large file support. I haven't heard of anyone actually trying to do it though and applications would still at best need to be recompiled with the define to enable the new "large time support" mode and at worst may require considerably more work.