In the last five years I must have read about at least fifty breakthroughs in battery technology, but nothing of it has reached the consumer (me) yet.
I believe that this is because researchers seem to exaggerate their research results for obvious reasons and seem to underestimate what it takes to make a successful product.
Regarding battery technology I completely stopped to believe anything that comes out of the research community.
The unit kilowatt is fine, but the number is ridiculously low. Cars breaks are typically designed to have a breaking power of four times the engine power, so we are talking about 500-1000 kW of breaking power in typical cars.
Modern high speed trains have a breaking power in the order of 10 - 20 MW using their engines for regenerative dynamic braking. The german ICE3 has about 16,4 MW dynamic breaking power, which is only slightly higher than the 16 MW propulsion power. It also has additional eddy current brakes, but they there breaking power is just 1600 kW, about 10% of the the dynamic breaking power. There are even additional disc brakes, but they are only used in emergency situatuions.
The largest burner on my gas stove has 5 kW power, but I don't think I can make a pan to glow in seconds, if at all.
Watching the video I would assume that they are talking about 4.6 MW oder 4600 kW of breaking power this disk can handle.
I do not see why the burden of proof that massive dissemination of poison is harmful should be with the public.
IMO those who manufacture and sell this stuff have to prove that it does not destroy our ecosystem.
I know, the stuff has been at some point been certified, but I think that every company that manufactures a product has an obligation to monitor if it is harmful even after it appears on the market. You simply can not determine the long term impact of wide use on the environment with a handful of studies,
If the game state is maintained and processed on "their" computer, it is a service. Otherwise I would regard it as a "good" that is executed on my computer.
The "Cubic Mile of Oil"-numbers do show the opposite of what you are suggesting.
Replacing one CMU with nuclear is the second most expensive alternative, only solar panels are more expensive. Building 2200 nuclear power plant is not only totally out of question, it would also deplete world uranium reserves within a decade.
On the other Hand, building 1.6 Mio. wind turbines is the cheapest viable alternative. Germany alone already has about 22.000 of them, and they are profitable.
Safe nuclear power is not commercially viable. The only reason why it looks cheap is because today's commercial reactors are unsafe by design and the risk costs are carried by the society.
With the cheap nuclear reactors that are still being build today we will have a major nuclear disaster every 20-30 years, and the economic damage to the affected country is huge.
Depending on the outcome, Fukushima probably will cost Japan more than has ever been saved by using nuclear power at all. Expect Japan to pay billions every year just to maintain the Fukushima ruin, and this might go on for hundreds of years.
Just one Fukushima or Chernobyl type disaster in Germany would cost more than the transition to renewable sources.
Of course a society can make the decision to take the gamble, but Germany has been hit hard from Chernobyl Fallout - in some areas in Bavaria deer still can not be consumed because the meat is too radioactive, and it will take about 200 years until the situation will be normal.
The U.S. may not have such big problem when a few thousand square miles get polluted with radioactive fallout - the country is big.
For a small country like Germany, one nuclear disaster might cause more damage than World War II.
My guess is that within the next 30 years we will see another major nuclear disaster somewhere in the world, most probably in the U.S. or in France.
I also expect that most planned nuclear reactors never will get built because even before Fukushima they were too expensive, and after Fukushima no sane Investor will sink money into it.
Like about 80% of the german population I regard the decision to end nuclear power in Germany as a good one - not because I am afraid, but because it is an economically sound decision.
Reading this article, I was thinking this security guy is exaggerating and playing down at the same time.
First of all, in the U.S. many companies use so much crap when it comes to IT that it makes me sick, so everything is possible. However, I think it is much more probable many systems will blow up on a large scale without any malice involved, but just due to incompetence and negligence.
At the same time this guy admits the U.S. is actively preparing and maybe even conducting cyber-warfare against other countries. I don't know how to comment on that. If all countries would stick to cyberwarfare instead of dropping bombs, this would not be *that* bad.
The talk about stolen intellectual property and trade secrets is mostly bullshit. Any business that requires a great deal of secrets to be kept is not sustainable anyway. The future belongs to companies who need very few secrets, if any at all, and are quite open about most aspects of their business. Secrets tend to get out sooner or later anyway.
For mission critical software the quality standards should not be very high, but insanely high. And when the life of people is on the line, software alone should never be able wreak havoc. Unfortunately there are too many people out there who don't have a clue and are just happy when things work. The only get wiser when after the shit hits fan a couple of times, but then they overreact. Professionals should have more courage and never let hazardous systems become operational.
However, I don't see a chance that most of those responsible for the bottom line would voluntarily invest in security and safety unless they are forced to do so, either by law, or by shitstorm.
I agree with you that the Browser Issue is a secondary one.
But no one would deny that Microsoft has a de-facto Monopoly on desktop operating systems and office applications.
Bill Gates did not become the richest person in the world because Microsoft did a good job on innovating, but they successfully succeeded the IBM Mainframe monopoly in the PC market on the back of IBM.
The only great thing Microsoft did was to royally kick IBMs testicles when they felt they could do so safely.
Then they operated for decades with an paranoid underdog mentality, and because they knew that their technology and their people were at best second-rate, they thought that every new small competitor might do to them what they had done to IBM, so they used every dirty trick in the book to eliminate upcomers.
And at the same time they enjoyed their monopoly and charged way too much for their software.
Software has a wonderful economy of scale, and with software you can create wonderful lock-in effects, and also abuse the patent and copyright system to keep competitors away.
By the number of windows licenses, a copy of windows should not cost more than a single digit amount of dollars, and Office as well.
Microsoft should have been forced to open its books and all the indecent profits should have been taxed away. Price controls should have been enforced based on the actual costs of developing and marketing the stuff, divided by the number of copies sold.
What happened here is a classical example of how laissez-faire capitalism hurts the economy, stiffles innovation and makes a few obscenely rich.
Now, before calling me a communist:
I regard Communism and Capitalism as equally inhumane because they both deny and suppress half of what makes us survive and prosper. Every sane human being has not only the desire to maximize his own profit, but also to give away, to share and to sacrifice for the prosperity of the society he lives in.
A society or culture that focuses on one side is simply doomed.
Yeah, sure, those terrorists capable of turning an iPhone into a cruise missile will be stopped in their tracks by this measure. The would never dare to smuggle in an iPhone *with* GPS, and would never get the idea to buy a $20 GPS module for this purpose.
No, your scenario is complete nonsense. There are much simpler and cheaper solutions when you want an programmable embedded GPS system.
However, what you can do with an "off the shelve" iPhone is to it use as tactical computer to coordinate the movement of you strike teams in real time. And with the iTerrorize-Plugin from the AppStore every iPhone-owner can sign up for a suicide mission on behalf of any organization. He just needs to bring his own weapons and explosives.
Seriously, a society that fails to provide a meaningful life in dignity for too many talented people will eventually get in serious trouble. You want the talented people to work for the establishment, not against it.
Restricting and regulating the use of some technology does make sense, but isolating your society from progressive and wide-spread technology used by the rest of the world will weaken your economy and make your country an easy prey for your neighbors in the long run. I don't believe that restricting the use of GPS in today's world can have any positive effect.
You are right, this claim might be a bit far fetched, but the "Universal Declaration of Human Rights" gives you the following rights:
Article 1 All human beings are born free and equal in dignity and rights. [...]
Article 19 Everyone has the right to freedom of opinion and expression; [...]
Article 23
3. Everyone who works has the right to just and favourable remuneration [...]
Article 27
2. Everyone has the right to the protection of the moral and material interests resulting from any scientific, literary or artistic production of which he is the author.
Article 29
1. Everyone has duties to the community in which alone the free and full development of his personality is possible.
2. In the exercise of his rights and freedoms, everyone shall be subject only to such limitations as are determined by law solely for the purpose of securing due recognition and respect for the rights and freedoms of others and of meeting the just requirements of morality, public order and the general welfare in a democratic society.
There is of course no right that says "All humans have the right to put Easter Eggs in their products", but taking all these above rights together, when you create something extraordinary, you are entitled to "protection of the moral interests", and matters like human dignity, freedom of expression and development of personality under circumstances may imply this right, unless of course you do no harm and respect the rights of others.
I don't believe you will be very successful in a U.S. court with this argumentation. Arguing with human rights in court generally does not work well in most countries of the world.
But the fact that there are so many easter eggs around and I never heard about anyone getting in trouble because of it indicates that most people, be it customers, superiors or colleagues do appreciate or tolerate it.
but if you are proud of what you have achieved, go ahead, it is a human right. I did it a couple of times when it seemed appropriate, and it was never a problem. If you are in a position where you are responsible for creating the system, it is just your decision.
If you need a rational argument for doing it: It might make the team more proud of the product, more productive for future releases, projects or products. A company is about people, and it is a people issue.
And when even weapon designers can put easter eggs into missiles, I can't imagine anything to be too serious to pull it off; a product just may be too boring or suck too much to deserve an easter egg.
An easter egg is just a sign that the people who made the product did care about it, and are proud of it. An easter egg is like a medal or decoration for the product.
This guy, Lutz Heilman, would be right now the least liked politician in whole Germany, if someone would care. Even his own party does not like him because he did not properly disclose that he worked for the Staatsicherheit, the secret police in Communist Germany. His political career is fucked since 2005, when the German news magazine "Der Spiegel" disclosed it, and now he probably has nothing to loose. The facts in the wikipedia are based on solid sources, so I don't think the court will decide in his favour. And not that it really matters in this case, but he is also avowed gay and a lawyer.
It is all more smoke than fire, only forwarding from www.wikipedia.de to de.wikipedia.org is affected, you can still access wikipedia under de.wikipedia.org from Germany. So actually not much is blocked, and when you go to www.wikipedia.de, you are even given the URL that still works.
And it is only a preliminary injunction, which probably will be lifted soon. It has happened already, but this futile symbolic act will probably just last a few days. In Germany, the courts do generally favor the right of every person to defend himself against defamatory speech, and are often quick with such injunctions to limit the damage the press can do to individuals.
It is a conflict between different human rights, the right to free speech, and the right of any human individual to live in dignity.
Actually the the Article 1 of the German constitution is one of the most beautiful and touching sentences I have ever seen in any piece of legislation:
"Human dignity is inviolable."
Free speech is important, but human dignity comes first.
Although the whole thing is stupid, I can not say the court did act on behalf of the powerful to suppress free speech, on the opposite, it did what it could to protect the rights of an politician who is not in high esteem with any of the established parties. The guy is from a small left wing party, the successor of the communist party of east germany. They get around 5% of the votes in whole Germany, and up to 20% in some parts of the former east.
The more you build on top of something, the more solid the base should be.
If your own stuff is just a few thousand LOC, it does not matter that much if you are locked into a particular framework. You can still do it again with another framwork.
However, if you investment is a major one, you better roll your own or use something rock solid, that even has different implementation with a similiar API or functionality, like a database or script interpreter.
And when in doubt, roll your own or built an abstraction layer.
Listening through the whole podcast after reading many comments here, I found most people here did not get the point.
Their stateless SYN-Handshake just makes the described types of attacks even more resource-efficient for the attacker, but I assume that it is even feasible when using "normal" SYN-ACK connection estabishing.
The point is, that the attack starts *after* a connection is set up, using all types of other flow-control mechanisms in order to exhaust specific resources of the tcp-stack.
As an example I just made up, after the connection is open, you could use a flood of ACKs combined with a particular order of out-of-sequence packets or fragmented packets to make the system use up a lot of timers until the whole tcp-stack or the kernel goes down the drain. But as mentioned in the podcast, Timers are just one of a large number of different resources involved in housekeeping after a connection is established, and you can craft attacks to exhaust a specific set of these.
This opens up a complete new can of worms and may turn out not to be easily thwarted because you have to guard *all* of your resources, not just those involved in establishing a connection like in a SYN-attack.
Noone currently has an idea how serious this will turn out to be in the wild, but to me it sounds serious enough not to be ignored, and it has at least the potential to kill the open internet as we know it. In some distant future, you might need an authorization to even send a packet to a particular host, or the whole internet will be so closely monitored that anyone interfering will be visited by black helicopters immediately.
Btw, there is a cool protocol called FLIP (Fast Local Internet Protocol), designed by Tanenbaum in the early 90s for the Amoeba operating system, that would kill many of the problems we have today on the Internet, and it seems to be immune against many types of attacks, including eavesdropping, but FLIP, as the term "Local" in the name says, works just on one local Ethernet, using TCP-Tunnels to connect to other Ethernets, so it is not a viable replacement in its current form and might not scale and has probably a number of issues on its own. What is however remarkable that in FLIP even the source and destination address and port are encrypted, so you can not get even a single packet through if you don't have the permission, and you don't know who is talking to whom even when you sniff the traffic. Although it won't be able to replace TCP/IP, the concepts are an interesting starting point when thinking about a future secure internet. Just imagine every router on the internet would be a kind of TOR-Node, and you could transparently adress every running process on the internet in a secure fashion, no matter where is physically resides, and you could even multicast and have group communication on the network level, reaching millions of host with just sending one packet, but it will arrive only with their prior permission, which you give by just joining a communication group. Imagine a kind of PGP on the network level. This is what a future internet could possibly provide.
For now, we are stuck with TCP/IP, and maybe for a very long time, unless this kind of problems will finally bring the Internet down, which I think is totally possible because TCP/IP is designed for a network of hosts requiring at least some trust and playing fair on the TCP level.
As far as I remember, we did offline data store checks whenever we encountered problems, which became more frequent as the database grew and aged (about once a month in the beginning, about every week in the end). We also moved the database to at least three different servers because we first assumed the hardware might be the problems.
Anyway, your experience seems to confirm my observations: Comparable Microsoft engineered systems seem to deteriorate faster than *nix-based systems. I find it unacceptable when a system needs so much attention and maintenance just to stay alive, especially when this kind of maintenance involves downtime, and manual maintenance always carries the risk that the operators make errors, and as they are just humans they sooner or later will.
Anyway, there is no excuse for a repair tool to crash, as it should be *designed* exactly for dealing with corrupt data. However, Microsoft has not a monopoly on such a behavior, many years ago I dropped ReiserFS because of the same combination of corruption and crashing repair tools.
I think that the hands-on approach and informal culture that gave Microsoft such an advantage over a scientific and buerocratic style of engineering that was commonplace in the mainframe and workstation world came back to haunt them and their users when it came to serious enterprise computing.
I think such fast and reliable system is always very difficult to build, and those guys who build are top notch engineers, I am sure about that. They guarantee 10ms transaction time and promise to go down to 3ms. Given that this involves network communication with a third party and some arbitration and transaction logging, this is quite cool.
However, no one in the world is able today to create a bug free system, and sometimes shit happens. My guess is that someone tried to save on maintenance costs, and when the system broke down, they obviously did not have the right guy in place who is able to find out whats wrong and fix it. Shit happens, but when it happens, you want a fast and efficicient troubleshooter right on action.
The problem might be windows indeed, but not because it is less realiable. I think with windows there are two problems:
1)It is difficult to understand because the source is not widely available, although the guys involved in this project would have access to it.
2) It accumulates entropy faster than a Unix-based system; this means you have to reboot a windows system more often to remove entropy and by putting the system into a more defined state. Linux and BSD share this problem, and while a typical linux or BSD is only rebooted to install kernel updates, you do more often reboot a windows system because it behaves strange and you have no idea why, especially with desktop systems, but even more with laptops in changing network environment.
On many *nix systems you can find out by reasoning how to fix a problem. In windows, you may find out by reasoning what is wrong, but you have to *know* what to do in this situation, you can often not find it out by reasoning.
This may have contributed to the problem in London, which did cost them two 9s of their five nines they were claiming.
Our company moved away from Microsoft Exchange when we had a two day mail outage because of Exchange database corruption. We had backups and logs and a proven recovery process, but this time it failed. By replaying the logs, corruption occured even on copies of several older backups of the database. We settled for one day of lost E-Mail traffic to bring back the system.
I wanted to stay with Microsoft Exchange because of the hazzle and costs of moving to a different mail and calendar system, but one thing was paramount: I was determinded to make that never happen again, not because of the material damage, which was negligable, but my reputation as a seasoned professional technician who moved into upper management was at stake.
I was very surprised when I found out that no matter how much money I would be willing to spend on hardware, licenses and the best experts, there would be no way to recover from Exchange database corruption once it occurs better than we did. When we did log replays, the Microsoft recovery tool crashed near completion after one hour of replaying logs, leaaving us with an unusable databese that caused the Exchange Server to crash when started. We finally had to go back many weeks in the backup cycle and replay Gigabytes of logs only to loose just one day of mail. Even after this recovery, the server continued crashingon different hardware, so we were sure it was the software and a possibly corrupted database and broken repair tools.
So all the experts I consulted told me: the situation we ran into was not supposed to ever happen, and it would propably never happen again, but they could not rule out it might happen some day again, and when it would happen again, we would be fucked again.
The first lessen for me was: When the software contains bugs that corrupt the database, and the repair tool is broken, and corruption is not detected by the running software or during backup, and only crashes the server when some someone accesses one of the broken items, then you are really, really fucked by Microsoft, without any chance. And I was lucky, my reputation was just slightly damanged, but other people might easily loose their job because of
According to http://arxiv.org/abs/astro-ph/0012399, the average earth in the universe is just 1.8 +/- 0.9 billion years older than our earth, which means that other earth-like planets have just a few billion years more to evolve, and while it took five billion years to create intelligent life on our earth, we can not tell whether this is a long or short time compared to other planets. There may be some special factors like our moon that could have accelerated this process, so it might be possible that planets without massive moons take 10 billion years to develop intelligent life.
I am not saying that this is very probable, but I think it is not "improbable enough to dismiss out of hand". If we were first in our galaxy, it would be indeed surprising, but it would be also a funny explanation for our situation.
You can also not take for granted that even if there is intelligent life, an industrial revolution will take place, and we do not know yet whether it is sustainable long enough to evolve sustainable space colonization.
- interstellar distances are simply to big to be overcome, or no one wants to pay for the energy
Doesn't seem likely. We can do some temporary cost analysis, and the cost isn't prohibitive in any way. Well, not for a civilization a little more advanced than ours any way. It seems reasonable that we will advance just a little more, and that if we are left alone we will probably conquer interstellar space. And then the black hole at the center of our galaxy wakes up. Does a minor Hiccup and goes back to bed. That hiccup sterilizes or galaxy though, tearing all organic molecules apart. Oooops. Sorry. Indigestion. Just swallowed a couple of neutron stars and didn't have any antacid close by. Sorry.
According to my own calculations, to accelerate a 600.000 ton spaceship (including fuel) to near light speed and decelerate, you need to turn about 400.000 tons of its mass into kinetic energy. A small robotic probe of course might be much cheaper, and travelling in a small craft in a hibernated state also might be the better option, but I would not want to travel in a conscient state for years in a structure that is smaller than a large ocean cruiser.
I agree that the cost is not prohibitive if we find a more efficient way to turn mass into energy than fusion or fission, like antimatter annihilation, but without the need of putting 400.000 tons of antimatter into our gas tank. (I think there would be a law against bringing substantial amounts of antimatter into a solar system.) However, I can envision a process where matter is heated to early big-bang temperatures and turned into a quark-gluon plasma, that will exhibit almost total matter-antimatter annihilation on cool-down. (Baryogenesis and Leptogenesis) We may however have some containment problems with our "reaction chamber" here, requiring some more "breakthrough technology".
Although this might be possible in theory, we do not know if this ever will be possible in practice, here or in any other solar system.
Btw, I do not believe that FTL-travel ever will be possible, all our knowledge speaks against it.
- evil alien predators are already on their way to blow us out of the sky before we become serious competitors
Improbable. All experience show that competitive cooperation is more profitable then warfare. It seems unreasonable to assume that a space faring civilization with interstellar capabilities has not discovered this.
I wish this would be true, but my experience and knowledge of history indicates otherwise. The industrial revolution happened in Europe not because of competitive cooperation, it was driven by four hundred year
Although improbable, there has to be a first civilization in our galaxy, and maybe we are the most advanced life form in our galaxy so far. Regarding the age of the universe, it is possible, because the matter of the solar system had to go through 2 or 3 supernovae before having enough high period elements, and it took the universe about 5 billion years alone to create us since our planet was created, which is more than a third of the age of the universe.
This leaves us with a number of possibilities:
- we are the first - intelligent life is very rare or very fragile and volatile in our universe - the guy is right and they are already here and among us (maybe they are criminals hiding on a primitive planet) - interstellar distances are simply to big to be overcome, or no one wants to pay for the energy - evil alien predators are already on their way to blow us out of the sky before we become serious competitors - we live in a zoo and are just being watched - everything is just a big illusion, and our reality is only virtual
that fortunately does not exist in Germany. Here the law is simple: A company that wants an N.C.A. to be enforceable, it has to pay at least 50% of the former wages of the employee, otherwise the N.C.A. is void. It also has to be very specific, the new company must be competitor, being an IT-company is not enough, you basically have to provide the same product to the same custumers. It is also limited in time to one year.
When I once left a company that didn't want to let me go I happily told them I would love to sign an N.C.A., but when they saw what it would cost them and would bring them (I would be gone anyway), they quickly reconsidered.
In a situation when someone acts like an asshole the best thing is to make this person to understand that immediately and without any chance to get it wrong.
Human are in possesion of archaic communication channels that get directly through and can not be ignored.
It is called emotion, and in such a situation it is appropriate to get just frantic. Take a deep breath and start to cry out as loud as you can that you are fed up with such an selfish and antisocial behaviour, and call this guy whatever actually comes into your mind.
Many people, especially geeks have problems to show their emotions and consider it a weekness, but emotions are a very powerful and efficient mean of communication and immediate persuasion.
We are beeig told to be cool and stay calm, and talk things over, but there are situations where you can talk for hours and you just don't get through. You state you irrefutable arguments over and over again, but the other person lives in a different reality and filters or bends off everything you say.
But then, if you show true emotions, you will get through, no matter how hard-boiled or ignorant the other person is. He does not even need to speak your language.
However, it is a good idea to practise a bit, to learn not how to suppress emotion, but how to let out controlled burst of varying intensity. With mastering three or four different levels between "beeing normal" and "totally freaking out" you have enough granularity to master most situations without excessive collateral damage.
You should also try to observe how you react to emotions of others and how hard or impossible it is to ignore them.
However, a beancounter needs some figures in terms of money.
A possible approach would be to estimate the value of the IT-Infrastructure to function properly, and also quantify the risk of security breaches or data loss.
Other factors are the number of servers, clients and users, the number of services, the number of applications, amount of data storage and the amount of internal and external network traffic and the diversity of the managed systems.
You can probably also only measure the performance of the whole team, I see no way how to judge individual performance except from peer reviews.
A simplified formula would look like the following:
A more elaborate formula would contain complexity classes (low, medium, high) for every service and application and add appropriate organization-dependent factors for different tasks.
A formula like the above has the advantage that it can also serve well to justify more people when the organizazion grows or faces rapid changes.
Regarding the risk, the monetary value of a risk can be expressed as CostOfDamage * ProbabilityOfOccurence.
You can now ask your bean-counter to come up with a list of downtime, data-loss and security-breach costs for all services, applications, databases and file servers, but he will surely find out very quickly that this does not really matter because it will easily amount to a sum that will put your company out of business.
This means that your organization has to spend any amount necessary to lower these risks to an acceptable minimum, and you can also point out that the management is personally liable that a reasonable data security and operational safety concept is in place.
Slashdot Mirror
...why can't I buy all these wonder batteries?
In the last five years I must have read about at least fifty breakthroughs in battery technology, but nothing of it has reached the consumer (me) yet.
I believe that this is because researchers seem to exaggerate their research results for obvious reasons and seem to underestimate what it takes to make a successful product.
Regarding battery technology I completely stopped to believe anything that comes out of the research community.
Unless I can buy it, it does not exist.
p.
The unit kilowatt is fine, but the number is ridiculously low. Cars breaks are typically designed to have a breaking power of four times the engine power, so we are talking about 500-1000 kW of breaking power in typical cars.
Modern high speed trains have a breaking power in the order of 10 - 20 MW using their engines for regenerative dynamic braking. The german ICE3 has about 16,4 MW dynamic breaking power, which is only slightly higher than the 16 MW propulsion power. It also has additional eddy current brakes, but they there breaking power is just 1600 kW, about 10% of the the dynamic breaking power. There are even additional disc brakes, but they are only used in emergency situatuions.
The largest burner on my gas stove has 5 kW power, but I don't think I can make a pan to glow in seconds, if at all.
Watching the video I would assume that they are talking about 4.6 MW oder 4600 kW of breaking power this disk can handle.
I do not see why the burden of proof that massive dissemination of poison is harmful should be with the public.
IMO those who manufacture and sell this stuff have to prove that it does not destroy our ecosystem.
I know, the stuff has been at some point been certified, but I think that every company that manufactures a product has an obligation to monitor if it is harmful even after it appears on the market. You simply can not determine the long term impact of wide use on the environment with a handful of studies,
p.
More precisely:
If the game state is maintained and processed on "their" computer, it is a service. Otherwise I would regard it as a "good" that is executed on my computer.
p.
The "Cubic Mile of Oil"-numbers do show the opposite of what you are suggesting.
Replacing one CMU with nuclear is the second most expensive alternative, only solar panels are more expensive. Building 2200 nuclear power plant is not only totally out of question, it would also deplete world uranium reserves within a decade.
On the other Hand, building 1.6 Mio. wind turbines is the cheapest viable alternative. Germany alone already has about 22.000 of them, and they are profitable.
Safe nuclear power is not commercially viable. The only reason why it looks cheap is because today's commercial reactors are unsafe by design and the risk costs are carried by the society.
With the cheap nuclear reactors that are still being build today we will have a major nuclear disaster every 20-30 years, and the economic damage to the affected country is huge.
Depending on the outcome, Fukushima probably will cost Japan more than has ever been saved by using nuclear power at all. Expect Japan to pay billions every year just to maintain the Fukushima ruin, and this might go on for hundreds of years.
Just one Fukushima or Chernobyl type disaster in Germany would cost more than the transition to renewable sources.
Of course a society can make the decision to take the gamble, but Germany has been hit hard from Chernobyl Fallout - in some areas in Bavaria deer still can not be consumed because the meat is too radioactive, and it will take about 200 years until the situation will be normal.
The U.S. may not have such big problem when a few thousand square miles get polluted with radioactive fallout - the country is big.
For a small country like Germany, one nuclear disaster might cause more damage than World War II.
My guess is that within the next 30 years we will see another major nuclear disaster somewhere in the world, most probably in the U.S. or in France.
I also expect that most planned nuclear reactors never will get built because even before Fukushima they were too expensive, and after Fukushima no sane Investor will sink money into it.
Like about 80% of the german population I regard the decision to end nuclear power in Germany as a good one - not because I am afraid, but because it is an economically sound decision.
p.
Reading this article, I was thinking this security guy is exaggerating and playing down at the same time.
First of all, in the U.S. many companies use so much crap when it comes to IT that it makes me sick, so everything is possible. However, I think it is much more probable many systems will blow up on a large scale without any malice involved, but just due to incompetence and negligence.
At the same time this guy admits the U.S. is actively preparing and maybe even conducting cyber-warfare against other countries. I don't know how to comment on that. If all countries would stick to cyberwarfare instead of dropping bombs, this would not be *that* bad.
The talk about stolen intellectual property and trade secrets is mostly bullshit. Any business that requires a great deal of secrets to be kept is not sustainable anyway. The future belongs to companies who need very few secrets, if any at all, and are quite open about most aspects of their business. Secrets tend to get out sooner or later anyway.
For mission critical software the quality standards should not be very high, but insanely high. And when the life of people is on the line, software alone should never be able wreak havoc. Unfortunately there are too many people out there who don't have a clue and are just happy when things work. The only get wiser when after the shit hits fan a couple of times, but then they overreact. Professionals should have more courage and never let hazardous systems become operational.
However, I don't see a chance that most of those responsible for the bottom line would voluntarily invest in security and safety unless they are forced to do so, either by law, or by shitstorm.
p.
No.
I read an interview with these guys some times ago.
- They basically sell very small bars, like 1 gram pieces, which are sold which high markup everywhere
- they don't target investors with these machines, but people who want to bring a gift home
- they want to raise interest for buying gold this way
p.
I agree with you that the Browser Issue is a secondary one.
But no one would deny that Microsoft has a de-facto Monopoly on desktop operating systems and office applications.
Bill Gates did not become the richest person in the world because Microsoft did a good job on innovating, but they successfully succeeded the IBM Mainframe monopoly in the PC market on the back of IBM.
The only great thing Microsoft did was to royally kick IBMs testicles when they felt they could do so safely.
Then they operated for decades with an paranoid underdog mentality, and because they knew that their technology and their people were at best second-rate, they thought that every new small competitor might do to them what they had done to IBM, so they used every dirty trick in the book to eliminate upcomers.
And at the same time they enjoyed their monopoly and charged way too much for their software.
Software has a wonderful economy of scale, and with software you can create wonderful lock-in effects, and also abuse the patent and copyright system to keep competitors away.
By the number of windows licenses, a copy of windows should not cost more than a single digit amount of dollars, and Office as well.
Microsoft should have been forced to open its books and all the indecent profits should have been taxed away. Price controls should have been enforced based on the actual costs of developing and marketing the stuff, divided by the number of copies sold.
What happened here is a classical example of how laissez-faire capitalism hurts the economy, stiffles innovation and makes a few obscenely rich.
Now, before calling me a communist:
I regard Communism and Capitalism as equally inhumane because they both deny and suppress half of what makes us survive and prosper. Every sane human being has not only the desire to maximize his own profit, but also to give away, to share and to sacrifice for the prosperity of the society he lives in.
A society or culture that focuses on one side is simply doomed.
p.
Yeah, sure, those terrorists capable of turning an iPhone into a cruise missile will be stopped in their tracks by this measure. The would never dare to smuggle in an iPhone *with* GPS, and would never get the idea to buy a $20 GPS module for this purpose.
No, your scenario is complete nonsense. There are much simpler and cheaper solutions when you want an programmable embedded GPS system.
However, what you can do with an "off the shelve" iPhone is to it use as tactical computer to coordinate the movement of you strike teams in real time. And with the iTerrorize-Plugin from the AppStore every iPhone-owner can sign up for a suicide mission on behalf of any organization. He just needs to bring his own weapons and explosives.
Seriously, a society that fails to provide a meaningful life in dignity for too many talented people will eventually get in serious trouble. You want the talented people to work for the establishment, not against it.
Restricting and regulating the use of some technology does make sense, but isolating your society from progressive and wide-spread technology used by the rest of the world will weaken your economy and make your country an easy prey for your neighbors in the long run. I don't believe that restricting the use of GPS in today's world can have any positive effect.
p.
You are right, this claim might be a bit far fetched, but the "Universal Declaration of Human Rights" gives you the following rights:
Article 1
All human beings are born free and equal in dignity and rights. [...]
Article 19
Everyone has the right to freedom of opinion and expression; [...]
Article 23
3. Everyone who works has the right to just and favourable remuneration [...]
Article 27
2. Everyone has the right to the protection of the moral and material interests resulting from any scientific, literary or artistic production of which he is the author.
Article 29
1. Everyone has duties to the community in which alone the free and full development of his personality is possible.
2. In the exercise of his rights and freedoms, everyone shall be subject only to such limitations as are determined by law solely for the purpose of securing due recognition and respect for the rights and freedoms of others and of meeting the just requirements of morality, public order and the general welfare in a democratic society.
There is of course no right that says "All humans have the right to put Easter Eggs in their products", but taking all these above rights together, when you create something extraordinary, you are entitled to "protection of the moral interests", and matters like human dignity, freedom of expression and development of personality under circumstances may imply this right, unless of course you do no harm and respect the rights of others.
I don't believe you will be very successful in a U.S. court with this argumentation. Arguing with human rights in court generally does not work well in most countries of the world.
But the fact that there are so many easter eggs around and I never heard about anyone getting in trouble because of it indicates that most people, be it customers, superiors or colleagues do appreciate or tolerate it.
p.
but if you are proud of what you have achieved, go ahead, it is a human right. I did it a couple of times when it seemed appropriate, and it was never a problem. If you are in a position where you are responsible for creating the system, it is just your decision.
If you need a rational argument for doing it: It might make the team more proud of the product, more productive for future releases, projects or products. A company is about people, and it is a people issue.
And when even weapon designers can put easter eggs into missiles, I can't imagine anything to be too serious to pull it off; a product just may be too boring or suck too much to deserve an easter egg.
An easter egg is just a sign that the people who made the product did care about it, and are proud of it. An easter egg is like a medal or decoration for the product.
p.
This guy, Lutz Heilman, would be right now the least liked politician in whole Germany, if someone would care. Even his own party does not like him because he did not properly disclose that he worked for the Staatsicherheit, the secret police in Communist Germany. His political career is fucked since 2005, when the German news magazine "Der Spiegel" disclosed it, and now he probably has nothing to loose. The facts in the wikipedia are based on solid sources, so I don't think the court will decide in his favour. And not that it really matters in this case, but he is also avowed gay and a lawyer.
p.
It is all more smoke than fire, only forwarding from www.wikipedia.de to de.wikipedia.org is affected, you can still access wikipedia under de.wikipedia.org from Germany. So actually not much is blocked, and when you go to www.wikipedia.de, you are even given the URL that still works.
And it is only a preliminary injunction, which probably will be lifted soon. It has happened already, but this futile symbolic act will probably just last a few days. In Germany, the courts do generally favor the right of every person to defend himself against defamatory speech, and are often quick with such injunctions to limit the damage the press can do to individuals.
It is a conflict between different human rights, the right to free speech, and the right of any human individual to live in dignity.
Actually the the Article 1 of the German constitution is one of the most beautiful and touching sentences I have ever seen in any piece of legislation:
"Human dignity is inviolable."
Free speech is important, but human dignity comes first.
Although the whole thing is stupid, I can not say the court did act on behalf of the powerful to suppress free speech, on the opposite, it did what it could to protect the rights of an politician who is not in high esteem with any of the established parties. The guy is from a small left wing party, the successor of the communist party of east germany. They get around 5% of the votes in whole Germany, and up to 20% in some parts of the former east.
Hope this makes a few thing more clear.
p.
The more you build on top of something, the more solid the base should be.
If your own stuff is just a few thousand LOC, it does not matter that much if you are locked into a particular framework. You can still do it again with another framwork.
However, if you investment is a major one, you better roll your own or use something rock solid, that even has different implementation with a similiar API or functionality, like a database or script interpreter.
And when in doubt, roll your own or built an abstraction layer.
p.
Listening through the whole podcast after reading many comments here, I found most people here did not get the point.
Their stateless SYN-Handshake just makes the described types of attacks even more resource-efficient for the attacker, but I assume that it is even feasible when using "normal" SYN-ACK connection estabishing.
The point is, that the attack starts *after* a connection is set up, using all types of other flow-control mechanisms in order to exhaust specific resources of the tcp-stack.
As an example I just made up, after the connection is open, you could use a flood of ACKs combined with a particular order of out-of-sequence packets or fragmented packets to make the system use up a lot of timers until the whole tcp-stack or the kernel goes down the drain. But as mentioned in the podcast, Timers are just one of a large number of different resources involved in housekeeping after a connection is established, and you can craft attacks to exhaust a specific set of these.
This opens up a complete new can of worms and may turn out not to be easily thwarted because you have to guard *all* of your resources, not just those involved in establishing a connection like in a SYN-attack.
Noone currently has an idea how serious this will turn out to be in the wild, but to me it sounds serious enough not to be ignored, and it has at least the potential to kill the open internet as we know it. In some distant future, you might need an authorization to even send a packet to a particular host, or the whole internet will be so closely monitored that anyone interfering will be visited by black helicopters immediately.
Btw, there is a cool protocol called FLIP (Fast Local Internet Protocol), designed by Tanenbaum in the early 90s for the Amoeba operating system, that would kill many of the problems we have today on the Internet, and it seems to be immune against many types of attacks, including eavesdropping, but FLIP, as the term "Local" in the name says, works just on one local Ethernet, using TCP-Tunnels to connect to other Ethernets, so it is not a viable replacement in its current form and might not scale and has probably a number of issues on its own. What is however remarkable that in FLIP even the source and destination address and port are encrypted, so you can not get even a single packet through if you don't have the permission, and you don't know who is talking to whom even when you sniff the traffic. Although it won't be able to replace TCP/IP, the concepts are an interesting starting point when thinking about a future secure internet. Just imagine every router on the internet would be a kind of TOR-Node, and you could transparently adress every running process on the internet in a secure fashion, no matter where is physically resides, and you could even multicast and have group communication on the network level, reaching millions of host with just sending one packet, but it will arrive only with their prior permission, which you give by just joining a communication group. Imagine a kind of PGP on the network level. This is what a future internet could possibly provide.
For now, we are stuck with TCP/IP, and maybe for a very long time, unless this kind of problems will finally bring the Internet down, which I think is totally possible because TCP/IP is designed for a network of hosts requiring at least some trust and playing fair on the TCP level.
p.
As far as I remember, we did offline data store checks whenever we encountered problems, which became more frequent as the database grew and aged (about once a month in the beginning, about every week in the end). We also moved the database to at least three different servers because we first assumed the hardware might be the problems.
Anyway, your experience seems to confirm my observations: Comparable Microsoft engineered systems seem to deteriorate faster than *nix-based systems. I find it unacceptable when a system needs so much attention and maintenance just to stay alive, especially when this kind of maintenance involves downtime, and manual maintenance always carries the risk that the operators make errors, and as they are just humans they sooner or later will.
Anyway, there is no excuse for a repair tool to crash, as it should be *designed* exactly for dealing with corrupt data. However, Microsoft has not a monopoly on such a behavior, many years ago I dropped ReiserFS because of the same combination of corruption and crashing repair tools.
I think that the hands-on approach and informal culture that gave Microsoft such an advantage over a scientific and buerocratic style of engineering that was commonplace in the mainframe and workstation world came back to haunt them and their users when it came to serious enterprise computing.
p.
I think such fast and reliable system is always very difficult to build, and those guys who build are top notch engineers, I am sure about that. They guarantee 10ms transaction time and promise to go down to 3ms. Given that this involves network communication with a third party and some arbitration and transaction logging, this is quite cool.
However, no one in the world is able today to create a bug free system, and sometimes shit happens. My guess is that someone tried to save on maintenance costs, and when the system broke down, they obviously did not have the right guy in place who is able to find out whats wrong and fix it. Shit happens, but when it happens, you want a fast and efficicient troubleshooter right on action.
The problem might be windows indeed, but not because it is less realiable. I think with windows there are two problems:
1)It is difficult to understand because the source is not widely available, although the guys involved in this project would have access to it.
2) It accumulates entropy faster than a Unix-based system; this means you have to reboot a windows system more often to remove entropy and by putting the system into a more defined state. Linux and BSD share this problem, and while a typical linux or BSD is only rebooted to install kernel updates, you do more often reboot a windows system because it behaves strange and you have no idea why, especially with desktop systems, but even more with laptops in changing network environment.
On many *nix systems you can find out by reasoning how to fix a problem. In windows, you may find out by reasoning what is wrong, but you have to *know* what to do in this situation, you can often not find it out by reasoning.
This may have contributed to the problem in London, which did cost them two 9s of their five nines they were claiming.
Our company moved away from Microsoft Exchange when we had a two day mail outage because of Exchange database corruption. We had backups and logs and a proven recovery process, but this time it failed. By replaying the logs, corruption occured even on copies of several older backups of the database. We settled for one day of lost E-Mail traffic to bring back the system.
I wanted to stay with Microsoft Exchange because of the hazzle and costs of moving to a different mail and calendar system, but one thing was paramount: I was determinded to make that never happen again, not because of the material damage, which was negligable, but my reputation as a seasoned professional technician who moved into upper management was at stake.
I was very surprised when I found out that no matter how much money I would be willing to spend on hardware, licenses and the best experts, there would be no way to recover from Exchange database corruption once it occurs better than we did. When we did log replays, the Microsoft recovery tool crashed near completion after one hour of replaying logs, leaaving us with an unusable databese that caused the Exchange Server to crash when started. We finally had to go back many weeks in the backup cycle and replay Gigabytes of logs only to loose just one day of mail. Even after this recovery, the server continued crashingon different hardware, so we were sure it was the software and a possibly corrupted database and broken repair tools.
So all the experts I consulted told me: the situation we ran into was not supposed to ever happen, and it would propably never happen again, but they could not rule out it might happen some day again, and when it would happen again, we would be fucked again.
The first lessen for me was: When the software contains bugs that corrupt the database, and the repair tool is broken, and corruption is not detected by the running software or during backup, and only crashes the server when some someone accesses one of the broken items, then you are really, really fucked by Microsoft, without any chance. And I was lucky, my reputation was just slightly damanged, but other people might easily loose their job because of
According to http://arxiv.org/abs/astro-ph/0012399, the average earth in the universe is just 1.8 +/- 0.9 billion years older than our earth, which means that other earth-like planets have just a few billion years more to evolve, and while it took five billion years to create intelligent life on our earth, we can not tell whether this is a long or short time compared to other planets. There may be some special factors like our moon that could have accelerated this process, so it might be possible that planets without massive moons take 10 billion years to develop intelligent life.
I am not saying that this is very probable, but I think it is not "improbable enough to dismiss out of hand". If we were first in our galaxy, it would be indeed surprising, but it would be also a funny explanation for our situation.
You can also not take for granted that even if there is intelligent life, an industrial revolution will take place, and we do not know yet whether it is sustainable long enough to evolve sustainable space colonization.
According to my own calculations, to accelerate a 600.000 ton spaceship (including fuel) to near light speed and decelerate, you need to turn about 400.000 tons of its mass into kinetic energy. A small robotic probe of course might be much cheaper, and travelling in a small craft in a hibernated state also might be the better option, but I would not want to travel in a conscient state for years in a structure that is smaller than a large ocean cruiser.
I agree that the cost is not prohibitive if we find a more efficient way to turn mass into energy than fusion or fission, like antimatter annihilation, but without the need of putting 400.000 tons of antimatter into our gas tank. (I think there would be a law against bringing substantial amounts of antimatter into a solar system.) However, I can envision a process where matter is heated to early big-bang temperatures and turned into a quark-gluon plasma, that will exhibit almost total matter-antimatter annihilation on cool-down. (Baryogenesis and Leptogenesis) We may however have some containment problems with our "reaction chamber" here, requiring some more "breakthrough technology".
Although this might be possible in theory, we do not know if this ever will be possible in practice, here or in any other solar system.
Btw, I do not believe that FTL-travel ever will be possible, all our knowledge speaks against it.
I wish this would be true, but my experience and knowledge of history indicates otherwise. The industrial revolution happened in Europe not because of competitive cooperation, it was driven by four hundred year
Although improbable, there has to be a first civilization in our galaxy, and maybe we are the most advanced life form in our galaxy so far. Regarding the age of the universe, it is possible, because the matter of the solar system had to go through 2 or 3 supernovae before having enough high period elements, and it took the universe about 5 billion years alone to create us since our planet was created, which is more than a third of the age of the universe.
This leaves us with a number of possibilities:
- we are the first
- intelligent life is very rare or very fragile and volatile in our universe
- the guy is right and they are already here and among us (maybe they are criminals hiding on a primitive planet)
- interstellar distances are simply to big to be overcome, or no one wants to pay for the energy
- evil alien predators are already on their way to blow us out of the sky before we become serious competitors
- we live in a zoo and are just being watched
- everything is just a big illusion, and our reality is only virtual
p.
that fortunately does not exist in Germany. Here the law is simple: A company that wants an N.C.A. to be enforceable, it has to pay at least 50% of the former wages of the employee, otherwise the N.C.A. is void. It also has to be very specific, the new company must be competitor, being an IT-company is not enough, you basically have to provide the same product to the same custumers. It is also limited in time to one year.
When I once left a company that didn't want to let me go I happily told them I would love to sign an N.C.A., but when they saw what it would cost them and would bring them (I would be gone anyway), they quickly reconsidered.
p.
It was more like:
US -> England, France, Soviet, and a bit Israel
France -> Israel
Israel -> South Africa
Israel, South Africa -> Khan Network
P.
just great. I'd mod you up if had points.
p.
In a situation when someone acts like an asshole the best thing is to make this person to understand that immediately and without any chance to get it wrong.
Human are in possesion of archaic communication channels that get directly through and can not be ignored.
It is called emotion, and in such a situation it is appropriate to get just frantic. Take a deep breath and start to cry out as loud as you can that you are fed up with such an selfish and antisocial behaviour, and call this guy whatever actually comes into your mind.
Many people, especially geeks have problems to show their emotions and consider it a weekness, but emotions are a very powerful and efficient mean of communication and immediate persuasion.
We are beeig told to be cool and stay calm, and talk things over, but there are situations where you can talk for hours and you just don't get through. You state you irrefutable arguments over and over again, but the other person lives in a different reality and filters or bends off everything you say.
But then, if you show true emotions, you will get through, no matter how hard-boiled or ignorant the other person is. He does not even need to speak your language.
However, it is a good idea to practise a bit, to learn not how to suppress emotion, but how to let out controlled burst of varying intensity. With mastering three or four different levels between "beeing normal" and "totally freaking out" you have enough granularity to master most situations without excessive collateral damage.
You should also try to observe how you react to emotions of others and how hard or impossible it is to ignore them.
p.
However, a beancounter needs some figures in terms of money.
A possible approach would be to estimate the value of the IT-Infrastructure to function properly, and also quantify the risk of security breaches or data loss.
Other factors are the number of servers, clients and users, the number of services, the number of applications, amount of data storage and the amount of internal and external network traffic and the diversity of the managed systems.
You can probably also only measure the performance of the whole team, I see no way how to judge individual performance except from peer reviews.
A simplified formula would look like the following:
sysadmin-productivity = (NumberOfSupportedUsers/NumberOfHelpRequest * numberOfNewUsersPerYear* log(NumberOfServers) * NumberOfNewServersPerYear * 1/serverDowntime * NumberOfDifferentOS * log(NumberOfServices) * NumerOfNewServicesPerYear * NumberOfDifferentServices * log(dataStorage) * log(dataStorageGrowth)* log(networkTraffic) * log(networkTrafficGrowth) * numberOfSupportedApplications) / numberOfSysadmins.
A more elaborate formula would contain complexity classes (low, medium, high) for every service and application and add appropriate organization-dependent factors for different tasks.
A formula like the above has the advantage that it can also serve well to justify more people when the organizazion grows or faces rapid changes.
Regarding the risk, the monetary value of a risk can be expressed as CostOfDamage * ProbabilityOfOccurence.
You can now ask your bean-counter to come up with a list of downtime, data-loss and security-breach costs for all services, applications, databases and file servers, but he will surely find out very quickly that this does not really matter because it will easily amount to a sum that will put your company out of business.
This means that your organization has to spend any amount necessary to lower these risks to an acceptable minimum, and you can also point out that the management is personally liable that a reasonable data security and operational safety concept is in place.
In short terms:
Having a good team of sys-admins is priceless.
p.