If you learn to type on Dvorak, even if it is better/faster/etc, if you want to use another PC it most likely has a Qwerty keyboard, so you either have to change the layout (may not always be possible) or use the keyboard as is, probably less efficiently than you would if you had learned to use Qwerty from the start.
If you use a measurement system that is different from anyone else, you will have problems communicating. Now, while the US system is standard in the US (so you can talk to some people without problems), as soon as you want to talk to someone from a different country (which uses metric) someone (maybe both) has to google what the measurements are. If you say that your car goes x mph and does y mpg I have to go to google or wolfram alpha to convert those numbers to km/h and L/100km. Then I will give you some measurements in metric and you will have to convert them. It makes the conversation slower and harder.
Interesting - I like laptop-style keyboards enough that I bought Logitech UltraX Flat when it was new and expensive (paid something like 24EUR for it). Before it I tried a few keyboards but none were as good (to me) as this. I do not type a lot, but a laptop style keyboard allows me to type faster (I sometimes look at the keys and sometimes don't, usually after some typing while looking I can stop looking and type just as accurately).The only problem with the UltraX Flat is that since the keys are low, it does not take a lot of dust (etc) to make the key no longer work and I have to clean it, especially the space bar.
Those kinds of failures can occur but with a sealed device, if you build it using good parts it is much less likely than with the average system.
Unless the component quality depend on whether the system is sealed (which I doubt), a non-sealed system with good components will last just as long without repairs, but will be easier to repair.
As for the battery, my guess is that the phone runs much cooler than the laptop and heat shortens the life of the battery. It also most likely depends on the power draw as the internal resistance of the battery increases when it ages, so, for example (not the same chemistry but still) a battery might be too old to use in an UPS but good enough for things that do not draw as much current. Same stuff with the laptop - it uses a lot of power compared to the phone so the effects of the increasing internal resistance are more obvious (higher resistance = lower voltage under load = laptops thinks that the battery is discharged). Cell phones use little power so the internal resistance can be higher without obvious effects (when it gets really high then weird stuff starts happening, like the indicator going nuts - almost empty one minute, full the next).
I can rapidly see the day coming when that will have to be replaced.
Why? The amp does not wear out and it does not use some proprietary protocols, inputs and outputs are analog, as for the rest of the system it depends on what you have. I have working reel-to-reel and cassette decks, record player,CD/LD player and VHS VCR, I use them all, since my music is in all those formats and I am not going to record it to a computer, since I see no problem with using cassettes or reels or records.
Still, SSDs do fail (it seems that they are more durable, that is, they can survive more physical abuse than hard drives, but still wear out or fail in other ways), batteries also fail (the battery of my laptop lasts something like 30 minutes now, compared to 3 hours 30 minutes when it was new, but instead of replacing it, in the rare event that I need to use the laptop away from 220V source, I bring an external 7Ah lead-acid battery (or two if I expect to be there longer), which is heavier but much cheaper).
More RAM is not just needed because the software uses more memory, but also bacause I want to have a lot of apps running at once (say Firefox, Mathcad, Word, Excel, mpc-hc), and while they eah do not use a lot of memory, all of them combined do. As I understand the iPad only allows one app at a time so I guess it is not a problem for it, but then again I could just keep the 512MB RAM and also be able to use one app at a time.
I like devices with user-replaceable parts (to me it means a part that does not require more advanced tools than a regular (de)soldering iron, so most SMD parts are out), since all devices sooner or later break (even if it's just a bad capacitor), I can fix them cheaper than buying a new device. This is part of the reason why I like old technology - those devices were built like tanks and when they develop a problem (as any device will, maybe after 40 years but still) it is much easier to fix.
When I bought my laptop in 2005 it had 512MB RAM and 80GB hard drive. Some time after I added a 1GB RAM stick (making total to be 1280MB). Last month I replaced the hard drive with a 160GB one(the only one I could find with an IDE interface), though the reason was not the lack of space, but the fact that the old hard drive was dying. The new drive is faster and a bit cooler though.
I also replaced the wireless card with one that was more suitable for wardriving, because the original one was not natively supported by Linux.
I do not understand why everybody has a problem with the stylus. I used a Psion Series 5 for a long time and it was great. By modern standards it is too slow to browse the web and so on, but I would really love to have an updated Psion with the same form factor but faster CPU and more storage (I am sure that it would be possible to do that in 2011, compared with what was possible in 1997). And maybe actually running normal Windows on it.
The fact that the Psion could work on AA batteries was also great - if I forget to recharge and the batteries run out, I can just go and buy alkaline ones, no need to look for a 220V outlet and wait a few hours for it to recharge. The form factor of the Psion is also good - small enough to fit in a (big) pocket, but big enough to have a keyboard that is big enough for my fingers.
XP is "Good Enough". My current PC is "Good Enough". Reinstalling Windows is a PITA, especially if you, like me, use a lot of small apps.
Put all of those together and the result is that I do not want to install a new version of Windows even if it worked OK on my current PC (it should) and I could get the new version for free (I can).
20 years ago it was different. Hardware was quite limited and in turn, the OSs were quite limited. I can do a lot of things with XP thawt I cannot do with 3.0 or DOS. However, all of the features that 7 has (that XP does not have) are not essential - prettier desktop (something that I would turn off right after autorun), UAC (probably would turn that off too, depending on how often it would annoy me) non-sucky 64bit version (could use 2003, also 3.25GB RAM is enough for me for now) different UI (so I would have to install ClassicShell) IE9 (I don't use IE, other than to install updates) DX11 (most games still support 9, so no problem there)
And now you have three devices with batteries that you need to charge and worry about running out. If I connect a mouse and/or keyboard to a laptop, there still is one power source to worry about.
Also, for a tablet+keyboard+mouse you need a desk, you can put a laptop on your, well, lap and only need a chair to sit on.
Yes, DST is perverted and should be eliminated. While for now for me it is not inconvenient other than having to reset a bunch of clocks, people who go to bed and get up at the same time every day are really inconvenienced because they either get up an hour early (thus missing some sleep) or have to get up an hour earlier than they are used to.
If you only communicate with those in your time zone, then wouldn't the current system be better? I mean why go trough all that change and the result would be the same. So what if the clock shows the same time everywhere in the world (or not) if the clocks of your time zone are the only ones that matter to you?
Sure, I'll know when the sun is over my head and what the clock says then. Now, the question is will you know whether it is day or night _for me_ when you want to call me. You will know what my clock says, but how do you know that 02:00 is the middle of the night for me and you should not call me.
When you want to call me now you go something like this: "my local time is 16:00, my time zone is GMT+9, his time zone is GMT+2, so his clock shows 09:00, he should be in the office, so it's OK to call". With the new system you would go something like this: "the time for both of us is 07:00, his work starts at 06:00, so he should be in the office, so it's OK to call".
Of course it is, unless you expect people to start work at, say, 10am irrespective of whether it will be light or dark during their work. This won't work. People synchronize themselves to the sun, even if their job does not require them being outside.
So, you would still need to know if the person you are going to call is asleep or not.
For some reason, most people sleep at night and are active (work etc) during the day. So, for you to say "meet me at 13:45", you still would have to know whether it will be day where I live. So, as you see, you do need to know the position of the sun, even if only to know whether other people are active or not.
Now you have to look up what your 13:45 is in my time zone. Under the new system you would have to look up whether it will be business hours at 13:45 in my timezone (maybe it's midnight here).
How your computer based system would work for, say, phone conversations? When I say "meet me at 6pm", what does it mean? 6pm UTC, 6pm local time or 6pm my personal time? How does the other person find out what time does that correspond to in his personal time zone?
I'd say keep the time zones how they are. The conversions are already done and only bother those who have frequent contacts far away. Local time works OK for local stuff and it also happens to correspond to the position of the sun. Even if you got rid of the time zones and managed to teach everyone that thay should begin/end work, go to speed/get up x hours earlier/later, it would still be the same as now. 6pm now would mean day in one country and night in another, so you would still need to know the time difference.
Officially, you aren't even supposed to say "1/3 meter," but rather "333 milliliters." For everyday uses, such as cooking, it is much more natural to use fractions.
Well, 333mm or 330mm, depending on what precision you need. As for the "more natural", I guess it depends on tradition, decimal is more natural to me - 0.25L or 250ml, but not 1/4 L.
The 2-liter bottle seems to have become "natural," but if you want to buy a single drink, it's easier to say "a pint" or even "a 12-ounce cup" rather than "400 milliliters."
Is the "pint" some absolute unit or is it an arbitrary one (~473.2mL). Could you drink 0.5L of beer? That's how you can get beer where I live (completely metric country) - 0.3L, 0.5L, 1L (if we are talking about a bar). I have no trouble asking for "one beer, zero five" (literally translated) or a "one big beer" if the bar offers only two options (very common, 0.3 and 0.5). The bottles are 0.5L, and cans are 0.3 or 0.5.
Stronger drinks are often measured in "grams", or at least when someone says 100g of vodka they mean 100mL (because 100mL of water would weigh 100g), gram is shorter than mililiter.
Currently I cannot think of any "folk" units in use (in my country) that do not correspond directly to some of the metric units.
The thing is that time of day is not the same everywhere.
What I mean is: any measurement system can be used anywhere, my metric tape measure would work just as well in the US as it does in my country. A day is 24 hours long everywhere, so a 24 hour clock would work anywhere without any problem at all. Now, no measurement system is going to change the fact that when it's mid-day in one country it is midnight in another. You can deal with this two ways:
1. (current) Assign midnight as 00:00, mid-day as 12:00 and split the world in time zones, so that 12:00 in each time zone means mid-day. Make one time zone the primary, tell all the others by their difference from the primary. To find out what time it is in another country, you need to add or subtract some time based on the difference between time zones. 2. (proposed). Synchronize all clocks to the primary time zone. Now all the clocks show the same time, so mid-day in one country is at 12:00, in some other at 9:00 and in another at 00:00. Create a table of countries and the time of mid-day (or midnight) in each of them. You know what time it is in another country, but you still do not know if it's, say, business hours or not, so you now have to calculate the difference between your time-of-midnight and their time-of-midnight, then add or subtract it from the time on your clock to know what actual time, compared to you, is there.
So, you have a system that is just as (if not more) complex, local time no longer corresponds to the position of the sun (without the calculations, anyway) and you gained absolutely nothing.
The OSI model is still useful to know in which order you want to do stuff.
For example, take the application data, if you need to, convert it to something that the recipient can read (XML, some encoding), then encrypt it and/or use whatever session management protocols you want, after that put it in a transport protocol, then a network protocol and pass it down to data link which will send it over a physical connection.
The fact that you can arbitrarily nest protocols inside one another is the result of the fact that each protocol cares only about its layer and does not care what is above and below it. I think that was the point of the OSI model.
Well, in that case ICMP is a transport layer protocol, I mean you can stuff arbitrary data inside an echo request packet, so you can use it as a way to send HTTP requests (and the recipient replies with the same data, so you can check whether it arrived correctly).
Well, another example - I take an HTTP packet and send it straight over the wire (let's say a serial or parallel port of a PC), now it only has two layers - physical and application, all others are null. Or if you want a network, try an I2C bus, it has adressing (and layer 2), but no layer 3 or 4 (you can use it if you want though).
HTTP can also drop to a lower layer if it is being used to transfer some other protocol, for example BitTorrent..
Well, you can imagine a "null" layer that does nothing, just passes the data unmodified to the next layer.
For example, HTTPS would be HTTP over SSL, SSL wouls be level 6 (presentation). If you use HTTP without SSL then level 6 is empty or uses the "null" protocol.
ICMP is part of IP, while you could say that the ICMP packet is inside an IP packet it is easier to imagine ICMP as just a part of IP, because it is used that way (for example, to signal that some other packet could not be delivered).
Just because I can send the HTTP packet inside an Ethernet frame (without IP or TCP), does not mean that the model is broken, it's just that "null" is a valid protocol.
You can mount a partition as a folder in a NTFS partition (like it is on Linux).
How?
Computer Management ->Disk Management -> Right click on the partition you want to mount ->Change drive letters and paths->Add->Mount in the following empty NTFS folder
When you install Windows you can specify that the user profiles be somewhere else other than "C:\Documents and Settings"
No you can't, or at least my install process didn't give that option.
No, but nLite lets you customize this, as well as other settings (for example my temp directory is C:\Temp and not c:\Documents and Settings\user name\Local Settings\Temp)
You can move the profile of a single user to another folder/drive
1. You can mount a partition as a folder in a NTFS partition (like it is on Linux). 2. When you install Windows you can specify that the user profiles be somewhere else other than "C:\Documents and Settings" 3. You can move the Documents and Settings folder after installation, but it's difficult. 4. You can move the profile of a single user to another folder/drive 5. On a single user systems (and a lot of PCs now are single user) there is no need to do any of that - just save the files to another drive.
Does Windows support other hard drives with GPT under normal BIOS? If so, then I do not see the problem. I would not need a 3TB drive as a system drive, I'd use a smaller one, either a fast HDD or a SSD.
I still do not get why current BIOSes cannot be made compatible with larger drives.
Now, granted, the MBR has the 2TB limit, but BIOS does not require that MBR. What I mean is, bios just loads the first (zeroth?) sector of the hard drive to memory and jumps to that location in memory. It does not care what is in it. From then the CPU executes whatever instructions were in that first sector.
Also, while boot code and the partition table are currently on the same sector, there is no particular need why they have to be so and why the partition table has to be in that format. As long as the OS and the boot code inside the first sector supports it the partition table can be whatever you want.
And yet, not all applications need drives that fast. For example, my movie collection does not need 100MB/s read speed. I do not need a 64GB SSD for the movie collection. If my collection is small enough to fit in 64GB then I could buy a used 80GB hard drive and save lots of money.
Optical requires the receiving device to have a power source, also, AFAIK the cables are fragile, you can abuse a regular audio cable quite a lot and it will still work, I suspect the optical cable has a minimum bend radius larger than a few mm.
As for the new interface - 3.5mm TRS connectors are common, the new interface would be incompatible with the current one and would not provide a lot of benefit - 3.5mm plugs are usually used to connect headphones and headphones do not care whether the ground is connected first or not. At least once Apple is doing something that would be compatible with other devices.
Slashdot Mirror
Compatibility.
If you learn to type on Dvorak, even if it is better/faster/etc, if you want to use another PC it most likely has a Qwerty keyboard, so you either have to change the layout (may not always be possible) or use the keyboard as is, probably less efficiently than you would if you had learned to use Qwerty from the start.
If you use a measurement system that is different from anyone else, you will have problems communicating. Now, while the US system is standard in the US (so you can talk to some people without problems), as soon as you want to talk to someone from a different country (which uses metric) someone (maybe both) has to google what the measurements are. If you say that your car goes x mph and does y mpg I have to go to google or wolfram alpha to convert those numbers to km/h and L/100km. Then I will give you some measurements in metric and you will have to convert them. It makes the conversation slower and harder.
Interesting - I like laptop-style keyboards enough that I bought Logitech UltraX Flat when it was new and expensive (paid something like 24EUR for it). Before it I tried a few keyboards but none were as good (to me) as this. I do not type a lot, but a laptop style keyboard allows me to type faster (I sometimes look at the keys and sometimes don't, usually after some typing while looking I can stop looking and type just as accurately).The only problem with the UltraX Flat is that since the keys are low, it does not take a lot of dust (etc) to make the key no longer work and I have to clean it, especially the space bar.
Those kinds of failures can occur but with a sealed device, if you build it using good parts it is much less likely than with the average system.
Unless the component quality depend on whether the system is sealed (which I doubt), a non-sealed system with good components will last just as long without repairs, but will be easier to repair.
As for the battery, my guess is that the phone runs much cooler than the laptop and heat shortens the life of the battery. It also most likely depends on the power draw as the internal resistance of the battery increases when it ages, so, for example (not the same chemistry but still) a battery might be too old to use in an UPS but good enough for things that do not draw as much current. Same stuff with the laptop - it uses a lot of power compared to the phone so the effects of the increasing internal resistance are more obvious (higher resistance = lower voltage under load = laptops thinks that the battery is discharged). Cell phones use little power so the internal resistance can be higher without obvious effects (when it gets really high then weird stuff starts happening, like the indicator going nuts - almost empty one minute, full the next).
I can rapidly see the day coming when that will have to be replaced.
Why? The amp does not wear out and it does not use some proprietary protocols, inputs and outputs are analog, as for the rest of the system it depends on what you have. I have working reel-to-reel and cassette decks, record player,CD/LD player and VHS VCR, I use them all, since my music is in all those formats and I am not going to record it to a computer, since I see no problem with using cassettes or reels or records.
Still, SSDs do fail (it seems that they are more durable, that is, they can survive more physical abuse than hard drives, but still wear out or fail in other ways), batteries also fail (the battery of my laptop lasts something like 30 minutes now, compared to 3 hours 30 minutes when it was new, but instead of replacing it, in the rare event that I need to use the laptop away from 220V source, I bring an external 7Ah lead-acid battery (or two if I expect to be there longer), which is heavier but much cheaper).
More RAM is not just needed because the software uses more memory, but also bacause I want to have a lot of apps running at once (say Firefox, Mathcad, Word, Excel, mpc-hc), and while they eah do not use a lot of memory, all of them combined do. As I understand the iPad only allows one app at a time so I guess it is not a problem for it, but then again I could just keep the 512MB RAM and also be able to use one app at a time.
I like devices with user-replaceable parts (to me it means a part that does not require more advanced tools than a regular (de)soldering iron, so most SMD parts are out), since all devices sooner or later break (even if it's just a bad capacitor), I can fix them cheaper than buying a new device. This is part of the reason why I like old technology - those devices were built like tanks and when they develop a problem (as any device will, maybe after 40 years but still) it is much easier to fix.
When I bought my laptop in 2005 it had 512MB RAM and 80GB hard drive. Some time after I added a 1GB RAM stick (making total to be 1280MB). Last month I replaced the hard drive with a 160GB one(the only one I could find with an IDE interface), though the reason was not the lack of space, but the fact that the old hard drive was dying. The new drive is faster and a bit cooler though.
I also replaced the wireless card with one that was more suitable for wardriving, because the original one was not natively supported by Linux.
I do not understand why everybody has a problem with the stylus. I used a Psion Series 5 for a long time and it was great. By modern standards it is too slow to browse the web and so on, but I would really love to have an updated Psion with the same form factor but faster CPU and more storage (I am sure that it would be possible to do that in 2011, compared with what was possible in 1997). And maybe actually running normal Windows on it.
The fact that the Psion could work on AA batteries was also great - if I forget to recharge and the batteries run out, I can just go and buy alkaline ones, no need to look for a 220V outlet and wait a few hours for it to recharge. The form factor of the Psion is also good - small enough to fit in a (big) pocket, but big enough to have a keyboard that is big enough for my fingers.
XP is "Good Enough". My current PC is "Good Enough". Reinstalling Windows is a PITA, especially if you, like me, use a lot of small apps.
Put all of those together and the result is that I do not want to install a new version of Windows even if it worked OK on my current PC (it should) and I could get the new version for free (I can).
20 years ago it was different. Hardware was quite limited and in turn, the OSs were quite limited. I can do a lot of things with XP thawt I cannot do with 3.0 or DOS. However, all of the features that 7 has (that XP does not have) are not essential -
prettier desktop (something that I would turn off right after autorun),
UAC (probably would turn that off too, depending on how often it would annoy me)
non-sucky 64bit version (could use 2003, also 3.25GB RAM is enough for me for now)
different UI (so I would have to install ClassicShell)
IE9 (I don't use IE, other than to install updates)
DX11 (most games still support 9, so no problem there)
Did I leave anything out?
And now you have three devices with batteries that you need to charge and worry about running out. If I connect a mouse and/or keyboard to a laptop, there still is one power source to worry about.
Also, for a tablet+keyboard+mouse you need a desk, you can put a laptop on your, well, lap and only need a chair to sit on.
Yes, DST is perverted and should be eliminated. While for now for me it is not inconvenient other than having to reset a bunch of clocks, people who go to bed and get up at the same time every day are really inconvenienced because they either get up an hour early (thus missing some sleep) or have to get up an hour earlier than they are used to.
If you only communicate with those in your time zone, then wouldn't the current system be better? I mean why go trough all that change and the result would be the same. So what if the clock shows the same time everywhere in the world (or not) if the clocks of your time zone are the only ones that matter to you?
Sure, I'll know when the sun is over my head and what the clock says then. Now, the question is will you know whether it is day or night _for me_ when you want to call me. You will know what my clock says, but how do you know that 02:00 is the middle of the night for me and you should not call me.
When you want to call me now you go something like this: "my local time is 16:00, my time zone is GMT+9, his time zone is GMT+2, so his clock shows 09:00, he should be in the office, so it's OK to call". With the new system you would go something like this: "the time for both of us is 07:00, his work starts at 06:00, so he should be in the office, so it's OK to call".
Of course it is, unless you expect people to start work at, say, 10am irrespective of whether it will be light or dark during their work. This won't work. People synchronize themselves to the sun, even if their job does not require them being outside.
So, you would still need to know if the person you are going to call is asleep or not.
For some reason, most people sleep at night and are active (work etc) during the day. So, for you to say "meet me at 13:45", you still would have to know whether it will be day where I live. So, as you see, you do need to know the position of the sun, even if only to know whether other people are active or not.
Now you have to look up what your 13:45 is in my time zone. Under the new system you would have to look up whether it will be business hours at 13:45 in my timezone (maybe it's midnight here).
How your computer based system would work for, say, phone conversations? When I say "meet me at 6pm", what does it mean? 6pm UTC, 6pm local time or 6pm my personal time? How does the other person find out what time does that correspond to in his personal time zone?
I'd say keep the time zones how they are. The conversions are already done and only bother those who have frequent contacts far away. Local time works OK for local stuff and it also happens to correspond to the position of the sun. Even if you got rid of the time zones and managed to teach everyone that thay should begin/end work, go to speed/get up x hours earlier/later, it would still be the same as now. 6pm now would mean day in one country and night in another, so you would still need to know the time difference.
Officially, you aren't even supposed to say "1/3 meter," but rather "333 milliliters." For everyday uses, such as cooking, it is much more natural to use fractions.
Well, 333mm or 330mm, depending on what precision you need. As for the "more natural", I guess it depends on tradition, decimal is more natural to me - 0.25L or 250ml, but not 1/4 L.
The 2-liter bottle seems to have become "natural," but if you want to buy a single drink, it's easier to say "a pint" or even "a 12-ounce cup" rather than "400 milliliters."
Is the "pint" some absolute unit or is it an arbitrary one (~473.2mL). Could you drink 0.5L of beer? That's how you can get beer where I live (completely metric country) - 0.3L, 0.5L, 1L (if we are talking about a bar). I have no trouble asking for "one beer, zero five" (literally translated) or a "one big beer" if the bar offers only two options (very common, 0.3 and 0.5). The bottles are 0.5L, and cans are 0.3 or 0.5.
Stronger drinks are often measured in "grams", or at least when someone says 100g of vodka they mean 100mL (because 100mL of water would weigh 100g), gram is shorter than mililiter.
Currently I cannot think of any "folk" units in use (in my country) that do not correspond directly to some of the metric units.
The thing is that time of day is not the same everywhere.
What I mean is: any measurement system can be used anywhere, my metric tape measure would work just as well in the US as it does in my country. A day is 24 hours long everywhere, so a 24 hour clock would work anywhere without any problem at all. Now, no measurement system is going to change the fact that when it's mid-day in one country it is midnight in another. You can deal with this two ways:
1. (current) Assign midnight as 00:00, mid-day as 12:00 and split the world in time zones, so that 12:00 in each time zone means mid-day. Make one time zone the primary, tell all the others by their difference from the primary. To find out what time it is in another country, you need to add or subtract some time based on the difference between time zones.
2. (proposed). Synchronize all clocks to the primary time zone. Now all the clocks show the same time, so mid-day in one country is at 12:00, in some other at 9:00 and in another at 00:00. Create a table of countries and the time of mid-day (or midnight) in each of them. You know what time it is in another country, but you still do not know if it's, say, business hours or not, so you now have to calculate the difference between your time-of-midnight and their time-of-midnight, then add or subtract it from the time on your clock to know what actual time, compared to you, is there.
So, you have a system that is just as (if not more) complex, local time no longer corresponds to the position of the sun (without the calculations, anyway) and you gained absolutely nothing.
The OSI model is still useful to know in which order you want to do stuff.
For example, take the application data, if you need to, convert it to something that the recipient can read (XML, some encoding), then encrypt it and/or use whatever session management protocols you want, after that put it in a transport protocol, then a network protocol and pass it down to data link which will send it over a physical connection.
The fact that you can arbitrarily nest protocols inside one another is the result of the fact that each protocol cares only about its layer and does not care what is above and below it. I think that was the point of the OSI model.
Well, in that case ICMP is a transport layer protocol, I mean you can stuff arbitrary data inside an echo request packet, so you can use it as a way to send HTTP requests (and the recipient replies with the same data, so you can check whether it arrived correctly).
Well, another example - I take an HTTP packet and send it straight over the wire (let's say a serial or parallel port of a PC), now it only has two layers - physical and application, all others are null. Or if you want a network, try an I2C bus, it has adressing (and layer 2), but no layer 3 or 4 (you can use it if you want though).
HTTP can also drop to a lower layer if it is being used to transfer some other protocol, for example BitTorrent..
Well, you can imagine a "null" layer that does nothing, just passes the data unmodified to the next layer.
For example, HTTPS would be HTTP over SSL, SSL wouls be level 6 (presentation). If you use HTTP without SSL then level 6 is empty or uses the "null" protocol.
ICMP is part of IP, while you could say that the ICMP packet is inside an IP packet it is easier to imagine ICMP as just a part of IP, because it is used that way (for example, to signal that some other packet could not be delivered).
Just because I can send the HTTP packet inside an Ethernet frame (without IP or TCP), does not mean that the model is broken, it's just that "null" is a valid protocol.
You can mount a partition as a folder in a NTFS partition (like it is on Linux).
How?
Computer Management ->Disk Management -> Right click on the partition you want to mount ->Change drive letters and paths->Add->Mount in the following empty NTFS folder
When you install Windows you can specify that the user profiles be somewhere else other than "C:\Documents and Settings"
No you can't, or at least my install process didn't give that option.
No, but nLite lets you customize this, as well as other settings (for example my temp directory is C:\Temp and not c:\Documents and Settings\user name\Local Settings\Temp)
You can move the profile of a single user to another folder/drive
How?
http://support.microsoft.com/default.aspx?scid=kb;en-us;236621
1. You can mount a partition as a folder in a NTFS partition (like it is on Linux).
2. When you install Windows you can specify that the user profiles be somewhere else other than "C:\Documents and Settings"
3. You can move the Documents and Settings folder after installation, but it's difficult.
4. You can move the profile of a single user to another folder/drive
5. On a single user systems (and a lot of PCs now are single user) there is no need to do any of that - just save the files to another drive.
Does Windows support other hard drives with GPT under normal BIOS? If so, then I do not see the problem. I would not need a 3TB drive as a system drive, I'd use a smaller one, either a fast HDD or a SSD.
I still do not get why current BIOSes cannot be made compatible with larger drives.
Now, granted, the MBR has the 2TB limit, but BIOS does not require that MBR. What I mean is, bios just loads the first (zeroth?) sector of the hard drive to memory and jumps to that location in memory. It does not care what is in it. From then the CPU executes whatever instructions were in that first sector.
Also, while boot code and the partition table are currently on the same sector, there is no particular need why they have to be so and why the partition table has to be in that format. As long as the OS and the boot code inside the first sector supports it the partition table can be whatever you want.
And yet, not all applications need drives that fast. For example, my movie collection does not need 100MB/s read speed. I do not need a 64GB SSD for the movie collection. If my collection is small enough to fit in 64GB then I could buy a used 80GB hard drive and save lots of money.
Optical requires the receiving device to have a power source, also, AFAIK the cables are fragile, you can abuse a regular audio cable quite a lot and it will still work, I suspect the optical cable has a minimum bend radius larger than a few mm.
As for the new interface - 3.5mm TRS connectors are common, the new interface would be incompatible with the current one and would not provide a lot of benefit - 3.5mm plugs are usually used to connect headphones and headphones do not care whether the ground is connected first or not. At least once Apple is doing something that would be compatible with other devices.