You miss the fact that using QT for proprietary apps means you'll have to pay, whereas that's not the case for Gnome. For most small or medium sized software companies that's no big deal.
As for support, I'm sure you'd find hundreds of people who know Gnome ready to enter into a support contract with you, including a significant amout of companies.
I don't agree. Pre-fabricated building components doesn't mean there are hardly any builders around. On the contrary: Before it might take decades to complete a large castle or cathedral, while today we can build skyscrapers that are many times larger in much shorter time. As a result building costs adjusted for inflation are dramatically down for comparable quality, and more building work is done.
A large set of types of work is limited not by a fixed demand, but by demand tempered by price, and when the price comes down demand tend to go up.
Look at all the new places computer systems have been introduced over the last decade, to a large extent because of rapidly falling costs of computing. This is just an extension of that.
I think the predicted shrinkage itself is unlikely. Possibly over the short term, but not for long. First of all, moving positions offshore increase the salary expectations in the target country, so the attractiveness of moving offshore will decrease until there is an equilibrium salary wise.
Secondly, moving positions offshore will always be restricted in certain ways: customer facing positions may need to remain in your original market, such as senior engineering positions and project managers, architects etc. Based on that it is unlikely that all of the rest of the team will be moved - it's hard to manage geographically split teams.
Thirdly, as long as there is a cost advantage to moving offshore, doing so will improve the competitiveness of the company doing the move and may even neccesitate more hiring locally to make partially or fully up for the positions being moved.
Fourth, moving offshore works mainly for larger companies where core engineering is functionally split from the people who do requirements gathering, onsite support etc. Most engineers are employed by small to medium sized companies that are unlikely to gain anything by moving anyone offshore (how do you move half a position? how efficient will it be to move 1,2,3 positions?).
I'm sure it will have an impact on salary expectations for low end engineering staff, but high end engineering staff will quickly become expensive in countries like India too, and as someone else has mentioned: fewer low level engineering positions might actually shorten the time it takes to rise up the career ladder for the really good entry level engineers who now often get stuck because the engineering teams often are very bottom heavy.
Another thing is that increased focus on making software development truly software engineering will result in a lot of new positions as companies need to put resources into improving quality through formalising development processes, introducing more stringent QA and documentation guidelines etc.
Last but not least, lowering the cost of custom software development will make it more affordable for companies to do custom development projects, which is where most software engineering jobs are - "shrinkwrap" software houses employ only a small percentage of engineers. Even if a lot of that new work will go offshore too, it will reduce the impact for engineers in the industrialized countries too partly by boosting offshore salaries and making it less attractive, and partly because some of it will end up locally as well.
This logic doesn't work in a free market. In a free market companies will not be competitive unless they choose the most efficient alternatives as long as there are competitors that will. Lack of open source only benefit large corporates, for which writing software in house for reuse can be done once with minimal cost averaged out over the business, while large corporates are not where the majority of people are employed.
Without open source, companies such as IBM, with hundreds of thousands of employees would share within the company and lower their costs, while the thousands and thousands of smaller companies that employ the majority of people would find it harder to compete because they would have to pay more salaries to write all this code themselves.
By reducing the competitiveness of small and medium sized companies, these companies would be less profitable and be able to pay fewer people.
While being inefficient will make a company need more people, it also reduces that companys chance of expanding and even of surviving, and hence is longer term bad for employment.
Society is much better off with increasing efficiency, as it increases capital return on investments which again makes it more worthwhile to invest in new ventures or in expanding existing ventures, and makes it more worthwhile to hire people.
Based on your arguments, developers should work as slow as they can, because it would result in a need for more people. However all that would achieve would be to drive those companies out of business or reduce their growth and prevent them from hiring more people in the long run.
You have a very limited idea about what a "useful" robot is. Yes, there are plenty of uses for robots that don't benefit from being humanoid. On the other hand, humands are used to interact with humanoids in a wide number of ways, and for uses where a robot will need extensive interaction with humans for instance, a humanoid robot could be very useful.
Large parts of our daily surroundings are also adapted for humanoids, including everything from height and width of doorways, to placement of doorhandles, placements of levers and buttons, shape of car seats etc. Robots designed to aid or accompany humans or serve humans as opposed to carrying out "industrial" tasks where the operator and the people the robot interact with can be assumed to be trained in operating it will need to be able to handle a substantial part of normal human surroundings and interact with them.
Your point being? You can reach continental Europe from London with the Eurostar. Train links to Russia from Europe are reasonably good. The Trans-Siberian railroad would presumably be extended to any hypotethical Bering Strait link (or the Bering Strait link would be useless). Similarly, on the other side, the Alaska side of a Berig strait link would have to be connected to the Canadian railroad for it to be useful, at which point you could in theory get quite far south in the Americas. Of course I say in theory, you'd have to deal with Amtrak on a significant part of the journey.
If you're an economist you should also realise that the tunnels ability to finance itself with through tickets is only a small part of the equation, and a major reason why it's unreasonable to assume that something like this will be based on private finance alone.
First of all, this will employ a lot of people. Salaries to contractors, sub-contractors and infrastructure around them will be a major factor in the total cost. All of these people pay tax. By reducing the number of unemployed in the area the Moroccan and Spanish government would be reducing the social welfare costs and increase their tax revenue.
Secondly, once completed the increased trade it brings will increase tax revenue, and also bring more employment.
These massively impact the real profitability of the project, while they won't show up in a profit and loss statement, and is a major reason why government involvement in these kinds of major infrastructure projects is warranted.
Looking only at moving people is short sighted. The real advantage is reduced cost of moving goods to and from all of western Africa, where the alternative is ship or plane. For goods shipped by boat today, a train link could save repackaging for delivery in continental Europe or parts of north western Africa.
The thing is registering copyright is irellevant in most cases, and WOULD HAVE in this case too if the product had been newer. Registering copyright is now practically only an issue if you wish to claim damages in a trial. Not registering your copyright is pointless if a) there is clear documentation available that you authored the work in question and b) you only intend to prevent further distribution in the case of infringement, not seek damages.
ObDisclaimer: I'm not a lawyer. This is not legal advice, and it's probably completely wrong:)
You have to realise the way this works: It's currently way too expensive research to be justified for personal transport. Military research budgets however are huge, and the safety demands (and hence costs) for unmanned vehicles are way lower than for pilotless planes with passengers. Hence the cost is lower and capital easier available for researching this as part of weapons programs. Once the technology is there, the pressure will start to commercialise the technology for civilian use.
I have problems enough tolerating the long boarding time and travel too/from the airport for ONE flight. Having to deal with connecting flight is a nightmare. Given the choice between one long flight and two shorter with transit time in between I'd choose the one long flight even if it ended up taking me the same amount of time or even slightly longer than the two shorter flights.
This is an area where pilotless planes and automated air traffic control could help greatly - anything that bring down the overhead of operating small planes with short runway needs that would be acceptable for small airports closer to city cores would make me a lot happier than the massive planned Airbus planes.
Re:35 min. NY to LA passenger flights? Keep dreami
on
The Future of Flight
·
· Score: 1
Oh, given current progress I fully expect to be able to get launched into orbit in casual clothes in my lifetime. "Just" flying three times faster than the SR-71 isn't much compared to that.
Sure it would be expensive, but we've already seen that several people have been willing to spend millions of dollars for a trip into space despite having to go through extensive training, being stuck in a cramped little Soyuz capsule, and not having anything to do up there.
I'd say you don't need to get it extremely cheap before you could start making money off it.
At 5 times the speed of Concorde, I assume that if you could get down to comparable costs you'd easily see the same passenger volume. And if it could achieve this high enough that sonic booms wouldn't be an issue it would be able to fly a hell of a lot more profitable long haul routes than London/Paris to New York and DC.
I fully expect to live at least another 40-50 years or so, and 50 years was enough to bring us from no aviation to transatlantic jets - I don't see something like hypersoar at "reasonable" (read "Concorde level") prices to be too unlikely within a timeframe like that. Especially given that there are multiple hypersonic projects already underway, and that there is enough private investment in private space travel that we should see results from that as well within the next decade or two.
Actually, you are wrong. Passports are identification, yes, but the biometrics on them is intended to help authenticate you as the rightful owner of the passport.
The assumption is that they can easily verify the passports validity against a database, so the problem is authenticating that you are who the passport says you are.
One of the goals is to AVOID having to do manual verification. Frequent travellers to the US has long been able to be fast tracked through immigration if they're willing to register with the INS and provide biometric information.
In that case, you get an identification card for identification, and use your biometrics as authentication that you are the holder of the card, just like the new passport systems.
As for your example, it's deeply flawed - the biometric information that is currently being considered (facial recognition for instance) will not uniquely identify you by far, so there would be huge numbers of non-unique entries in the databases.
And even with fingerprints, they are far from safe - a reasonable number of people (several percent) have fingerprints that are too worn down to be usable. More people share enough characteristics of their fingerprints that collisions WILL occur. Fingerprints works in forensics because they can do quick checks and then manually spend a lot of time analysing the details. It does not work to uniquely identify an individual without a significant risk of errors in a short timespan.
Facial recognition and hand recognition, which both rely on shapes, is quite unreliable if used for identification, and as such certainly not foolproof to use for authentication as well.
However it means you won't have to put your fingers on some plate, or step up to a scanner to have your irises scanned, which apparently causes more resistance because it reminds people too much about dystopian sci-fi and makes people think about police (the fingerprints) treatment of suspects, and brings up all kinds of worries about how criminals will want to rip out their eyeballs and stuff...
In other words, the selection of facial recognition is in part a PR excercise to make biometrics more palatable to the general public.
(I just got this image of taking this to the extreme, where the immigration officer smiling and in a silly faked happy voice asking if you'd just like to freshen up in the nice mirror right there and then getting very insistent if you say you're ok because thats where they've hidden the camera they pretend isn't there)
In some ways perhaps - CCTV cameras are very common here and used in ways that would never have been accepted in Norway for instance where I'm from. As an example, in central London a large part of the major roads are covered by CCTV, while in Norway there was an extended public debate about the installation of CCTV to cover just the area around the main railway station in Oslo.
In other areas, not. Contrary to many other countries, the UK doesn't have a central database of all citizens for instance. You don't need to have a passport. You don't need to be registered for national insurance (equivalent to being registered for social security in the US). You don't need to be registered with Inland Revenue (equivalent to the IRS), and in fact the Inland Revenue really don't do much if they don't have your right address (in Norway it's a criminal offense not to report to the tax authorities when you move...). You don't need to be on the electoral roll (needed if you want to vote). You don't need to have a drivers license. And even if you do, these registers aren't cross linked, and no non-governmental agencies have access to any of them except for the electoral roll, and if you choose it will only be accessible for very limited use such as credit checks.
You can easily build credit based on fake details, because most of the credit scoring is done based on your time living at a particular address (or rather, the amount of time you have been able to receive mail addressed to a particular address), and your behaviour towards other creditors, but if you manage to get one account with false details you'd easily get a new accounts based on your credit records.
The Inland Revenue doesn't need to, or want to, know about your bank accounts. You are responsible for reporting your revenue, not the details about what you have in which accounts (whereas in many other countries, including Norway, the banks are required to give the tax authorities details about all accounts you hold)
In general, the UK system is a whole lot looser coupled than what you'll find in quite a few countries. From the above, for instance, you can see that there is no one safe unique identifier you can use to identify a UK citizen, and since there is no one complete registry you can obtain most of the documents above if you manage to get hold of a couple of faked documents, and use them to build on eachother, while in Norway for instance you would need to find a valid, registered, personal identification number that match you reasonably well (birth year and gender is part of the number) to get anywhere, or you can live without most or all of the documents if you want to stay anonymous.
There is one significant difference: We have a reasonable expectation that in most cases that only limited information is kept about us. If the passport system contains detailed biometrics, there are no more technical barriers (such as lack of data with the current system) stopping the government of whatever country we're entering (including our own) from tracking our movements and in general invading our privacy in any way they see fit.
Yeah, I'm sure refugees running from oppressive regimes with strong security force presences in their embassies would see it as an "advantage" if their embassy automatically knew they were in they country. Israel and it's history of assassinations outside their own borders springs to mind (including at least one case where Mossad, the Israeli intelligence agency, murdered the wrong man in Norway because of a mistake).
Now there would be an incentive for faked passports or human smuggling.
The reason the UK, and the rest of Europe, is moving to biometric information is passport is that the US government demands that passports with biometric information start being issued to all citizens of countries that can enter the US without a visa.
So in the near future it's either biometrics, or having to apply for a visa to get into the US.
You miss the point. Implementing raytracing is "trivial". You can spend lots of effort at adding effects and handling materials realistically, and optimizing for speed, but a basic raytracing algorithm can be outlined on a sheet of A4 paper with sufficient detail for someone to implement it.
The news is that Blender now has built in support for it, so that you can raytrace your models to see how they'll look without messing with a separate program.
It's trivial to make a protocol reliable. You just increase the overhead A LOT by one or more of mechanisms such as ack's, checksums/hashes to verify integrity, resends, small packet sizes etc. The reason X/Y/Z modem got popular was exactly because they threw a lot of the overhead out, increased packet sizes etc. because for almost all uses it won't buy you anything.
But getting that level of reliability is dead simple - I wrote a custom protocol implementation for data transfer between two GSM handsets many years ago that had to cope with nastyness such as frequently dropped calls (due to one of the handsets being on a ship that was continuously circling over an autonomous submersible, and the other handset being on another ship, both of them well off shore:) ), and small packet sizes and predictive resends (packages would be resent automatically if it didn't get an ack or nack within a reasonable amount of time was all that was needed to make it "rock stable" (except for the delay caused by the GSM phones reconnecting).
Serial protocols aren't exactly hard to do unless you "need" to squeeze every last byte out of the theoretical maximum transfer speed available.
I hate to ask this, but WHY? Any FTP client worth anything can support server to server transfers (hmm... bet someone is going to take offence at that). It's the only thing that even remotely justifies the way the pile of shit that is the FTP protocol is structured.
Slashdot Mirror
As for support, I'm sure you'd find hundreds of people who know Gnome ready to enter into a support contract with you, including a significant amout of companies.
Heh. Mentioning Windows and criticizing KDE and Gnome's API's in one paragraph without truckloads of sarcasm is quite impressive.
A large set of types of work is limited not by a fixed demand, but by demand tempered by price, and when the price comes down demand tend to go up.
Look at all the new places computer systems have been introduced over the last decade, to a large extent because of rapidly falling costs of computing. This is just an extension of that.
Secondly, moving positions offshore will always be restricted in certain ways: customer facing positions may need to remain in your original market, such as senior engineering positions and project managers, architects etc. Based on that it is unlikely that all of the rest of the team will be moved - it's hard to manage geographically split teams.
Thirdly, as long as there is a cost advantage to moving offshore, doing so will improve the competitiveness of the company doing the move and may even neccesitate more hiring locally to make partially or fully up for the positions being moved.
Fourth, moving offshore works mainly for larger companies where core engineering is functionally split from the people who do requirements gathering, onsite support etc. Most engineers are employed by small to medium sized companies that are unlikely to gain anything by moving anyone offshore (how do you move half a position? how efficient will it be to move 1,2,3 positions?).
I'm sure it will have an impact on salary expectations for low end engineering staff, but high end engineering staff will quickly become expensive in countries like India too, and as someone else has mentioned: fewer low level engineering positions might actually shorten the time it takes to rise up the career ladder for the really good entry level engineers who now often get stuck because the engineering teams often are very bottom heavy.
Another thing is that increased focus on making software development truly software engineering will result in a lot of new positions as companies need to put resources into improving quality through formalising development processes, introducing more stringent QA and documentation guidelines etc.
Last but not least, lowering the cost of custom software development will make it more affordable for companies to do custom development projects, which is where most software engineering jobs are - "shrinkwrap" software houses employ only a small percentage of engineers. Even if a lot of that new work will go offshore too, it will reduce the impact for engineers in the industrialized countries too partly by boosting offshore salaries and making it less attractive, and partly because some of it will end up locally as well.
Without open source, companies such as IBM, with hundreds of thousands of employees would share within the company and lower their costs, while the thousands and thousands of smaller companies that employ the majority of people would find it harder to compete because they would have to pay more salaries to write all this code themselves.
By reducing the competitiveness of small and medium sized companies, these companies would be less profitable and be able to pay fewer people.
While being inefficient will make a company need more people, it also reduces that companys chance of expanding and even of surviving, and hence is longer term bad for employment.
Society is much better off with increasing efficiency, as it increases capital return on investments which again makes it more worthwhile to invest in new ventures or in expanding existing ventures, and makes it more worthwhile to hire people.
Based on your arguments, developers should work as slow as they can, because it would result in a need for more people. However all that would achieve would be to drive those companies out of business or reduce their growth and prevent them from hiring more people in the long run.
"I, Robot". As in the Isaac Asimov short story collection.
Large parts of our daily surroundings are also adapted for humanoids, including everything from height and width of doorways, to placement of doorhandles, placements of levers and buttons, shape of car seats etc. Robots designed to aid or accompany humans or serve humans as opposed to carrying out "industrial" tasks where the operator and the people the robot interact with can be assumed to be trained in operating it will need to be able to handle a substantial part of normal human surroundings and interact with them.
Deceased Film Four? I watched Film Four as late as last night... Unless you're talking about a different Film Four they are very much alive.
Your point being? You can reach continental Europe from London with the Eurostar. Train links to Russia from Europe are reasonably good. The Trans-Siberian railroad would presumably be extended to any hypotethical Bering Strait link (or the Bering Strait link would be useless). Similarly, on the other side, the Alaska side of a Berig strait link would have to be connected to the Canadian railroad for it to be useful, at which point you could in theory get quite far south in the Americas. Of course I say in theory, you'd have to deal with Amtrak on a significant part of the journey.
First of all, this will employ a lot of people. Salaries to contractors, sub-contractors and infrastructure around them will be a major factor in the total cost. All of these people pay tax. By reducing the number of unemployed in the area the Moroccan and Spanish government would be reducing the social welfare costs and increase their tax revenue.
Secondly, once completed the increased trade it brings will increase tax revenue, and also bring more employment.
These massively impact the real profitability of the project, while they won't show up in a profit and loss statement, and is a major reason why government involvement in these kinds of major infrastructure projects is warranted.
Looking only at moving people is short sighted. The real advantage is reduced cost of moving goods to and from all of western Africa, where the alternative is ship or plane. For goods shipped by boat today, a train link could save repackaging for delivery in continental Europe or parts of north western Africa.
RTFA. The area chosen for the crossing was chosen specifically because the strait is only 300m deep there.
ObDisclaimer: I'm not a lawyer. This is not legal advice, and it's probably completely wrong :)
You have to realise the way this works: It's currently way too expensive research to be justified for personal transport. Military research budgets however are huge, and the safety demands (and hence costs) for unmanned vehicles are way lower than for pilotless planes with passengers. Hence the cost is lower and capital easier available for researching this as part of weapons programs. Once the technology is there, the pressure will start to commercialise the technology for civilian use.
This is an area where pilotless planes and automated air traffic control could help greatly - anything that bring down the overhead of operating small planes with short runway needs that would be acceptable for small airports closer to city cores would make me a lot happier than the massive planned Airbus planes.
Sure it would be expensive, but we've already seen that several people have been willing to spend millions of dollars for a trip into space despite having to go through extensive training, being stuck in a cramped little Soyuz capsule, and not having anything to do up there.
I'd say you don't need to get it extremely cheap before you could start making money off it.
At 5 times the speed of Concorde, I assume that if you could get down to comparable costs you'd easily see the same passenger volume. And if it could achieve this high enough that sonic booms wouldn't be an issue it would be able to fly a hell of a lot more profitable long haul routes than London/Paris to New York and DC.
I fully expect to live at least another 40-50 years or so, and 50 years was enough to bring us from no aviation to transatlantic jets - I don't see something like hypersoar at "reasonable" (read "Concorde level") prices to be too unlikely within a timeframe like that. Especially given that there are multiple hypersonic projects already underway, and that there is enough private investment in private space travel that we should see results from that as well within the next decade or two.
The assumption is that they can easily verify the passports validity against a database, so the problem is authenticating that you are who the passport says you are.
One of the goals is to AVOID having to do manual verification. Frequent travellers to the US has long been able to be fast tracked through immigration if they're willing to register with the INS and provide biometric information. In that case, you get an identification card for identification, and use your biometrics as authentication that you are the holder of the card, just like the new passport systems.
As for your example, it's deeply flawed - the biometric information that is currently being considered (facial recognition for instance) will not uniquely identify you by far, so there would be huge numbers of non-unique entries in the databases.
And even with fingerprints, they are far from safe - a reasonable number of people (several percent) have fingerprints that are too worn down to be usable. More people share enough characteristics of their fingerprints that collisions WILL occur. Fingerprints works in forensics because they can do quick checks and then manually spend a lot of time analysing the details. It does not work to uniquely identify an individual without a significant risk of errors in a short timespan.
Facial recognition and hand recognition, which both rely on shapes, is quite unreliable if used for identification, and as such certainly not foolproof to use for authentication as well.
However it means you won't have to put your fingers on some plate, or step up to a scanner to have your irises scanned, which apparently causes more resistance because it reminds people too much about dystopian sci-fi and makes people think about police (the fingerprints) treatment of suspects, and brings up all kinds of worries about how criminals will want to rip out their eyeballs and stuff...
In other words, the selection of facial recognition is in part a PR excercise to make biometrics more palatable to the general public.
(I just got this image of taking this to the extreme, where the immigration officer smiling and in a silly faked happy voice asking if you'd just like to freshen up in the nice mirror right there and then getting very insistent if you say you're ok because thats where they've hidden the camera they pretend isn't there)
In other areas, not. Contrary to many other countries, the UK doesn't have a central database of all citizens for instance. You don't need to have a passport. You don't need to be registered for national insurance (equivalent to being registered for social security in the US). You don't need to be registered with Inland Revenue (equivalent to the IRS), and in fact the Inland Revenue really don't do much if they don't have your right address (in Norway it's a criminal offense not to report to the tax authorities when you move...). You don't need to be on the electoral roll (needed if you want to vote). You don't need to have a drivers license. And even if you do, these registers aren't cross linked, and no non-governmental agencies have access to any of them except for the electoral roll, and if you choose it will only be accessible for very limited use such as credit checks.
You can easily build credit based on fake details, because most of the credit scoring is done based on your time living at a particular address (or rather, the amount of time you have been able to receive mail addressed to a particular address), and your behaviour towards other creditors, but if you manage to get one account with false details you'd easily get a new accounts based on your credit records.
The Inland Revenue doesn't need to, or want to, know about your bank accounts. You are responsible for reporting your revenue, not the details about what you have in which accounts (whereas in many other countries, including Norway, the banks are required to give the tax authorities details about all accounts you hold)
In general, the UK system is a whole lot looser coupled than what you'll find in quite a few countries. From the above, for instance, you can see that there is no one safe unique identifier you can use to identify a UK citizen, and since there is no one complete registry you can obtain most of the documents above if you manage to get hold of a couple of faked documents, and use them to build on eachother, while in Norway for instance you would need to find a valid, registered, personal identification number that match you reasonably well (birth year and gender is part of the number) to get anywhere, or you can live without most or all of the documents if you want to stay anonymous.
There is one significant difference: We have a reasonable expectation that in most cases that only limited information is kept about us. If the passport system contains detailed biometrics, there are no more technical barriers (such as lack of data with the current system) stopping the government of whatever country we're entering (including our own) from tracking our movements and in general invading our privacy in any way they see fit.
Now there would be an incentive for faked passports or human smuggling.
And to add to that, the reason you got checked in the UK is that the UK isn't a member of Schengen, which is the passport cooperation.
So in the near future it's either biometrics, or having to apply for a visa to get into the US.
The news is that Blender now has built in support for it, so that you can raytrace your models to see how they'll look without messing with a separate program.
But getting that level of reliability is dead simple - I wrote a custom protocol implementation for data transfer between two GSM handsets many years ago that had to cope with nastyness such as frequently dropped calls (due to one of the handsets being on a ship that was continuously circling over an autonomous submersible, and the other handset being on another ship, both of them well off shore :) ), and small packet sizes and predictive resends (packages would be resent automatically if it didn't get an ack or nack within a reasonable amount of time was all that was needed to make it "rock stable" (except for the delay caused by the GSM phones reconnecting).
Serial protocols aren't exactly hard to do unless you "need" to squeeze every last byte out of the theoretical maximum transfer speed available.
I hate to ask this, but WHY? Any FTP client worth anything can support server to server transfers (hmm... bet someone is going to take offence at that). It's the only thing that even remotely justifies the way the pile of shit that is the FTP protocol is structured.