Alright, now what about recent reports about how most equipment in Hospitals are infected by some sort of Malware?. It's safe to say when these Robots are placed in those facilities, they will become infected along with other equipment. You really want this infected Robot poking around your circulatory system?.
Devices like this are usually not connected to any network in that sense.... For this one it would only have to be connected if they wanted the vein images from the machine... There already exists many types of automated systems in hospitals today, like heart-lung machines or how about pace-makers that people are walking around with outside... If they manage to keep the heart-lung machines working i think they can manage to keep this one running too...
Not only this, in my area I've been seeing advertisements on TV from Law Offices whom are seeking applicants who have been injured by Robotic Equipment. So the injuries have already started. You think about those things. What would be nice would be a safer, non-automated device for Blood extraction.
ehmm... "Robotic equipment"... quite broad statement... can be anything from someone tripping over an automated vaccum-cleaner to someone walking into the operating-zone of a factory robot...
And most cars are controlled by computers... so why does not the break's fail or the gas locked to the maximum all the time?
You can design safe systems... If something invalid is detected then go to fail mode and safely stop the procedure. Failure detection can be a combination of both software and physical sensors and could be a secondary system that would operate completely separate from the main one.
And Therac-25 is a really bad example of a poorly designed system.... But there is one interesting thing about it on wikipedia... "The system noticed that something was wrong and halted the X-ray beam, but merely displayed the word "MALFUNCTION" followed by a number from 1 to 64. The user manual did not explain or even address the error codes, so the operator pressed the P key to override the warning and proceed anyway." So here it was user-fault again... If something says "MALFUNCTION" and displays an undocumented error-message it might be a good idea to check with the manufacturer what is actually is instead of just proceeding...
Also: "The software was written in assembly language that might require more attention for testing and good design. However the choice of language by itself is not listed as a primary cause in the report. The machine also used its own operating system."
Why did they write this in assembly??? Make it much harder to do reviews on.... If they did it because they had to little cpu-resources then should have added more resources.
Designing software for safety and/or graceful failures can be tricky, but it's not hard...
3'rd party site that wants more visitors creates a booking page for many popular resturants... People fill in their information and then they will get a mail if they managed to get the reservation...
It's not so hard to imagine... And a great way to attract more people to your page...
Well.. People that says these things have never been to a 3-star restaurant...... The thing is that the $50 per meal restaurants cannot afford good (good and good... they are probably not bad anyway.. ) ingredients for those prices.... Good wagyu beef costs about $50 for the restaurant to buy for one portion (150grams or so)... Also there is a big thing on how the animals have been raised so you actually should look at the farm where the meat came from.. (Btw, if you have never tried it then don't... it will ruin all other meat for you:)
Never buy chicken that have been grown in 7-9 weeks and basically force-fed with corn.. Buy a free-ranging chicken that has been grown in the natural speed.. It tastes so much better.
Never buy salmon (or any fish) from fish-farms. To start with they contain less Omega 3 (the good one) and much higher concentration of Omega 6 (one of the bad ones). They are also fed antibiotics and dye to give the flesh an appealing red color... Also fish from fishfarms contain much higher concentrations of PCB/pesticides and other harmful chemicals.
Never buy beef that have only been fed corn, as most meat-animals are. Try and go for free-ranging cattle from a fairly tempered areas with lots of good food...
There is a difference in food and food... And going for quantity over quality is always bad..... Buy a good meal and enjoy the experience... If you are still hungry afterwards then have some fruit or grab a sandwich..
And about the linked article. To start with i do not believe their statement of 15CPS they stated since they said that this was about double the normal amount, but normal backround radioation is ~20-80CPM so they are off by a factor of altleast 80. If it would really be 15CPS that would be 900CPM or about 9microseivert per hour and that would be quite huge... But even at those levels And as they write they found areas with up to 50CPS and that would be 30microseivert per hour and that would be spectacular if nobody except one small paper would have picked up on... But on the other hand, even with those extremly blown up figures you still only have an increase of cancer risk of 1 in a 1,000 per 12.5 milliSievert (mSv) (or 12500microsieverts) so that would still take about 416 hours to increase the risk by one in 1000 at the most extreme value they reported.
But then the next question is quite important.. What type of radiation is it? Gamma/Beta/Alpha??? Gamma can be quite bad, but Alpha is quite safe unless you manage to inhale highly radioactive material that i doubt exists on that beach with those levels, even if there was 15CPS.....
Just a few simple things that can bring up the background radiation: - Soot/ash from coal-plants, actually it can be quite high in really bad stuff. - Spill of potassium (fertilizer) - Concrete - Stone from a quarry that contained a bit higher PPM of uranium.
"That is double the amount of radiation normally found in the atmosphere in Britain" this is comparison is also completely invalid.. You don't measure the atmosphere, you measure the ground or material around you.
Normal background radiation is between 25-75 in the US, so 15 CPM is nothing...
Read http://xkcd.com/radiation/ and then take into account that 100 cpm is about equal to 1 microseivert per hour.
From above poster: A trip from NY to LA would give you 40microsieverts/hr... During a normal day you will recieve about 10microsieverts/hr...
Or you could do a google yourself on "number of deaths coal oil nuclear"
The thing with nuclear-power is that everything happens at the same place and affects more people in one go..... And i prefer something that kills ~90 people per year over for example Oil that kills ~36000 per year... Or natural gas that kills ~4000 per year.. Even wind-power kills ~150 per year....
The problem is that it's public opinion that drives the direction of how we generate power, but the problem is that the general population don't have the knowledge to actually make an informed decision, and neither can i fully. The thing is that nuclear-power, and there are many types of technologies, is probably the only thing that will be able to sustain the human population for the next 50 years until we can perfect fusion-power or something else that do not have the same impact.
If nuclear-power would still be seen positive by the general population we would also build new reactors that are safer instead of staying with the old reactors that have known safety issues.
Makes me sick of a justice system where a 11'yr child could be charged with a crime for something like this.... Does not matter if he was autistic or not, but autistic kids might even flip out a bit more if someone grabs them (not sure if that's the same for everyone?)....
Any police should be able to handle a 11'yr old kid verbally, and worst case taking a hold of his arms and then getting a kick or two from a kid is not really that bad.. If someone work as a police-officer they should at least be able to handle a few bruises when handling kids..
Kids are kids.. They flip out from time to time and it's normal..... The strange thing seems to be that the police that are sent to schools don't get any proper training on how to handle kids..... And working in a place where you have autistic (and other things too?) should require quite a bit more of education...
92233720368547800/1296.30 = 71151523851383 ( $1296.3 was price per ounce i could find ) (71151523851383*28.3495231) / 1000 = 2017111769024 Kg's of gold. ( One ounce 28.3495231gram... )
= 0.115 cubic kilometers... Or a cube with 486 meter sides.....
Will probably be quite hard to grow anything on it... But the biggest problem would probably be all the leprechauns running around on that island... Not to mention all the rainbows that would end there causing rainbow-blindness:)
Pay off?? For that amount i could basically buy any country in the world... Offer every citizen $5000000 to move out of the country, but only if at least 99% of the population leaves. Even if there are a few wealthy people that chose to stay there would be nobody that would take care of the infrastructure for them;)
$92,233,720,368,547,800 / 6 billion = $15372286 per person in the world..
But there is a problem with having that much money and starting to spread it around would cause a devaluation of money.. And since it's so much it would spread and would affect all currencies in the world...... Ie... those $15372286 per person might be able to buy some dinner after some time...
And it would not matter if it's a fiat currency or gold-backed currency... The wanted resources (food/water/electricity etc) stay the same, it's just how the money is spread out.
If the problem is cyber-warfare then there is a easy solution to the problem... Just don't frecking connect sensitive machines to it... If you want a secure tunnel via the internet that is also 'easy' to make secure and more or less hacker-proof... The issue is when you want the sensitive machines/networks to be both secure *and* connected to the rest at the same time...
One line specification of something like that is not really real world so adding a few things...
- What about during initialization of the frame-buffer that will take 50ms? Is it ok to open screen and clear and set pixel or does the framebuffer have to be initialized with the specified pixel before output is activated?
- We have 3 hardware layers where layer 2 has this pixel and layer 3 displays error message. Error message X is defined as fullscreen and during error the application is paused. Should the error-screen set alpha on pixel 0,0 to show it or do we not have to conform to this since the application is in pause-mode?
- Hardware requires alpha value to be set. Layer one will show video. Spec also says that all graphics should have a alpha of 200, does this include this pixel? Is it 128,128,128 when having a black layer 1 or is it 128,128,128 when having a white layer 1 in that case?
Even a dead-simple specification like that can be broken down so much when adding a few more requirements that makes it impossible for anyone to think of all corner-cases.. And if they would have written a spec to that degree of accuracy no developer would be able read the whole specification and remember everything during development. You usually also divide work up on multiple developers where each developer gets a part of the specification that he should implement, but that developer only sees his part and don't want or have time to read and understand the full spec of the software.
For each overlapping feature you have an exponential growth of sub-feature-specifications that would have to be written to be able to handle all corner-cases...
It all depends on the available funds for the potential customers and the cost-difference between thin/thick-client and a real PC... So the prime-time for some specific tech might differ with many years between places... It all boils down to what people can afford to pay and the benefits they might get with it.
My bad experiences with thin clients are that sometimes an app that everyone is running on the server end crashes somehow, and there is no way to correct the situation except to restart the server the app runs on. Sometimes that could be done gracefully, other times, more frequently, everyone would need to exit their remote sessions, wait until further instructions, then log in again. If there are lots of folks logged in, and they all have to exit because one app on the back end crashed that can cause serious consternation.
That is why running a thick-client can be a very good alternative... The app runs on the actual client but without the need of local administration and storage..
Paying for a simple service like this can be good in places where people cannot afford to buy a PC, or don't want to shell out with lots (for them) of money.
The benefits of this is that even poor families can get access to a whole new world of possibilities and learning new things.... Education for the population is *always* good. Take a look at the OLPC project and what it's all about...
The wealth of a country does usually match the curve of the it's population education-level.. So as the population gains new skills their wages will go up and they can start afford real computers and this business-model will be over and/or converted to a ISP model for internet-connectivity..
Making a 3 server redundant cluster should be easy... Just depends on how the solution will look.. Dumb-clients can be tricky... Thick-clients just need some remote cluster-fs they run everything from and a file-share for user-contents.
With thin-clients: (dumb-clients) Simple UPS - 1k Euro 6x Servers 2k each (3 extra for failures) 2x storage 2k each. VMWare or similar solution with that has HA so each user actually have their session running on 2 machines in case of failures... (no idea of cost)
With thick-clients: Simple UPS - 1k Euro (probably overkill) 3x Servers 2k each. (one extra for failures..) 2x Storage 2k each. (6Tb should be enough for 200 users unless they are storing movies and stuff) = 11k Euro
No need for 24/7 tech's... Running a system like this would probably take up ~1-2h per day, on average, in administration (backup's/upgrades etc).... The rest would be on-call for system failures, and with a HA solution there would not really be a need of more than 8-24h 'time to fix'.... That's the beauty of thin/thick clients... If something happens for someone they just reboot and reload the standard image... User contents are just files on a file-share...
Oh, forgot... There are also some quite nice things that can be done.. Like time-staping the input of the user and having the server do corrections to correct for the lag.. Can result in some graphics-glitches, but if you have a latency of 30ms that should be minimal for the user..
Also, you could stream some of the graphics as pure OpenGL commands for better graphics for the most visible items..
H.265 could be a very nice addition to reduce the bandwidth needed for good video.. 4Mbit should be more than enough for good 720p.
Terminals can be good, but it all depends on how it's all deployed and what things that are used...
There are several versions... (my own alias'es for the different versions, don't yell:] ) 1. Thin-client.. All it does it display an image sent from a server.. 2. Thick-Client... Everything is run on the client but all code is stored on the shared server. CPU hungry app's can be run on the server. 3. Fat-Client.. Everything runs on the client and everything is stored on the shared server. The system-image is cached on the client for better performance. (overlay-fs or similar) 4. Thin-PC.. Everything runs on the client and everything is stored on the shared server. The system-image is cached on the client for better performance and user-data is stored on the client as well but mirrored to the shared server via a gluster/AFS or other cluster based file-system.
In terms of costs 1,2,3 can be fairly cheap.. In terms of performance options 3 or 4 are very good... Option 4 can also be used in the case when the main server is down.
With option 2,3 the user can also choose to install extra applications, but would need some type of block-deduplication to reduce storage-costs. With option 4 the user can install extra applications into their local storage.
With option 1,2,3 the administration can be done centrally, with option 4 client-specific administration might be needed depending on the setup.
With option 1,2,3 all user-data is stored on centralized servers so replacing a terminal is simple and will not result in data-loss. With option 4 user-data is stored on centralized server and on local HD. Can result in data-loss and can cause strange issues for the users depending on usage.
The balance you will have to make is if it's cheaper to get good nice and fast servers with redundancy or if it's cost-beneficial to have slower servers and faster clients and also how much the administration will cost for the different setups.. The more client's you have the more money you save on option 1 but also the less performance for the users..
You are probably thinking of putting a server-rack in the building or something? Not really needed... Lets say you have a big building with say 200 apartments... That could easily be taken care of with a couple of simple servers... dual cpu with 24-32 cores each.. 144Gb ram... A couple of internal disks for booting and applications and then some shared storage..... We bought a couple of servers (dual cpu with 24 cores and 144Gb ram 2x500Gb disks with raid-1 setup ) refurbished for about 4k euro about a year ago.....
If people are using simple terminals for web/mail and other such tasks it requires soo little in terms of performance you could get away with quite a small setup...
Some rough numbers with everything in the building.. 10000/12/3/200 = 1.38 Eur for server-cost. (Counting on less than 1k Euro for cooling of those 2 servers. (2 x simple AC) ) 300/12/3 = 8.3 Eur for terminal 100/12/3 = 2.7 Eur for cabeling (paid off after 3 years) Salary for the admin 10 Eur (no idea what a normal salary is there) and he will earn (200*12*10) = 24000 Euro per year if everyone joins in. Internet connectivity for the building 300/200 = 1.5 Eur. (you should be able to share a 100Mbit connection on 200 apartments. Don't know the price so estimated it to 300Eur..) 5Eur for misc things like equipment replacement, billing, rent of room for servers etc. (5*200*12 = 12000 Euro per year) So total of 28.88 Euro per apartment per month.. Compare that to buying a computer for ~600Eur (16.6Euro per month + internet 15Euro = 31.6 Euro)...
Don't know the price for electricity there, but that should be fairly minimal and could probably be included in the "misc" group...
What i have not counted on here is licenses, but for mail+web+word-processing and other such simpler tasks you could get everything going with free software.... Even the "terminal" could in that case be used for some of the simpler processing like the web-browser.. Cheap Intel Atom with 2Gb of ram + monitor could set you back maybe 200-250Eur including monitor.. when bought in larger quantities... No idea about their ARM based machines the article wrote about...
I think i used some bloated numbers in my calculation, except for the salary since i have no clue about what would be normal there, but you could probably cut the server-cost in half.. 100
Also for the ones that buy's their own computers he could sell internet-connectivity for extra profit...Ie no terminals to lease etc, no need for extra cpu/memory in the server(s) etc....
Some other extra services he could provide at a small extra cost: - Internet connectivity to people with their own computers.. (almost pure profit..) - WiFi access in the whole building for phones/tablets etc. (almost pure profit.... add some costs for wifi-routers... Could be done on the cheap with commercial OpenWRT based solutions..) - VoIP service to the whole building. (He could even rent them the wired/wireless phone..) - VoIP-based Fax-service. If fax arrives it's printed on a shared fax and delivered to the apartment, or sent out as PDF to the person's mailbox. For outgoing faxes it can be done as PDF being sent or going to the shared printer/fax room and scanning the paper into the account the person has on the servers. - Shared printer-service. 0.1Eur per sheet.. - Collaborating with other such residences offering WiFi to a large area and then sharing the profits for this extra service. - Collaborating with other such residences buying a cheaper/faster connection, that is then shared between them. - Offering backup-services of user-data. (1Euro per month or so... 1*200*12*3 = 7200Euro.. For that he could buy quite a nice disk-based solution + offsite tapes..) - Collaborating with other such residences that are close by and sharing more of the infrastructure for lower costs..
But even if you don't have genetic patents they still have an incentive... Patent for drug X that affect gene Y in this manner.. Or Patent for cheaply and accurately detecting gene Z.
Genetic patents are not really good because they patent the actual gene so nobody else can make similar and better drugs or tests for that specific thing... There in lies the evil...
If we take genetically modified crops... They are patented and they are suing people that reuse seeds from their last harvest...... Now lets say Company A gets a patent where they insert some new genetic sequence into a person... If that person then gets a kid then he will be infringing too?? Or how about if a virus would absorb a genetic-sequence and then spread over a population... Will the infected be patent-infringers too??
Patents on genes should not exist. It's a really dangerous thing that actually hinders development of new treatments and blocks actual research.
The power-outlet differs between countries, but they all use 240V so using a simple adaptor will work... Laptop's AC adaptors are usually designed for anything between 110-240V...
Internet in most countries in europe is quite good, and you can buy fairly cheap 3g prepaid cards for pure data too..... and loads of cafe's have free wifi too... Some prices for Sweden : http://prepaidwithdata.wikia.com/wiki/Sweden So buy a prepaid sim for the wanted duration and plug it into a cheap android phone and you have a perfect way to get online where ever you might be..
Slashdot Mirror
Alright, now what about recent reports about how most equipment in Hospitals are infected by some sort of Malware?. It's safe to say when these Robots are placed in those facilities, they will become infected along with other equipment. You really want this infected Robot poking around your circulatory system?.
Devices like this are usually not connected to any network in that sense.... For this one it would only have to be connected if they wanted the vein images from the machine...
There already exists many types of automated systems in hospitals today, like heart-lung machines or how about pace-makers that people are walking around with outside... If they manage to keep the heart-lung machines working i think they can manage to keep this one running too...
Not only this, in my area I've been seeing advertisements on TV from Law Offices whom are seeking applicants who have been injured by Robotic Equipment. So the injuries have already started. You think about those things. What would be nice would be a safer, non-automated device for Blood extraction.
ehmm... "Robotic equipment"... quite broad statement... can be anything from someone tripping over an automated vaccum-cleaner to someone walking into the operating-zone of a factory robot...
And most cars are controlled by computers... so why does not the break's fail or the gas locked to the maximum all the time?
You can design safe systems... If something invalid is detected then go to fail mode and safely stop the procedure. Failure detection can be a combination of both software and physical sensors and could be a secondary system that would operate completely separate from the main one.
And Therac-25 is a really bad example of a poorly designed system.... But there is one interesting thing about it on wikipedia...
"The system noticed that something was wrong and halted the X-ray beam, but merely displayed the word "MALFUNCTION" followed by a number from 1 to 64. The user manual did not explain or even address the error codes, so the operator pressed the P key to override the warning and proceed anyway."
So here it was user-fault again... If something says "MALFUNCTION" and displays an undocumented error-message it might be a good idea to check with the manufacturer what is actually is instead of just proceeding...
Also:
"The software was written in assembly language that might require more attention for testing and good design. However the choice of language by itself is not listed as a primary cause in the report. The machine also used its own operating system."
Why did they write this in assembly??? Make it much harder to do reviews on.... If they did it because they had to little cpu-resources then should have added more resources.
Designing software for safety and/or graceful failures can be tricky, but it's not hard...
Possibility
3'rd party site that wants more visitors creates a booking page for many popular resturants... People fill in their information and then they will get a mail if they managed to get the reservation...
It's not so hard to imagine... And a great way to attract more people to your page...
Then don't watch http://www.imdb.com/title/tt1467304/ or http://www.imdb.com/title/tt1530509/ :)
Well.. People that says these things have never been to a 3-star restaurant...... The thing is that the $50 per meal restaurants cannot afford good (good and good... they are probably not bad anyway .. ) ingredients for those prices.... Good wagyu beef costs about $50 for the restaurant to buy for one portion (150grams or so)... Also there is a big thing on how the animals have been raised so you actually should look at the farm where the meat came from.. (Btw, if you have never tried it then don't... it will ruin all other meat for you :)
Never buy chicken that have been grown in 7-9 weeks and basically force-fed with corn.. Buy a free-ranging chicken that has been grown in the natural speed.. It tastes so much better.
Never buy salmon (or any fish) from fish-farms. To start with they contain less Omega 3 (the good one) and much higher concentration of Omega 6 (one of the bad ones). They are also fed antibiotics and dye to give the flesh an appealing red color... Also fish from fishfarms contain much higher concentrations of PCB/pesticides and other harmful chemicals.
Never buy beef that have only been fed corn, as most meat-animals are. Try and go for free-ranging cattle from a fairly tempered areas with lots of good food...
There is a difference in food and food... And going for quantity over quality is always bad..... Buy a good meal and enjoy the experience... If you are still hungry afterwards then have some fruit or grab a sandwich..
To start with you should read up on this. Radioactivity from coal-plants are actually quite high in comparison.
http://en.wikipedia.org/wiki/Background_radiation
And about the linked article.
To start with i do not believe their statement of 15CPS they stated since they said that this was about double the normal amount, but normal backround radioation is ~20-80CPM so they are off by a factor of altleast 80.
If it would really be 15CPS that would be 900CPM or about 9microseivert per hour and that would be quite huge... But even at those levels
And as they write they found areas with up to 50CPS and that would be 30microseivert per hour and that would be spectacular if nobody except one small paper would have picked up on...
But on the other hand, even with those extremly blown up figures you still only have an increase of cancer risk of 1 in a 1,000 per 12.5 milliSievert (mSv) (or 12500microsieverts) so that would still take about 416 hours to increase the risk by one in 1000 at the most extreme value they reported.
But then the next question is quite important.. What type of radiation is it? Gamma/Beta/Alpha??? Gamma can be quite bad, but Alpha is quite safe unless you manage to inhale highly radioactive material that i doubt exists on that beach with those levels, even if there was 15CPS.....
Just a few simple things that can bring up the background radiation:
- Soot/ash from coal-plants, actually it can be quite high in really bad stuff.
- Spill of potassium (fertilizer)
- Concrete
- Stone from a quarry that contained a bit higher PPM of uranium.
"That is double the amount of radiation normally found in the atmosphere in Britain" this is comparison is also completely invalid.. You don't measure the atmosphere, you measure the ground or material around you.
Normal background radiation is between 25-75 in the US, so 15 CPM is nothing...
Read http://xkcd.com/radiation/ and then take into account that 100 cpm is about equal to 1 microseivert per hour.
From above poster: ...
A trip from NY to LA would give you 40microsieverts/hr
During a normal day you will recieve about 10microsieverts/hr...
Well, nuclear causes less death's than any other energy-source..
http://www.forbes.com/sites/jamesconca/2012/06/10/energys-deathprint-a-price-always-paid/
http://www.geekosystem.com/coal-oil-nuclear-deaths-chart/
Or you could do a google yourself on "number of deaths coal oil nuclear"
The thing with nuclear-power is that everything happens at the same place and affects more people in one go..... And i prefer something that kills ~90 people per year over for example Oil that kills ~36000 per year... Or natural gas that kills ~4000 per year.. Even wind-power kills ~150 per year....
The problem is that it's public opinion that drives the direction of how we generate power, but the problem is that the general population don't have the knowledge to actually make an informed decision, and neither can i fully.
The thing is that nuclear-power, and there are many types of technologies, is probably the only thing that will be able to sustain the human population for the next 50 years until we can perfect fusion-power or something else that do not have the same impact.
If nuclear-power would still be seen positive by the general population we would also build new reactors that are safer instead of staying with the old reactors that have known safety issues.
Makes me sick of a justice system where a 11'yr child could be charged with a crime for something like this.... Does not matter if he was autistic or not, but autistic kids might even flip out a bit more if someone grabs them (not sure if that's the same for everyone?)....
Any police should be able to handle a 11'yr old kid verbally, and worst case taking a hold of his arms and then getting a kick or two from a kid is not really that bad.. If someone work as a police-officer they should at least be able to handle a few bruises when handling kids..
Kids are kids.. They flip out from time to time and it's normal..... The strange thing seems to be that the police that are sent to schools don't get any proper training on how to handle kids..... And working in a place where you have autistic (and other things too?) should require quite a bit more of education...
Well... would not really call that a continent...
92233720368547800/1296.30 = 71151523851383 ( $1296.3 was price per ounce i could find ) ... )
(71151523851383*28.3495231) / 1000 = 2017111769024 Kg's of gold. ( One ounce 28.3495231gram
http://www.wolframalpha.com/input/?i=92233720368547800USD+in+gold+volume
= 0.115 cubic kilometers... Or a cube with 486 meter sides.....
Will probably be quite hard to grow anything on it... But the biggest problem would probably be all the leprechauns running around on that island... Not to mention all the rainbows that would end there causing rainbow-blindness :)
Pay off?? For that amount i could basically buy any country in the world... Offer every citizen $5000000 to move out of the country, but only if at least 99% of the population leaves. Even if there are a few wealthy people that chose to stay there would be nobody that would take care of the infrastructure for them ;)
$92,233,720,368,547,800 / 6 billion = $15372286 per person in the world..
But there is a problem with having that much money and starting to spread it around would cause a devaluation of money.. And since it's so much it would spread and would affect all currencies in the world...... Ie... those $15372286 per person might be able to buy some dinner after some time...
And it would not matter if it's a fiat currency or gold-backed currency... The wanted resources (food/water/electricity etc) stay the same, it's just how the money is spread out.
Well... "You still have to pay income-tax on it this year... Next year you can deduct it if the issue has been resolved"
http://wiki.cyanogenmod.org/w/Devices#vendor="Motorola";
well... cyanogenmod is running on quite a few....
If the problem is cyber-warfare then there is a easy solution to the problem... Just don't frecking connect sensitive machines to it...
If you want a secure tunnel via the internet that is also 'easy' to make secure and more or less hacker-proof... The issue is when you want the sensitive machines/networks to be both secure *and* connected to the rest at the same time...
One line specification of something like that is not really real world so adding a few things...
- What about during initialization of the frame-buffer that will take 50ms? Is it ok to open screen and clear and set pixel or does the framebuffer have to be initialized with the specified pixel before output is activated?
- We have 3 hardware layers where layer 2 has this pixel and layer 3 displays error message. Error message X is defined as fullscreen and during error the application is paused. Should the error-screen set alpha on pixel 0,0 to show it or do we not have to conform to this since the application is in pause-mode?
- Hardware requires alpha value to be set. Layer one will show video. Spec also says that all graphics should have a alpha of 200, does this include this pixel? Is it 128,128,128 when having a black layer 1 or is it 128,128,128 when having a white layer 1 in that case?
Even a dead-simple specification like that can be broken down so much when adding a few more requirements that makes it impossible for anyone to think of all corner-cases.. And if they would have written a spec to that degree of accuracy no developer would be able read the whole specification and remember everything during development.
You usually also divide work up on multiple developers where each developer gets a part of the specification that he should implement, but that developer only sees his part and don't want or have time to read and understand the full spec of the software.
For each overlapping feature you have an exponential growth of sub-feature-specifications that would have to be written to be able to handle all corner-cases...
I was commenting on what the actual article was about... And saying that some games could be run this way on thin/thick clients...
It all depends on the available funds for the potential customers and the cost-difference between thin/thick-client and a real PC... So the prime-time for some specific tech might differ with many years between places... It all boils down to what people can afford to pay and the benefits they might get with it.
My bad experiences with thin clients are that sometimes an app that everyone is running on the server end crashes somehow, and there is no way to correct the situation except to restart the server the app runs on. Sometimes that could be done gracefully, other times, more frequently, everyone would need to exit their remote sessions, wait until further instructions, then log in again. If there are lots of folks logged in, and they all have to exit because one app on the back end crashed that can cause serious consternation.
That is why running a thick-client can be a very good alternative... The app runs on the actual client but without the need of local administration and storage..
Paying for a simple service like this can be good in places where people cannot afford to buy a PC, or don't want to shell out with lots (for them) of money.
The benefits of this is that even poor families can get access to a whole new world of possibilities and learning new things.... Education for the population is *always* good. Take a look at the OLPC project and what it's all about...
The wealth of a country does usually match the curve of the it's population education-level.. So as the population gains new skills their wages will go up and they can start afford real computers and this business-model will be over and/or converted to a ISP model for internet-connectivity..
Making a 3 server redundant cluster should be easy... Just depends on how the solution will look.. Dumb-clients can be tricky... Thick-clients just need some remote cluster-fs they run everything from and a file-share for user-contents.
With thin-clients: (dumb-clients)
Simple UPS - 1k Euro
6x Servers 2k each (3 extra for failures)
2x storage 2k each.
VMWare or similar solution with that has HA so each user actually have their session running on 2 machines in case of failures... (no idea of cost)
With thick-clients:
Simple UPS - 1k Euro (probably overkill)
3x Servers 2k each. (one extra for failures..)
2x Storage 2k each. (6Tb should be enough for 200 users unless they are storing movies and stuff)
= 11k Euro
No need for 24/7 tech's... Running a system like this would probably take up ~1-2h per day, on average, in administration (backup's/upgrades etc).... The rest would be on-call for system failures, and with a HA solution there would not really be a need of more than 8-24h 'time to fix'.... That's the beauty of thin/thick clients... If something happens for someone they just reboot and reload the standard image... User contents are just files on a file-share...
Oh, forgot... There are also some quite nice things that can be done.. Like time-staping the input of the user and having the server do corrections to correct for the lag.. Can result in some graphics-glitches, but if you have a latency of 30ms that should be minimal for the user..
Also, you could stream some of the graphics as pure OpenGL commands for better graphics for the most visible items..
H.265 could be a very nice addition to reduce the bandwidth needed for good video.. 4Mbit should be more than enough for good 720p.
Gaming in the cloud does not have to suck.. Strategy-games don't really require that low latency... It's mainly the FPS games...
Terminals can be good, but it all depends on how it's all deployed and what things that are used...
There are several versions... (my own alias'es for the different versions, don't yell :] )
1. Thin-client.. All it does it display an image sent from a server..
2. Thick-Client... Everything is run on the client but all code is stored on the shared server. CPU hungry app's can be run on the server.
3. Fat-Client.. Everything runs on the client and everything is stored on the shared server. The system-image is cached on the client for better performance. (overlay-fs or similar)
4. Thin-PC.. Everything runs on the client and everything is stored on the shared server. The system-image is cached on the client for better performance and user-data is stored on the client as well but mirrored to the shared server via a gluster/AFS or other cluster based file-system.
In terms of costs 1,2,3 can be fairly cheap.. In terms of performance options 3 or 4 are very good... Option 4 can also be used in the case when the main server is down.
With option 2,3 the user can also choose to install extra applications, but would need some type of block-deduplication to reduce storage-costs.
With option 4 the user can install extra applications into their local storage.
With option 1,2,3 the administration can be done centrally, with option 4 client-specific administration might be needed depending on the setup.
With option 1,2,3 all user-data is stored on centralized servers so replacing a terminal is simple and will not result in data-loss.
With option 4 user-data is stored on centralized server and on local HD. Can result in data-loss and can cause strange issues for the users depending on usage.
The balance you will have to make is if it's cheaper to get good nice and fast servers with redundancy or if it's cost-beneficial to have slower servers and faster clients and also how much the administration will cost for the different setups.. The more client's you have the more money you save on option 1 but also the less performance for the users..
You are probably thinking of putting a server-rack in the building or something? Not really needed... Lets say you have a big building with say 200 apartments... That could easily be taken care of with a couple of simple servers... dual cpu with 24-32 cores each.. 144Gb ram... A couple of internal disks for booting and applications and then some shared storage..... We bought a couple of servers (dual cpu with 24 cores and 144Gb ram 2x500Gb disks with raid-1 setup ) refurbished for about 4k euro about a year ago.....
If people are using simple terminals for web/mail and other such tasks it requires soo little in terms of performance you could get away with quite a small setup...
Some rough numbers with everything in the building..
10000/12/3/200 = 1.38 Eur for server-cost. (Counting on less than 1k Euro for cooling of those 2 servers. (2 x simple AC) )
300/12/3 = 8.3 Eur for terminal
100/12/3 = 2.7 Eur for cabeling (paid off after 3 years)
Salary for the admin 10 Eur (no idea what a normal salary is there) and he will earn (200*12*10) = 24000 Euro per year if everyone joins in.
Internet connectivity for the building 300/200 = 1.5 Eur. (you should be able to share a 100Mbit connection on 200 apartments. Don't know the price so estimated it to 300Eur..)
5Eur for misc things like equipment replacement, billing, rent of room for servers etc. (5*200*12 = 12000 Euro per year)
So total of 28.88 Euro per apartment per month.. Compare that to buying a computer for ~600Eur (16.6Euro per month + internet 15Euro = 31.6 Euro)...
Don't know the price for electricity there, but that should be fairly minimal and could probably be included in the "misc" group...
What i have not counted on here is licenses, but for mail+web+word-processing and other such simpler tasks you could get everything going with free software.... Even the "terminal" could in that case be used for some of the simpler processing like the web-browser.. Cheap Intel Atom with 2Gb of ram + monitor could set you back maybe 200-250Eur including monitor.. when bought in larger quantities... No idea about their ARM based machines the article wrote about...
I think i used some bloated numbers in my calculation, except for the salary since i have no clue about what would be normal there, but you could probably cut the server-cost in half.. 100
Also for the ones that buy's their own computers he could sell internet-connectivity for extra profit...Ie no terminals to lease etc, no need for extra cpu/memory in the server(s) etc....
Some other extra services he could provide at a small extra cost:
- Internet connectivity to people with their own computers.. (almost pure profit..)
- WiFi access in the whole building for phones/tablets etc. (almost pure profit.... add some costs for wifi-routers... Could be done on the cheap with commercial OpenWRT based solutions..)
- VoIP service to the whole building. (He could even rent them the wired/wireless phone..)
- VoIP-based Fax-service. If fax arrives it's printed on a shared fax and delivered to the apartment, or sent out as PDF to the person's mailbox. For outgoing faxes it can be done as PDF being sent or going to the shared printer/fax room and scanning the paper into the account the person has on the servers.
- Shared printer-service. 0.1Eur per sheet..
- Collaborating with other such residences offering WiFi to a large area and then sharing the profits for this extra service.
- Collaborating with other such residences buying a cheaper/faster connection, that is then shared between them.
- Offering backup-services of user-data. (1Euro per month or so... 1*200*12*3 = 7200Euro.. For that he could buy quite a nice disk-based solution + offsite tapes..)
- Collaborating with other such residences that are close by and sharing more of the infrastructure for lower costs..
But even if you don't have genetic patents they still have an incentive... Patent for drug X that affect gene Y in this manner.. Or Patent for cheaply and accurately detecting gene Z.
Genetic patents are not really good because they patent the actual gene so nobody else can make similar and better drugs or tests for that specific thing... There in lies the evil...
If we take genetically modified crops... They are patented and they are suing people that reuse seeds from their last harvest...... Now lets say Company A gets a patent where they insert some new genetic sequence into a person... If that person then gets a kid then he will be infringing too?? Or how about if a virus would absorb a genetic-sequence and then spread over a population... Will the infected be patent-infringers too??
Patents on genes should not exist. It's a really dangerous thing that actually hinders development of new treatments and blocks actual research.
Do a google and read up a bit on the issues with genetic-patents... To get you started...
http://www.guardian.co.uk/commentisfree/2013/apr/24/gene-patents-scientific-research-innovation
The power-outlet differs between countries, but they all use 240V so using a simple adaptor will work... Laptop's AC adaptors are usually designed for anything between 110-240V...
Internet in most countries in europe is quite good, and you can buy fairly cheap 3g prepaid cards for pure data too..... and loads of cafe's have free wifi too...
Some prices for Sweden : http://prepaidwithdata.wikia.com/wiki/Sweden
So buy a prepaid sim for the wanted duration and plug it into a cheap android phone and you have a perfect way to get online where ever you might be..