It's not just that it is cheap (though it isn't hugely expensive it can over the lifetime of a device become comparable to the cost of the device itself), it's also that the cost is distributed and hidden (not to mention difficult to calculate because determining average power draw requires information on the ammount of time that will be spent in each state). If the customer was fully aware of the cost of an appliances lifetime electricity use then differences in average power consumption should make a substantial difference to the price the customer is willing to pay at least in warmer climates (in cooler climates the difference is less significant because of displaced heating but still looking at my bills natural gas is arround 1/3 the price per kwh that electricity is).
Oh sure if want to put the effort into figuring out if my hardware is suitable (and if necessary upgrading) I can but for someone who doesn't take an interest in computer hardware anyway it's going to be a PITA. It doesn't help that most review sites seem to focus on the expensive overkill hardware rather than the reasonable priced and sufficient hardware.
P.S. I personally game on both consoles and PCs, which port of a given game I choose depends on many factors including what platform the game was designed for and what hardware I own at the time and how draconian the DRM situation is.
Were you shocked when your I-Pod didn't charge when it was plugged into a powered off PC?
No because USB ports on PCs were not intended for charging.
Are you shocked when your car battery drains when the engine is off?
No because there are good technical reasons for that. Further the car is designed for the battery to be charged while driving and the battery is large enough that having it run out is rare unless there is a fault with the car.
However I don't think either of these cases are relavent to the PS3. Lets consider the specific situation of the PS3.
* The PS3 already has provision for a standby mode where most of the hardware is powered off but some remains powered on. * The power needed to charge controllers is very small compared to the consoles total normal operating power and well within what it would be reasonable to expect a standby section of a PSU to provide. * The cable supplied with the controller is far shorter than the cable supplied with previous playstation controllers and while it is possible to play with the controller attached by the cable the short cable strongly discourages this (yes you can buy a longer cable but at least when the PS3 first came out long A-mini B cables were somewhat tricky to find). * The controller and the PS3 were designed and sold as a package, this clearly isn't a case of unfortunate interactions between products designed and sold seperately.
IMO given these facts not allowing controllers to charge while the system is in standby mode is clearly a design flaw.
There's a reason why they sell charging stations for....... PS3 controllers.
Yes sony fucked up the design of the console so third parties had to step in to fill the gap.
many of them target every major console and non-console platform these days.
Many AAA games come out on consoles first. They may or may not come out later on PC and when they do the PC versions that seem like an afterthought (poorly optimised, poor controller configuration options etc). Of course at the time of the console release there is usually no indication as to whether or not a PC version will come out later.
What exactly is the purpose of gaming consoles today?
When a game developer targets a console they will generally design their game to run well on that console. So if I buy a console early in it's generation I can be reasonablly confident that new games (with a few exceptions from shitty developers) will continue to run well on that console through it's lifecycle. Reviewers will be using the same hardware specs as players will so if a game plays well on the reviewers system it will play well for users too.
With PC gaming it's far more of a crapshoot, yes the graphics etc can be better than consoles but if your hardware specs aren't high enough the experience can be a lot worse. Furthermore neither CPU or GPU vendors label their products in a way that makes it easy for the customer to determine whether his CPU/GPU is better or worse than the one a reviewer used.
Oh and most PC games now have some form of online activation which often includes anti-resale measures (at the very least you don't know if the previous owner has posted the key somewhere online that could result in it being blacklisted). Console games can for the most part still be resold (some console games are starting to make DLC and/or online multiplayer access free for the original owner and chargable for subsequent owners but i've not yet seen a console game where the main single player game can't be resold).
I do wonder what apple are running in their datacenter anyway. They dropped the xserve and are refusing to license their server product for anyone elses hardware. If they are running mac pros that would certainly explain a low power density.
However while the channel offered by those ancient copper wires isn't great it is a dedicated channel to each property. So it may only be able to deliver a few megabits per second to each property but (provided the backhaul is available) it can deliver that to all properties at the same time. The only technical justification for any caps is backhaul congestion.
Wireless OTOH is great when only a few people are using it at once but benig a shared channel it breaks down quickly if you get too many heavy users in the same space. This can be mitigated to some degree by adding more cells but that gets expensive and there are limits to how tight you can pack the cells (you need the signal to drop off sufficiently between two cells that are using the same frequencies). So caps will remain low to discourage congestion of the expensive wireless channel.
As you say the only real soloution to increasing bandwidth availability is more fiber (whether that fiber is used to hook up new cell sites, reduce the length of DSL links with FTTC or provide a purely fiber soloution)
Not to be too pedantic, but it's not exactly true that noise limits your SNR. It's a ratio. You can increase SNR by either decreasing N or increasing S. I'll grant you that there are tradeoffs and practical limits to increasing S.
Indeed there are. For example radio transmitters leak somewhat to other parts of the spectrum (perfect filters are physically impossible) and most modern wireless networks rely on spacial re-use of spectrum. The result is that one user's signal is another user's noise and beyond a certain point increasing signal doesn't have any significant impact on SNR.
Also note that logrithm in shannons law. If you double your transmit power and noise stays the same you only get one more bit per symbol.
Not really, you can do SOME logical operations with them but to make a proper logic system you need devices that can invert signals (turn whatever you have chosen to represent a 1 into whatever you have chosen to represent a 0 and vice-versa) and devices that can boost signals (take a weaker incoming signal and produce a stronger output signal).
Diode logic as described in the wikipedia article can only provide and/or and worse it will SERIOUSLY weaken the signal when one of those is followed by the other.
Also note that electrical logic systems typically use voltage to represent stuff while optical systems typically use presense/absense of light. This means you can't really build an "and" gate out of optical diodes even though you can build one out of electrical diodes.
"Newell is ex-Microsoft employee" -- your link seems to support that claim "who was "put off" to destroy free software movement" -- you have not provided anything that supports this claim "by only releasing Steam for Windows". -- They have now released steam for the mac If they are really working for MS then would they do that? and why would they allow steam to work under wine? "There is no Linux version! And you know why?" -- I don't know for sure but I suspect what keeps most commercial game devs away from native linux development is a combination of platform fragmentation, small userbase and a userbase who are used to stuff being free (unlike apple users who are generally pretty rich, apple hardware is generally pretty expensive*) "Because Valve works with Microsoft to increase and keep Windows market share and so that no one would try better OS like Linux." -- if you want to convince people of this claim you will have to provide something more substantial than him being an ex MS employee.
*I don't deny that apple hardware is pretty nice and that finding exact equivalents in the PC world is often equally expensive. But I DO believe if you start from a set of requirements and find both a mac and a PC to fill them then the mac will generally be far more expensive. **Which you are presenting as fact but I suspect is just a crackpot theory of yours
What I meant to highlight is that pixel-based design is stupid, because you have no idea how big that pixel will be when displayed
You also have "no idea" how far the user is sitting from the screen or how good their eyesight is.
However what you DO know is that a pixel is the smallest thing you control. Further you know that on pretty much every contemporary computer setup a single pixel wide line can been seen. Given these facts pixel based design makes reasonable sense. If the user wants more stuff on their screen they need more pixels.
It's not like we coundn't use perfectly scalable vector graphics
For a truly resoloution independent layout to work well the size of the smallest thing in the layout must be significantly bigger than the size of a pixel. That works well when the individual pixels are tiny (e.g. on a laser printer) it doesn't work so well when your pixels are large enough to be seen individually and you want to take advantage of all of them (hell with subpixel rendering we are even trying to take advantage of PARTS of pixels).
You can get around this problem by doing a psuedo-resoloution-independent layout where large structures vary smoothly with the resoloution while small features (such as the width of lines) change in increments of whole pixels so you don't try to draw a 1.5 pixel wide line but it's harder to get this right than EITHER pixel based design or resolution independent design for high DPI devices.
Still even with such tricks there will be a minimum "assumed DPI" below which a given layout won't render acceptably because there simply aren't enough pixels to render the intricacies of the layout. Once you can no longer decrease the "assumed DPI" your only option to get more "real estate" is to increase the number of dots.
and define size based on a percentage of the screen.
You do want some way for the user to change the size of things to suit their appearance and preferences whether that is setting the "assumed DPI"* directly or setting values that are used to calculate the "assumed DPI"
* The "assumed DPI" being the scale factor used to translate from design units to pixels. This may or may not match the real DPI (and some users may not WANT it to match the real DPI)
It seems silly to blame Java when the entire purpose of Java is to serve as an execution platform for general purpose software.
That was one purpose of it but not the only one and not the one that has caused the controversy.
Another purpose from java was to provide a SANDBOXED execution platform for running untrusted software (such as applets from the web) while preventing it from damaging the users system. The problem is getting a sandbox like this right is hard and every so often a flaw is discovered that lets malicious code break out of the sandbox.
Firstly most programs at least in the windows world are designed in pixels. So if you want to fit more on the screen at once you need a higher resoloution. You can't easilly decouple resoloution and "real-estate" (in theory you can change the "assumed DPI" but stuff tends to break if you change it and text quickly becomes difficult to read if you take the assumed DPI below the default, modern fonts just aren't designed for displaying at such small sizes) so to get more "real-estate" you need to up the resoloution.
Secondly developers use 1024x768 as their baseline assumption. Not surprising since it has been the most common resolution for many years and (with the exception of the recent netbook craze that noone was expecting) lower resoloutions were becoming pretty rare.
What this means is
If you have less than 768 pixels of height you will find some software is basically unusable (mindstorms NXT for example) and even in some software that is generally usable (iTunes for example) you will have to find dialog boxes where you have to move them partially off the screen to see the whole box (and that is assuming you know how to move the top edge of a window off the screen). If you have exactly 768 pixels of height you will be mostly ok provided your setup follows their assumptions but if your setup doesn't follow their assumptions (for example you are running a multi-row taskbar or you are running some stuff in a maximised but not fullscreen VM which has it's own taskbard) you are back to having the problems that those with less than 768 vertical pixels have. If you have significantly more than 768 pixels of height you have some buffer left over for ways in which your setup violates their assumptions.
Also, the most-common size I recall seeing for PC monitors is 1980x1200, though maybe that's changing with 'HD' sizes for TVs taking over.
A few years ago 1920x1200 was pretty common but nowadays all the monitor manufacturers are pushing the smaller 1920x1080. You can still get 1920x1200 monitors but the starting price is over double and there is far less choice.
We do not need more ports and cables. Thunderbolt's claim is that by using it we'll have LESS cables. This is what they always say and it's always bullshit.
Not less cables total but certainly cables in different places.
With thunderbolt you can have a single cable from your laptop to your monitor then plug all your perhipherals into your monitor. Much less hassle than hooking up every perhipheral directly and hopefully without the hugely model specific characteristics of existing docking stations.
Still I agree thunderbolt is the new firewire, apple users are being pushed into using it (apple has not yet adopted USB3) and some other niches will use it because of it's strengths but I don't see it ever becoming maintstream.
Looking at the comparisons I've found, seems that Thunderbolt is likely to put a spanner in the works for USB 3.0 support.
The facts I see don't seem to support your assertion.
Firstly: thunderbolt gear has at least so far been EXTREMELY expensive. Yeah some of that is probablly the apple tax but other parts of it come down to design decisions and the sheer data rates of the protocol. It doesn't help that pretty much every device has to have two ports because most hosts only have one and it doubles as the monitor port.
Secondly: USB3 is already becoming pretty common. Afaict most current motherboards now have it.
Thirdly Panther Point (intel's latest chipset family) has integrated USB3. It does not have integrated thunderbolt (despite earlier rumours). So USB3 will be vitually free for computer vendors (just a few extra PCB traces and different sockets) while thunderbolt will require an expensive extra chip.
Fourthly we haven't seen any clear information on how thunderbolt will fit in with machines that don't use onboard graphics. Will the graphics functionality simply be disabled? will there be a way to route video from a graphics card to the thunderbolt cop?
Now don't get me wrong thunderbolt has it's uses, for example it can reduce the number of connectors a laptop user needs to hook up when they bring the machine to their desk to two (power and thunderbolt) while still providing plenty of bandwidth but I think it will remain a niche product.
The first mover has certain advantages in a market but they also have disadvantages. Namely that they get locked into descisions that made sense at the time but no longer do so. Afaict 10GBASE-T manages speeds comparable to thunderbolt without the need for active cables or fiber optics and over much longer distances, the hardware is expensive now but afaict it is coming down.
I'm pretty sure there will be a USB4, the main questions are 1: when and 2: will they decide to merely match thunderbolt or to surpass it.
Afaict the main thing people need out of "support" contracts from OS vendors at a late point in the operating system's lifecycle is security updates. If a functionality issue hasn't become apparent in the first year or so then it's probablly something you can live with but a security update that has just become known to your adversaries is a different matter. Sure you can TRY to isolate the networks but as stuxnet has shown that isn't exactly a bulletproof strategy if your adversary is determined enought.
Can you imagine the shitstorm if it was known that MS had backported the security patch that blocked the latest worm to XP but you couldn't get it without signing an agreement that cost $$$. Further do you think that shitstorm would be worth the money they could make out of selling extended support contracts.
I doubt many XP machines will still be working two years from now.
Computer hardware seems to be something of a lottery. Some of it fails after a few years but in my experiance if quality components were used it generally keeps going for much longer than that. Hard drives seem to be the most failure prone components but lukilly their interfaces have remained standard so replacing them is not generally a problem (though supplies of IDE drives do seem to be finally dwindling now, most suppliers still have them but usually only a handful of models).
Furthermore with vista being widely regarded as a failure and the general hesitation surrounding new versions of software plenty of people were buying computers with XP preinstalled right up until MS took that option away in late 2010 (you can still downgrade yourself but you can no longer buy off the shelf computers pre-downgraded unless they are old stock).
People understandably resent having to pay a big chunk of time and/or money on upgrading just to get security updates. I bet many won't bother and will just hope they can live without them just as many people continue to drive older cars that don't have modern safety freatures.
No and no linux distro i'm aware of is either. I think some of the big unix vendors may be comparable or better though.
WinXP was released in 2001 so that would be equivalent to OS 9.2 in the Apple world.
IMO support lifecycles should be measured not from when the OS came out but from when it's successor came out. Still MS is better than pretty much everyone else by that measure too.
If you really need to protect your conversations from tapping (whether by the government or by sufficiently motivated criminals) you shouldn't be using either skype or the regular phone network. You should be using an encyrpted system where you manage the keys yourself and where the datastream is constant bitrate (to prevent attacks based on packet size analysis)
AIUI using a VPN, proxy or similar service to access the internet is not in itself illegal. However if you use it to access sites that are supposed to be geographically restricted then you are likely to be breaking the terms of service of those sites. This then raises the quesiton of whether those terms of service form a legally enforceable contract.
For pay services you are also likely to need a US payment method.
Here in the UK (which is pressumablly what the GP meant by "people who use the word "petrol"") we buy petrol in litres but we measure our cars fuel economy in miles per (imperial) gallon.
Slashdot Mirror
Electricity is too cheap.
It's not just that it is cheap (though it isn't hugely expensive it can over the lifetime of a device become comparable to the cost of the device itself), it's also that the cost is distributed and hidden (not to mention difficult to calculate because determining average power draw requires information on the ammount of time that will be spent in each state). If the customer was fully aware of the cost of an appliances lifetime electricity use then differences in average power consumption should make a substantial difference to the price the customer is willing to pay at least in warmer climates (in cooler climates the difference is less significant because of displaced heating but still looking at my bills natural gas is arround 1/3 the price per kwh that electricity is).
Oh sure if want to put the effort into figuring out if my hardware is suitable (and if necessary upgrading) I can but for someone who doesn't take an interest in computer hardware anyway it's going to be a PITA. It doesn't help that most review sites seem to focus on the expensive overkill hardware rather than the reasonable priced and sufficient hardware.
P.S. I personally game on both consoles and PCs, which port of a given game I choose depends on many factors including what platform the game was designed for and what hardware I own at the time and how draconian the DRM situation is.
Were you shocked when your I-Pod didn't charge when it was plugged into a powered off PC?
No because USB ports on PCs were not intended for charging.
Are you shocked when your car battery drains when the engine is off?
No because there are good technical reasons for that. Further the car is designed for the battery to be charged while driving and the battery is large enough that having it run out is rare unless there is a fault with the car.
However I don't think either of these cases are relavent to the PS3. Lets consider the specific situation of the PS3.
* The PS3 already has provision for a standby mode where most of the hardware is powered off but some remains powered on.
* The power needed to charge controllers is very small compared to the consoles total normal operating power and well within what it would be reasonable to expect a standby section of a PSU to provide.
* The cable supplied with the controller is far shorter than the cable supplied with previous playstation controllers and while it is possible to play with the controller attached by the cable the short cable strongly discourages this (yes you can buy a longer cable but at least when the PS3 first came out long A-mini B cables were somewhat tricky to find).
* The controller and the PS3 were designed and sold as a package, this clearly isn't a case of unfortunate interactions between products designed and sold seperately.
IMO given these facts not allowing controllers to charge while the system is in standby mode is clearly a design flaw.
There's a reason why they sell charging stations for ....... PS3 controllers.
Yes sony fucked up the design of the console so third parties had to step in to fill the gap.
many of them target every major console and non-console platform these days.
Many AAA games come out on consoles first. They may or may not come out later on PC and when they do the PC versions that seem like an afterthought (poorly optimised, poor controller configuration options etc). Of course at the time of the console release there is usually no indication as to whether or not a PC version will come out later.
What exactly is the purpose of gaming consoles today?
When a game developer targets a console they will generally design their game to run well on that console. So if I buy a console early in it's generation I can be reasonablly confident that new games (with a few exceptions from shitty developers) will continue to run well on that console through it's lifecycle. Reviewers will be using the same hardware specs as players will so if a game plays well on the reviewers system it will play well for users too.
With PC gaming it's far more of a crapshoot, yes the graphics etc can be better than consoles but if your hardware specs aren't high enough the experience can be a lot worse. Furthermore neither CPU or GPU vendors label their products in a way that makes it easy for the customer to determine whether his CPU/GPU is better or worse than the one a reviewer used.
Oh and most PC games now have some form of online activation which often includes anti-resale measures (at the very least you don't know if the previous owner has posted the key somewhere online that could result in it being blacklisted). Console games can for the most part still be resold (some console games are starting to make DLC and/or online multiplayer access free for the original owner and chargable for subsequent owners but i've not yet seen a console game where the main single player game can't be resold).
IIRC in c++ accessing private stuff is as easy as "#define private public" before including the header that defines the class.
gnome3? unity?
I do wonder what apple are running in their datacenter anyway. They dropped the xserve and are refusing to license their server product for anyone elses hardware. If they are running mac pros that would certainly explain a low power density.
However while the channel offered by those ancient copper wires isn't great it is a dedicated channel to each property. So it may only be able to deliver a few megabits per second to each property but (provided the backhaul is available) it can deliver that to all properties at the same time. The only technical justification for any caps is backhaul congestion.
Wireless OTOH is great when only a few people are using it at once but benig a shared channel it breaks down quickly if you get too many heavy users in the same space. This can be mitigated to some degree by adding more cells but that gets expensive and there are limits to how tight you can pack the cells (you need the signal to drop off sufficiently between two cells that are using the same frequencies). So caps will remain low to discourage congestion of the expensive wireless channel.
As you say the only real soloution to increasing bandwidth availability is more fiber (whether that fiber is used to hook up new cell sites, reduce the length of DSL links with FTTC or provide a purely fiber soloution)
Not to be too pedantic, but it's not exactly true that noise limits your SNR. It's a ratio. You can increase SNR by either decreasing N or increasing S. I'll grant you that there are tradeoffs and practical limits to increasing S.
Indeed there are. For example radio transmitters leak somewhat to other parts of the spectrum (perfect filters are physically impossible) and most modern wireless networks rely on spacial re-use of spectrum. The result is that one user's signal is another user's noise and beyond a certain point increasing signal doesn't have any significant impact on SNR.
Also note that logrithm in shannons law. If you double your transmit power and noise stays the same you only get one more bit per symbol.
you can do logic with diodes:
Not really, you can do SOME logical operations with them but to make a proper logic system you need devices that can invert signals (turn whatever you have chosen to represent a 1 into whatever you have chosen to represent a 0 and vice-versa) and devices that can boost signals (take a weaker incoming signal and produce a stronger output signal).
Diode logic as described in the wikipedia article can only provide and/or and worse it will SERIOUSLY weaken the signal when one of those is followed by the other.
Also note that electrical logic systems typically use voltage to represent stuff while optical systems typically use presense/absense of light. This means you can't really build an "and" gate out of optical diodes even though you can build one out of electrical diodes.
And still everything I said is true.
Lets see
"Newell is ex-Microsoft employee" -- your link seems to support that claim
"who was "put off" to destroy free software movement" -- you have not provided anything that supports this claim
"by only releasing Steam for Windows". -- They have now released steam for the mac If they are really working for MS then would they do that? and why would they allow steam to work under wine?
"There is no Linux version! And you know why?" -- I don't know for sure but I suspect what keeps most commercial game devs away from native linux development is a combination of platform fragmentation, small userbase and a userbase who are used to stuff being free (unlike apple users who are generally pretty rich, apple hardware is generally pretty expensive*)
"Because Valve works with Microsoft to increase and keep Windows market share and so that no one would try better OS like Linux." -- if you want to convince people of this claim you will have to provide something more substantial than him being an ex MS employee.
*I don't deny that apple hardware is pretty nice and that finding exact equivalents in the PC world is often equally expensive. But I DO believe if you start from a set of requirements and find both a mac and a PC to fill them then the mac will generally be far more expensive.
**Which you are presenting as fact but I suspect is just a crackpot theory of yours
What I meant to highlight is that pixel-based design is stupid, because you have no idea how big that pixel will be when displayed
You also have "no idea" how far the user is sitting from the screen or how good their eyesight is.
However what you DO know is that a pixel is the smallest thing you control. Further you know that on pretty much every contemporary computer setup a single pixel wide line can been seen. Given these facts pixel based design makes reasonable sense. If the user wants more stuff on their screen they need more pixels.
It's not like we coundn't use perfectly scalable vector graphics
For a truly resoloution independent layout to work well the size of the smallest thing in the layout must be significantly bigger than the size of a pixel. That works
well when the individual pixels are tiny (e.g. on a laser printer) it doesn't work so well when your pixels are large enough to be seen individually and you want to take advantage of all of them (hell with subpixel rendering we are even trying to take advantage of PARTS of pixels).
You can get around this problem by doing a psuedo-resoloution-independent layout where large structures vary smoothly with the resoloution while small features (such as the width of lines) change in increments of whole pixels so you don't try to draw a 1.5 pixel wide line but it's harder to get this right than EITHER pixel based design or resolution independent design for high DPI devices.
Still even with such tricks there will be a minimum "assumed DPI" below which a given layout won't render acceptably because there simply aren't enough pixels to render the intricacies of the layout. Once you can no longer decrease the "assumed DPI" your only option to get more "real estate" is to increase the number of dots.
and define size based on a percentage of the screen.
You do want some way for the user to change the size of things to suit their appearance and preferences whether that is setting the "assumed DPI"* directly or setting values that are used to calculate the "assumed DPI"
* The "assumed DPI" being the scale factor used to translate from design units to pixels. This may or may not match the real DPI (and some users may not WANT it to match the real DPI)
It seems silly to blame Java when the entire purpose of Java is to serve as an execution platform for general purpose software.
That was one purpose of it but not the only one and not the one that has caused the controversy.
Another purpose from java was to provide a SANDBOXED execution platform for running untrusted software (such as applets from the web) while preventing it from damaging the users system. The problem is getting a sandbox like this right is hard and every so often a flaw is discovered that lets malicious code break out of the sandbox.
Why?
Firstly most programs at least in the windows world are designed in pixels. So if you want to fit more on the screen at once you need a higher resoloution. You can't easilly decouple resoloution and "real-estate" (in theory you can change the "assumed DPI" but stuff tends to break if you change it and text quickly becomes difficult to read if you take the assumed DPI below the default, modern fonts just aren't designed for displaying at such small sizes) so to get more "real-estate" you need to up the resoloution.
Secondly developers use 1024x768 as their baseline assumption. Not surprising since it has been the most common resolution for many years and (with the exception of the recent netbook craze that noone was expecting) lower resoloutions were becoming pretty rare.
What this means is
If you have less than 768 pixels of height you will find some software is basically unusable (mindstorms NXT for example) and even in some software that is generally usable (iTunes for example) you will have to find dialog boxes where you have to move them partially off the screen to see the whole box (and that is assuming you know how to move the top edge of a window off the screen).
If you have exactly 768 pixels of height you will be mostly ok provided your setup follows their assumptions but if your setup doesn't follow their assumptions (for example you are running a multi-row taskbar or you are running some stuff in a maximised but not fullscreen VM which has it's own taskbard) you are back to having the problems that those with less than 768 vertical pixels have.
If you have significantly more than 768 pixels of height you have some buffer left over for ways in which your setup violates their assumptions.
Also, the most-common size I recall seeing for PC monitors is 1980x1200, though maybe that's changing with 'HD' sizes for TVs taking over.
A few years ago 1920x1200 was pretty common but nowadays all the monitor manufacturers are pushing the smaller 1920x1080. You can still get 1920x1200 monitors but the starting price is over double and there is far less choice.
How is that docking station use working out for all those thunderbolt-equipped macs?
The "thunderbolt cinema display" is basically a monitor with integrated docking station.
We do not need more ports and cables. Thunderbolt's claim is that by using it we'll have LESS cables. This is what they always say and it's always bullshit.
Not less cables total but certainly cables in different places.
With thunderbolt you can have a single cable from your laptop to your monitor then plug all your perhipherals into your monitor. Much less hassle than hooking up every perhipheral directly and hopefully without the hugely model specific characteristics of existing docking stations.
Still I agree thunderbolt is the new firewire, apple users are being pushed into using it (apple has not yet adopted USB3) and some other niches will use it because of it's strengths but I don't see it ever becoming maintstream.
Looking at the comparisons I've found, seems that Thunderbolt is likely to put a spanner in the works for USB 3.0 support.
The facts I see don't seem to support your assertion.
Firstly: thunderbolt gear has at least so far been EXTREMELY expensive. Yeah some of that is probablly the apple tax but other parts of it come down to design decisions and the sheer data rates of the protocol. It doesn't help that pretty much every device has to have two ports because most hosts only have one and it doubles as the monitor port.
Secondly: USB3 is already becoming pretty common. Afaict most current motherboards now have it.
Thirdly Panther Point (intel's latest chipset family) has integrated USB3. It does not have integrated thunderbolt (despite earlier rumours). So USB3 will be vitually free for computer vendors (just a few extra PCB traces and different sockets) while thunderbolt will require an expensive extra chip.
Fourthly we haven't seen any clear information on how thunderbolt will fit in with machines that don't use onboard graphics. Will the graphics functionality simply be disabled? will there be a way to route video from a graphics card to the thunderbolt cop?
Now don't get me wrong thunderbolt has it's uses, for example it can reduce the number of connectors a laptop user needs to hook up when they bring the machine to their desk to two (power and thunderbolt) while still providing plenty of bandwidth but I think it will remain a niche product.
The first mover has certain advantages in a market but they also have disadvantages. Namely that they get locked into descisions that made sense at the time but no longer do so. Afaict 10GBASE-T manages speeds comparable to thunderbolt without the need for active cables or fiber optics and over much longer distances, the hardware is expensive now but afaict it is coming down.
I'm pretty sure there will be a USB4, the main questions are 1: when and 2: will they decide to merely match thunderbolt or to surpass it.
Afaict the main thing people need out of "support" contracts from OS vendors at a late point in the operating system's lifecycle is security updates. If a functionality issue hasn't become apparent in the first year or so then it's probablly something you can live with but a security update that has just become known to your adversaries is a different matter. Sure you can TRY to isolate the networks but as stuxnet has shown that isn't exactly a bulletproof strategy if your adversary is determined enought.
Can you imagine the shitstorm if it was known that MS had backported the security patch that blocked the latest worm to XP but you couldn't get it without signing an agreement that cost $$$. Further do you think that shitstorm would be worth the money they could make out of selling extended support contracts.
I doubt many XP machines will still be working two years from now.
Computer hardware seems to be something of a lottery. Some of it fails after a few years but in my experiance if quality components were used it generally keeps going for much longer than that. Hard drives seem to be the most failure prone components but lukilly their interfaces have remained standard so replacing them is not generally a problem (though supplies of IDE drives do seem to be finally dwindling now, most suppliers still have them but usually only a handful of models).
Furthermore with vista being widely regarded as a failure and the general hesitation surrounding new versions of software plenty of people were buying computers with XP preinstalled right up until MS took that option away in late 2010 (you can still downgrade yourself but you can no longer buy off the shelf computers pre-downgraded unless they are old stock).
People understandably resent having to pay a big chunk of time and/or money on upgrading just to get security updates. I bet many won't bother and will just hope they can live without them just as many people continue to drive older cars that don't have modern safety freatures.
Is Apple any better?
No and no linux distro i'm aware of is either. I think some of the big unix vendors may be comparable or better though.
WinXP was released in 2001 so that would be equivalent to OS 9.2 in the Apple world.
IMO support lifecycles should be measured not from when the OS came out but from when it's successor came out. Still MS is better than pretty much everyone else by that measure too.
Sure they can.
If you really need to protect your conversations from tapping (whether by the government or by sufficiently motivated criminals) you shouldn't be using either skype or the regular phone network. You should be using an encyrpted system where you manage the keys yourself and where the datastream is constant bitrate (to prevent attacks based on packet size analysis)
AIUI using a VPN, proxy or similar service to access the internet is not in itself illegal. However if you use it to access sites that are supposed to be geographically restricted then you are likely to be breaking the terms of service of those sites. This then raises the quesiton of whether those terms of service form a legally enforceable contract.
For pay services you are also likely to need a US payment method.
Here in the UK (which is pressumablly what the GP meant by "people who use the word "petrol"") we buy petrol in litres but we measure our cars fuel economy in miles per (imperial) gallon.