Apple has already been highly successful in alienating all the geeks (I include myself in this) and pushing them over to android (and to developing for it, and recommending it to their friends). Policy decisions which drive the "love the product; hate the company" would include:
- constant lockdown of iDevices. (yes, we understand that jailbreaking should only be for the techies, should be warranty-voiding, and should not be easy for
"grandma" to do by accident and then get upset about the consequences, but if we really want to, we should be able to).
- making it so hard for iPods to work on Linux - why can't they help out the libgpod devels by publishing specs.
- support for patents in general, and litigating against the competition rather than competing fairly. Also, DRM (though they've now mostly learnt that lesson).
- iTunes not working on Linux (or under Wine).
- not giving back as much as they take. Yes, Darwin is BSD, so it's legal, but it's really not cool to give back so little.
- killing off the "hacker" culture that they began with. Apple's hardware is really hard to tinker with: of course some of this is just because it's harder to experiment with a BGA A4 CPU than a DIL socketed 68k, but at least making parts available to hobbyists, and not suing them, would be a good start!
Finally, a personal gripe: Apple have lead the industry into making sure that only shortscreen-LCDs are available on any new hardware. I want 16:12 aspect, not 16:9 (and no doubt soon to be 16:8) !
Once you're down to the platters only, you can probably simply keep them in a filing cabinet. But folding the platter in half is a pretty good way to make it unreadable - at least without someone spending stupidly large amounts of money [if they care enough about your data!]. If you want to physically destroy them, it's probably cheaper to use a chemical means than thermal means. Finally, a single pass with "dd if=/dev/zero" is generally sufficient.
This problem has been solved before, for clock distribution on large chips. The way to make it slower, and *fairer* is:
1. All transactions execute at 00 seconds past the minute.
2. Between 10 seconds - 30 seconds past, the results of the transactions are confirmed.
3. Beween 30-50 seconds past, people or algorithms decide on and submit their next bids.
4. Then there is 10 seconds to get the next bid in, and the executions happen simultaneously.
It also means that the speed of light needn't disadvantage more distant traders, and gives the humans a chance against the machines. Also, because the total trades in an hour goes from ~ 100k to just 60, the speed of market-crashing is reduced.
Not so... if we are powerless the change the policy, we at least have a responsibility to fight back where we can. The best most passengers can do is to contribute slightly to making the job of the TSA agent more unpleasant: if we make their work unpleasant, and show the appropriate contempt, at least it will increase the job turnover, and raise the cost of the TSA. So do make the TSA agents realise that we don't respect them, don't trust them, and see them as unprofessional hired thugs...then they will quit their jobs a little faster than they otherwise would. Being "just a footsoldier" doesn't exempt people from guilt.
You're right, except for one thing...it's very tricky to have an impact that way. Newspapers like to break stories, so if the interesting material is just put online, it's surprisingly hard to get the "mainstream" press to take an interest in republishing it!
Indeed. Thanks a lot for a wonderful resource. I've been on "/." since 2000, and it taught me a lot. I've read it almost every day since then. Well done. P.S. The/. name still confuses non-geeks:-)
If it performs a phase change at a specified temperature, it means that there is a lot of extra thermal "capacitance" at one particular temperature. If we can choose that temperature to be, say 20 degrees C, this is useful. It would be a bit like water ice. As the freezer temperature falls from +1 to -1 degree, the ice releases a *lot* of stored heat (latent heat of fusion).
So, yes, this is a bit like building a large stone structure, which stays at a constant comfortable temperature by averaging out cold nights and hot days...but we don't need so much mass, and we can choose the temperature we want.
Remember that many functions are very much more efficient if they know the string length. For example, strlen() and strcat() both have to walk the string 1 byte at a time. This would easily compensate for the overhead of having an initial length field. Anyway, why not have both? Keep the null-terminated strings for the close-to-the-metal code, and implement a sensible string library for human-scale text processing, where most strings are tens to hundreds of characters in length.
BTW, the commonly quoted figures are that CFLs are "6-7 x better" in energy consumption (only true if you compare with a 40W bulb) and last "5x longer" (only if you allow the CFL to finally burn out, rather than replacing it after the light-output has fallen to 70% of the original value).
It's worthwhile comparing the Lumens/Watt ratio of the best-of-breed in both Halogen-Incandescent and CFL. I'm looking at the ones I can currently buy in my local store, and find the best Halogen is the "Philips Halogen Energy Saver" at 2100 lumens for 105 Watt. The best CFL is about 1200lm for 18W. In general the Halogen bulb uses about 3.3x the energy of the CFL. BUT...
* CFLs fall to about 50% efficiency over their lifetime; Incandescents remain at 99%
* The light quality of the Halogen is vastly better. Think colour-rendering, warmth, no high-frequency flicker. [Also, strobe effect is dangerous for rotating tools]
* Women look prettier under real light; under CFL (even the best of them), skin complexion looks poor; LEDs are usually even worse.
* Electric heating from lamps makes the wasted energy much lower; and the electricity is potentially green (nuclear/renewable). Although this isn't as efficient as gas heating, people turn on/off the lights as they enter/leave each room, but centrally-heat the whole house.
* These figures are for the UK; the lower voltage in the USA favours halogens slightly.
* CFLs are rarely disposed of properly, and if broken, the mercury is quite hazardous, especially to kids.
* The main problem with CFLs is the emitted spectrum, but the bulbs are ugly too; especially in chandeliers etc, which cannot sparkle.
So...the overall merit figure is quite neutral. Personally, I'd rather stick to tungsten-halogen; be good about turning off the lights that I don't use; and buy nuclear energy. Given this, we should let the consumer choose, and tax the actual CO2 emissions (burning things) rather than specific products.
It really is about time that American citizens had the self respect to tell the TSA thugs where to shove it. What will it take to get rid of the TSA? Already 40 BILLION dollars wasted on "security"- if you'd spent that on Education Aid in Pakistan and Saudi, it might have done some good. Instead, all that is done is to guarantee that other countries avoid the USA wherever possible. The TSA has directly caused our family not to visit - we've spent at least $100k in other countries over the last decade that would have been spent in the USA, where we used to have some wonderful vacations.
As an academic, I've always thought that it is completely immoral the way that the journals lock up one's work - though unfortunately they still have a significant monopoly because of the requirement to publish in a peer reviewed journal in order to get funding.
The current process is dreadful for 'official' academics (we have this wonderful tool called google, which is completely blinded to most of the science we might want to read, and even if we can search the abstract (not the full-text), we rarely get something so simple as a link to click and read.) It's ridiculously expensive for University libraries too. Even with unlimited official access, what I might be able to achieve in an hour on google will take a day. As for embedded hyperlinks, or the ability to comment on papers (blog style), forget it.
Worse than that, teachers, students, amateur scientists, and the entire third world are completely locked out.
For what it's worth, I always add something CC-license-like to all my papers, allowing complete reproduction provided attribution is given.
Well, you're right that bash is an acquired taste, for example using '$' as an operator to print the value of a variable. But the upside is the enormous number of Linux CLI commands available, all of which can do amazing things. What I'm really arguing for is to teach the use of the CLI, and the tools that are available, with just the merest smattering of shell-scripting to go with it. I speak from some experience - I've written 3000-line monsters in shell, and it's not especially pretty. Incidentally, the quirks of bash are much less severe if one teaches it as bash-scripting, rather than sh-scripting with some bash extensions.
Lastly, I'd argue that what programming students most want is immediate feedback, at the prompt. Bash is as good as BASIC in that regard. Eg:
echo -n "What's your name: "; read NAME; echo "HELLO $NAME" Admittedly other languages have shells (eg Python), but what Bash suffers in quirkiness, it makes up for by being so tightly integrated with every other command.
Bash isn't an especially pretty language, but it allows huge flexibility, by chaining together all the CLI commands. This makes it really easy to program quite advanced behaviour. Things like:
echo hello | festival --tts or
zenity --info --text "`fortune`" or
cat document.pdf | ssh hostname lpr
It would seem that they want to remain permanently in Alpha. As soon as any Linux Application (or Desktop environment) begins to become stable, feature-complete, and polished, someone decides to rewrite it from scratch, instead of finishing the one that is almost there. This means that we have about 50 half-done apps, and very few that are complete. As a longtime Linux user, I find this truly frustrating. For example, KDE 3.5.10 was almost perfect, but KDE 4 is still incomplete. Likewise Gnome. Amarok is less good now than it was. ALSA with dmix worked; pulseaudio still has gotchas. There are multiple different network management apps; none of them is complete, and they fight one another (eg Network Manager cannot be disconnected from gnome-keyring; this means that, even if the wifi keys are defined as system-wide properties, a laptop can cease to be remotely accessible). Why couldn't Software Center be built on top of Synaptic, rather than instead of it? Why waste time working on Wayland, when X is great (but just needs the drivers fixing - kudos to Nouveau, but they need more help). Why is it that when I put in an Audio CD, and immediately open an app to play it; the app gets a "disc not present" error unless I wait 10 seconds. (The fix would be for the app or kernel to notice that the drive has a new disc, but hasn't yet spun up, and to handle the error better). And there are about 17 different 3/4 done 1/2 stable VoIP apps, none of which is as resilient as Skype, and all of which fail in creatively different ways. Even the terminal bell doesn't work any more (there seems to be no way on earth to enable the pcspkr within Gnome).
Of course one could extend the figures further by including all human activity...for example, a young couple with a Prius and a baby are far worse for the environment than a bachelor in a gas guzzler!
I agree - we should be building the designs to incorporate what we learned over the last 40 years. Incidentally, nuclear waste is a solved problem - look at the Integral Fast Reactor (http://en.wikipedia.org/wiki/Integral_fast_reactor) - it's passively safe, and has no long-lived waste products. Typically, the wastes have decayed back to the same levels as the ore from which they were mined within 200 years.
Bit late now, but it's rather a pity that the Metric system couldn't have had imperial equivalences. There is no good reason why the metre couldn't have been 100 inches, or why the gram couldn't have been defined as 1/1000 pound. Actually, some engineers do: the "mil" is 0.001 inch - most electronic components use a 0.1" pin-spacing.
What about a way to display tooltips? You can't really hover a finger over a device (though it would be rather neat if the screen was sensitive to pressure, or to a finger nearly touching it). Also, why do so few devices implement long-press to get a (right-click) menu.
An event is a point in space-time: some marker that has coordinates (ct, x,y, z) in a particular frame of reference, S. In a different frame, S', the coordinates of that same event will be (ct', x', y', z'). Relativity lets us do the maths: if we know (ct,x,y,z) and the relative velocity of S' and S, then we can calculate (ct',x',y',z') (and vice versa).
In the standard twin-paradox, event A would be the twins saying goodbye, and B would be them saying hello again later. The time for the stationary twin is then
t_B - t_A, whereas the time for the moving twin has to be calculated in two frames, S' and S''.
However, to measure elapsed time, (B-A), both twins have to agree on the co-ordinates of the events. You can synchronise one event as the twins pass each other, but if they remain in intertial frames, you cannot define a second event to the satisfaction of both twins.
> But how do we know it was the "moving" twin that stopped and returned? > From the point of view of that twin the "stationary" twin is the one who moved away and came back.
The symmetry is broken, so we can tell. The stationary twin feels no acceleration (and remains in one inertial frame of reference throughout). The moving twin must accelerate away, turn around (decelerate + accelerate) and then brake on arrival. So he switches frame at least 3 times.
In SR, there is no special "rest" frame, and you can't tell what your own absolute speed is (it's not even defined). However, you *can* tell that you are accelerating, because of F = m.a . If you experience a force, you must be accelerating. The 2nd twin can feel himself being pushed back in his spacecraft's seat.
[in SR there is no gravity; once gravitation enters the fray, we have the equivalence principle (briefly, inertia === gravity) and then we must resort to GR]
Not quite sure you've got this right. Switching inertial frames doesn't "make the clocks converge". What happens is that, because the twin who moves away and then returns is doing an intertial-frame swap (he has to accelerate when he turns round to come back), we can break the symmetry and define him as the one who "moves". this symmetry breaking is why it isn't just "all relative". So the one who moves is the younger. (But time passes more slowly for him; he doesn't get to have "any more fun"). BTW, SR is actually quite straightforward - you just have to understand it as a way of doing co-ordinate transformation between two frames.
Seriously, why can't we have a chemical process such that I empty the lawnmower in, and get milk/cheese/cream out? That would be hugely beneficial, and have major benefits for environment, food supply, and vegans.
Sorry - you've got this wrong! It *is* possible to build an array of synchronised clocks at widely separated positions, provided that they are all in the same inertial frame (no relative velocities). [On earth, there will be effects due to the earth's rotation (various latitudes) and differences in height above sea-level, but these really are absolutely tiny]
Slashdot Mirror
Apple has already been highly successful in alienating all the geeks (I include myself in this) and pushing them over to android (and to developing for it, and recommending it to their friends). Policy decisions which drive the "love the product; hate the company" would include:
- constant lockdown of iDevices. (yes, we understand that jailbreaking should only be for the techies, should be warranty-voiding, and should not be easy for
"grandma" to do by accident and then get upset about the consequences, but if we really want to, we should be able to).
- making it so hard for iPods to work on Linux - why can't they help out the libgpod devels by publishing specs.
- support for patents in general, and litigating against the competition rather than competing fairly. Also, DRM (though they've now mostly learnt that lesson).
- iTunes not working on Linux (or under Wine).
- not giving back as much as they take. Yes, Darwin is BSD, so it's legal, but it's really not cool to give back so little.
- killing off the "hacker" culture that they began with. Apple's hardware is really hard to tinker with: of course some of this is just because it's harder to experiment with a BGA A4 CPU than a DIL socketed 68k, but at least making parts available to hobbyists, and not suing them, would be a good start!
Finally, a personal gripe: Apple have lead the industry into making sure that only shortscreen-LCDs are available on any new hardware. I want 16:12 aspect, not 16:9 (and no doubt soon to be 16:8) !
Once you're down to the platters only, you can probably simply keep them in a filing cabinet. But folding the platter in half is a pretty good way to make it unreadable - at least without someone spending stupidly large amounts of money [if they care enough about your data!]. If you want to physically destroy them, it's probably cheaper to use a chemical means than thermal means.
Finally, a single pass with "dd if=/dev/zero" is generally sufficient.
This problem has been solved before, for clock distribution on large chips. The way to make it slower, and *fairer* is:
1. All transactions execute at 00 seconds past the minute.
2. Between 10 seconds - 30 seconds past, the results of the transactions are confirmed.
3. Beween 30-50 seconds past, people or algorithms decide on and submit their next bids.
4. Then there is 10 seconds to get the next bid in, and the executions happen simultaneously.
It also means that the speed of light needn't disadvantage more distant traders, and gives the humans a chance against the machines. Also, because the total trades in an hour goes from ~ 100k to just 60, the speed of market-crashing is reduced.
Not so... if we are powerless the change the policy, we at least have a responsibility to fight back where we can. The best most passengers can do is to contribute slightly to making the job of the TSA agent more unpleasant: if we make their work unpleasant, and show the appropriate contempt, at least it will increase the job turnover, and raise the cost of the TSA. So do make the TSA agents realise that we don't respect them, don't trust them, and see them as unprofessional hired thugs...then they will quit their jobs a little faster than they otherwise would. Being "just a footsoldier" doesn't exempt people from guilt.
You're right, except for one thing...it's very tricky to have an impact that way. Newspapers like to break stories, so if the interesting material is just put online, it's surprisingly hard to get the "mainstream" press to take an interest in republishing it!
Indeed. Thanks a lot for a wonderful resource. I've been on "/." since 2000, and it taught me a lot. I've read it almost every day since then. Well done. /. name still confuses non-geeks :-)
P.S. The
If it performs a phase change at a specified temperature, it means that there is a lot of extra thermal "capacitance" at one particular temperature. If we can choose that temperature to be, say 20 degrees C, this is useful. It would be a bit like water ice. As the freezer temperature falls from +1 to -1 degree, the ice releases a *lot* of stored heat (latent heat of fusion).
So, yes, this is a bit like building a large stone structure, which stays at a constant comfortable temperature by averaging out cold nights and hot days...but we don't need so much mass, and we can choose the temperature we want.
Remember that many functions are very much more efficient if they know the string length. For example, strlen() and strcat() both have to walk the string 1 byte at a time. This would easily compensate for the overhead of having an initial length field. Anyway, why not have both? Keep the null-terminated strings for the close-to-the-metal code, and implement a sensible string library for human-scale text processing, where most strings are tens to hundreds of characters in length.
BTW, the commonly quoted figures are that CFLs are "6-7 x better" in energy consumption (only true if you compare with a 40W bulb) and last "5x longer" (only if you allow the CFL to finally burn out, rather than replacing it after the light-output has fallen to 70% of the original value).
It's worthwhile comparing the Lumens/Watt ratio of the best-of-breed in both Halogen-Incandescent and CFL. I'm looking at the ones I can currently buy in my local store, and find the best Halogen is the "Philips Halogen Energy Saver" at 2100 lumens for 105 Watt. The best CFL is about 1200lm for 18W. In general the Halogen bulb uses about 3.3x the energy of the CFL. BUT...
* CFLs fall to about 50% efficiency over their lifetime; Incandescents remain at 99%
* The light quality of the Halogen is vastly better. Think colour-rendering, warmth, no high-frequency flicker. [Also, strobe effect is dangerous for rotating tools]
* Women look prettier under real light; under CFL (even the best of them), skin complexion looks poor; LEDs are usually even worse.
* Electric heating from lamps makes the wasted energy much lower; and the electricity is potentially green (nuclear/renewable). Although this isn't as efficient as gas heating, people turn on/off the lights as they enter/leave each room, but centrally-heat the whole house.
* These figures are for the UK; the lower voltage in the USA favours halogens slightly.
* CFLs are rarely disposed of properly, and if broken, the mercury is quite hazardous, especially to kids.
* The main problem with CFLs is the emitted spectrum, but the bulbs are ugly too; especially in chandeliers etc, which cannot sparkle.
So...the overall merit figure is quite neutral. Personally, I'd rather stick to tungsten-halogen; be good about turning off the lights that I don't use; and buy nuclear energy. Given this, we should let the consumer choose, and tax the actual CO2 emissions (burning things) rather than specific products.
It really is about time that American citizens had the self respect to tell the TSA thugs where to shove it. What will it take to get rid of the TSA? Already 40 BILLION dollars wasted on "security"- if you'd spent that on Education Aid in Pakistan and Saudi, it might have done some good. Instead, all that is done is to guarantee that other countries avoid the USA wherever possible. The TSA has directly caused our family not to visit - we've spent at least $100k in other countries over the last decade that would have been spent in the USA, where we used to have some wonderful vacations.
As an academic, I've always thought that it is completely immoral the way that the journals lock up one's work - though unfortunately they still have a significant monopoly because of the requirement to publish in a peer reviewed journal in order to get funding.
The current process is dreadful for 'official' academics (we have this wonderful tool called google, which is completely blinded to most of the science we might want to read, and even if we can search the abstract (not the full-text), we rarely get something so simple as a link to click and read.) It's ridiculously expensive for University libraries too. Even with unlimited official access, what I might be able to achieve in an hour on google will take a day. As for embedded hyperlinks, or the ability to comment on papers (blog style), forget it.
Worse than that, teachers, students, amateur scientists, and the entire third world are completely locked out.
For what it's worth, I always add something CC-license-like to all my papers, allowing complete reproduction provided attribution is given.
Well, you're right that bash is an acquired taste, for example using '$' as an operator to print the value of a variable. But the upside is the enormous number of Linux CLI commands available, all of which can do amazing things. What I'm really arguing for is to teach the use of the CLI, and the tools that are available, with just the merest smattering of shell-scripting to go with it. I speak from some experience - I've written 3000-line monsters in shell, and it's not especially pretty. Incidentally, the quirks of bash are much less severe if one teaches it as bash-scripting, rather than sh-scripting with some bash extensions.
Lastly, I'd argue that what programming students most want is immediate feedback, at the prompt. Bash is as good as BASIC in that regard. Eg:
echo -n "What's your name: "; read NAME; echo "HELLO $NAME"
Admittedly other languages have shells (eg Python), but what Bash suffers in quirkiness, it makes up for by being so tightly integrated with every other command.
Bash isn't an especially pretty language, but it allows huge flexibility, by chaining together all the CLI commands.
This makes it really easy to program quite advanced behaviour. Things like:
echo hello | festival --tts
or
zenity --info --text "`fortune`"
or
cat document.pdf | ssh hostname lpr
It would seem that they want to remain permanently in Alpha. As soon as any Linux Application (or Desktop environment) begins to become stable, feature-complete, and polished, someone decides to rewrite it from scratch, instead of finishing the one that is almost there. This means that we have about 50 half-done apps, and very few that are complete. As a longtime Linux user, I find this truly frustrating. For example, KDE 3.5.10 was almost perfect, but KDE 4 is still incomplete. Likewise Gnome. Amarok is less good now than it was. ALSA with dmix worked; pulseaudio still has gotchas. There are multiple different network management apps; none of them is complete, and they fight one another (eg Network Manager cannot be disconnected from gnome-keyring; this means that, even if the wifi keys are defined as system-wide properties, a laptop can cease to be remotely accessible). Why couldn't Software Center be built on top of Synaptic, rather than instead of it? Why waste time working on Wayland, when X is great (but just needs the drivers fixing - kudos to Nouveau, but they need more help). Why is it that when I put in an Audio CD, and immediately open an app to play it; the app gets a "disc not present" error unless I wait 10 seconds. (The fix would be for the app or kernel to notice that the drive has a new disc, but hasn't yet spun up, and to handle the error better). And there are about 17 different 3/4 done 1/2 stable VoIP apps, none of which is as resilient as Skype, and all of which fail in creatively different ways. Even the terminal bell doesn't work any more (there seems to be no way on earth to enable the pcspkr within Gnome).
Of course one could extend the figures further by including all human activity...for example, a young couple with a Prius and a baby are far worse for the environment than a bachelor in a gas guzzler!
I agree - we should be building the designs to incorporate what we learned over the last 40 years. Incidentally, nuclear waste is a solved problem - look at the Integral Fast Reactor (http://en.wikipedia.org/wiki/Integral_fast_reactor) - it's passively safe, and has no long-lived waste products. Typically, the wastes have decayed back to the same levels as the ore from which they were mined within 200 years.
See: http://nextbigfuture.com/2011/03/deaths-per-twh-by-energy-source.html
Bit late now, but it's rather a pity that the Metric system couldn't have had imperial equivalences. There is no good reason why the metre couldn't have been 100 inches, or why the gram couldn't have been defined as 1/1000 pound. Actually, some engineers do: the "mil" is 0.001 inch - most electronic components use a 0.1" pin-spacing.
What about a way to display tooltips?
You can't really hover a finger over a device (though it would be rather neat if the screen was sensitive to pressure, or to a finger nearly touching it).
Also, why do so few devices implement long-press to get a (right-click) menu.
That only has one event in it, rather than 2.
An event is a point in space-time: some marker that has coordinates (ct, x,y, z) in a particular frame of reference, S.
In a different frame, S', the coordinates of that same event will be (ct', x', y', z').
Relativity lets us do the maths: if we know (ct,x,y,z) and the relative velocity of S' and S, then we can calculate (ct',x',y',z') (and vice versa).
In the standard twin-paradox, event A would be the twins saying goodbye, and B would be them saying hello again later. The time for the stationary twin is then
t_B - t_A, whereas the time for the moving twin has to be calculated in two frames, S' and S''.
However, to measure elapsed time, (B-A), both twins have to agree on the co-ordinates of the events. You can synchronise one event as the twins pass each other, but if they remain in intertial frames, you cannot define a second event to the satisfaction of both twins.
> But how do we know it was the "moving" twin that stopped and returned?
> From the point of view of that twin the "stationary" twin is the one who moved away and came back.
The symmetry is broken, so we can tell. The stationary twin feels no acceleration (and remains in one inertial frame of reference throughout).
The moving twin must accelerate away, turn around (decelerate + accelerate) and then brake on arrival. So he switches frame at least 3 times.
In SR, there is no special "rest" frame, and you can't tell what your own absolute speed is (it's not even defined). However, you *can* tell that you are accelerating, because of F = m.a . If you experience a force, you must be accelerating. The 2nd twin can feel himself being pushed back in his spacecraft's seat.
[in SR there is no gravity; once gravitation enters the fray, we have the equivalence principle (briefly, inertia === gravity) and then we must resort to GR]
Not quite sure you've got this right. Switching inertial frames doesn't "make the clocks converge". What happens is that, because the twin who moves away and then returns is doing an intertial-frame swap (he has to accelerate when he turns round to come back), we can break the symmetry and define him as the one who "moves". this symmetry breaking is why it isn't just "all relative". So the one who moves is the younger. (But time passes more slowly for him; he doesn't get to have "any more fun"). BTW, SR is actually quite straightforward - you just have to understand it as a way of doing co-ordinate transformation between two frames.
Seriously, why can't we have a chemical process such that I empty the lawnmower in, and get milk/cheese/cream out? That would be hugely beneficial, and have major benefits for environment, food supply, and vegans.
Sorry - you've got this wrong!
It *is* possible to build an array of synchronised clocks at widely separated positions, provided that they are all in the same inertial frame (no relative velocities).
[On earth, there will be effects due to the earth's rotation (various latitudes) and differences in height above sea-level, but these really are absolutely tiny]