...and also the Germans. If I remember my history, Finland was fighting multiple enemies throughout WW2. Sometimes walls work. Even if it just slows the invasion down, that is at least SOMETHING.
If the interview was in 1991 instead of 1999, then that would really be something. Bowie's side would have been much more prophetic and Paxman would have seemed much less of an idiot.
Maybe I misread the summary, but it made it sound like QKD has nothing to do with encryption. It sounds like a way to determine if someone is listening or tapped into the line. Or did I misunderstand this statement: "If any of the pulses' paths are interrupted and they don't arrive at the endpoint at the expected nanosecond, the sender and receiver know their communication has been compromised."
How about encrypting the data so that you don't care how many people capture the 1s and 0s going over the wire (be it electrical or optical) since none of them can make any sense of them without the decryption keys?
Exactly my thoughts. I really like my SSDs for data I use every day, but for the TBs of stuff I want to keep around but only access once in a while (or backups); nothing beats the good old HDD. SSD storage is still about 10x the $ per TB as HDD. In spite of all the predictions that they would cross over by now, they are still a big premium. Sure SDD prices have fallen, but HDD prices have as well so the premium has largely remained unchanged. Even if they somehow managed to get SDD or XPoint technology down to where it was just double the price of HDD, I doubt I would abandon my hard drives entirely. Data always seems to expand to fill the space we have to store it.
...their 'Assistant' can detect that it is a scam artist or robo-caller on the line and can try and waste as much of their time as possible. They simply will not go away until it becomes very expensive to find those 'suckers' that P.T. Barnum told us about. If it costs them 10 minutes of a real person's time (even if it is some poor guy in India) every single time they robo-call someone and they get nothing in return (because you were just pretending to look for your 'lost' credit card), then they will eventually stop trying. Call blockers or Do Not Call lists or laws will do nothing to stop them.
A lot of government sites currently publish data that can be easily downloaded, but the average user has trouble making sense of it. For example, the city of Chicago has a website where you can download crime data https://data.cityofchicago.org... for the last 18 years. You can pick from formats like CSV or XML (but they don't have Json yet) to download. Their website visualization tools are getting better, but I wish they were much more flexible.
I am building an analytic tool that makes it really easy to create relational tables from CSV or Json files and do all kinds of analytics using it. See a quick 4 minute demo video at https://www.youtube.com/watch?... to see a few things we can do with that Chicago data. It is just as easy to do similar things against any data you can download. In this instance, trying to load the 6.5 million row table into Excel is not very practical.
Let's see....they paid $10 million for it two years ago. Since they are the government, they probably paid 3 times what it was worth by going through 'approved vendors' who are used to soaking it to the taxpayer. Computer equipment tends to depreciate at a quick pace, so it could easily lose half its value in a couple of years. Still, it should be worth a couple million dollars to somebody if there was any decent hardware in there that wasn't so specialized that it doesn't have any practical use outside of its original intent (e.g. anything built by NASA).
'Computer equipment' could mean anything. If it were practical things like disk drives, SSD, or tons of memory that could be easily used in other hardware then I think they could have gotten some decent bids on it. The fact that they couldn't attract any bids tells me either it is all junk, or they are not giving out proper information so they can unload it for pennies on the dollar to some crony friend who will make a killing on the deal at taxpayer expense.
You know that system that is behaving badly and just seems to work again the minute you reboot? That might not work if engineers get non-volatile RAM to work really well. Everything will be in memory with no need to load it off disk. But those slow memory leaks, weird data corruption bugs, or software that stops working when it gets into a certain state; will not just magically disappear when you reboot!
Everyone knows that switching DB vendors can be a real pain, even if you are just moving between mainstream, similar systems. A company I once worked for spent a couple years (and lots of money) to move from MS SQL Server to PostgreSQL because of licensing costs. Cost savings are certainly one consideration, but there are others as well so I would like to know what has made you switch.
I am building a whole new kind of data management system that also does some relational DB functions. Getting early adopters can really be a challenge so I want to focus on things that are the most painful for people today. Our system can handle some pretty big tables (tested to 300 Million rows with 50 columns), and most queries have been faster than other systems (benchmarked against SQL Server, MySQL, and Postgres). We don't need to index our tables to get really fast queries either! In short, it is an entirely new kind of system that uses a completely different architecture than other systems. It is a columnar store with a flexible schema so we can add or delete entire columns on big tables without needing to reload data. It can do both transactional operations and analytics very quickly. See a 4 minute demo video at https://www.youtube.com/watch?... if you want to see it in action.
When I said it was 'solved' I didn't mean that somebody magically put a bunch of money into your 401(k) account or that everybody who qualifies for it takes advantage of it. You have to be disciplined to contribute regularly to it over the coarse of your career. If the median balance is only 75k then obviously few people are actually doing that. If you are talking about average balances, that is different than only counting balances of people ready to retire.
Double check what numbers? The numbers I pulled out of my hat to give as an example? Nobody claimed they were actual, valid numbers. The only real number to consider is 'Did overall tax revenues (you know, the amount the government takes in) increase or decrease after the tax rate cuts?' I believe the answer is INCREASE.
Unfortunately, I am working on my own startup that doesn't have the revenues yet where I can pay myself anything. I can't wait until I can say that where the number I am multiplying it by is not zero.
You won't do anything personally to try and fix this situation (learn a new skill, find a different job, make yourself more valuable to your current employer, etc.) but you probably want some politician to 'Fix it', by punishing your employer or making them form a union or something. Big government advocates love to have more people like that.
Yes it really sucks when your retirement and health care is tied to your employer. But it is silly to think that the only solution to that is a nationalized health care system. We basically solved the retirement problem with the 401(k) system. Now you don't have to work for GM for 40 years to get a pension. You can bounce from job to job and have each employer contribute something to your private retirement fund that you control. I have worked for 7 different companies over my career and put some money into a 401(k) from each of them.
We could do the same thing with our healthcare. Each person/family could have a personal health plan where your current employer pays some or all of its costs. You can control what coverage you want (instead of some government bureaucracy) or what deductibles you are comfortable with (gasp, the same as you currently do with life, home, and auto insurance). If you leave your current employer, those contributions may stop but the plan is still in force. You can pay them yourself, or get your next employer to chip in as a condition for getting you to work for them.
You should never measure tax cuts against deficits and debt. That is just a political game that gets played every time and propagandized in the media. They should only be measured against TAX REVENUES. If you cut taxes and overall revenues go up, (because the economy is stimulated and growth drives it) then the tax cut is a benefit to the country as a whole. Just because the politicians decided to spend an extra $500 Billion during the same year that tax revenues only went up by $200 Billion, does not mean the tax cuts causes an extra $300 Billion deficit.
This is precisely the kind of project I am working on. It is a new data management system that performs many operations in parallel. Some databases, for example are multi-threaded but only in the sense that they can run multiple, separate queries in parallel. They can't break up a single query and run parts of it on separate cores for faster processing. If you run a big query on a quad-core CPU it will take the same amount of time as running it on a 10-core CPU (assuming similar clock speeds).
My system on the other hand can run parts of the query in separate threads so I get about a 50% speed improvement when processing a query against a relational table on a hex-core vs a quad-core CPU. It does the same thing when dealing with things like lists, bitmaps, or key-value stores. All is implemented in the library, so applications get the speed improvement without needing to create the threads themselves.
There is a sucker born every minute. Unfortunately, the rest of us make it too easy for these scammers to find them. As long as there are a healthy number of suckers out there (just.01% = 700,000 people), these scammers will never quit. Laws won't stop them. Blockers won't stop them. Do Not Call lists will not stop them. As long as they get more money out of the scam than they have to invest in it, they will do it (kind of like Bitcoin miners). The rest of us have to make it extremely expensive for the scammers to find those suckers who will fall for it. Once it costs the scammers even $1.01 for every $1 they get out of it, they will go the way of the Dodo bird and the rest of us won't be bugged anymore. So waste their time! Get a real person on the line and keep them talking. Robocalls are almost free. Paying these foreign call centers may be cheap, but they are far from free.
The problem is almost nobody is building high quality PC software these days. Most new development is out in the 'cloud' with the PC being treated like the old 'dumb terminals' from the Main Frame era. The stuff running on the PC now is mostly front end client software that just sends commands to the servers in the cloud.
Once upon a time, all the cool new software was running on a PC that was not connected at all to the internet. Local databases, spreadsheets, word processors, etc. were all designed to work on a standalone PC with you controlling all your data. I think the pendulum can swing back and bring demand for more HEDT machines if compelling PC software once again is created. The backlash against big cloud providers that either abuse your data for their own profit, or fail to protect it from hackers will be a major factor in the pendulum swing if it happens.
This is precisely why I do buy IoT devices and do not subscribe to 'services' that send my personal information to somebody's 'cloud'. They will sell you out the first chance they get to make a buck at your expense. If you can control it with your phone, then you don't own the data it creates. It is sitting on somebody else's server and they can do whatever they want with it. Even if they promise (in writing no less) that they will never share it, they will. This goes for your video doorbell, your alarm system, your smart lighting system, your sprinkling system, etc., etc..
I fully realize that is not according to FOSS requirements, but that does not stop companies from wanting now to put restrictions on it. They will simply have a hard time doing that after the genie was let out of the bottle. Once you release your source code as open source, you have given up your rights to all kinds of things. Now these same companies are having 'donor's remorse' when they find out they did it with one business model in mind, but now the rules have changed and they want to turn back the clock.
I face this dilemma myself. I have been working on a new data management system (kind of a database/file system hybrid that is way better than WinFS was supposed to be) and have been advised by many to release it as open source. That would undoubtedly increase its exposure and might promote wide adoption, but once you do that there is no going back. If I actually want to be paid for all that hard work I have done, then I may never be able to do that, so I am very hesitant. Until I see a clear path to actual money, I am not going to chase after some mirage.
What these companies are realizing is that it doesn't matter at all if you have a big slice or a small slice if the pie itself is worth ZERO. Anything multiplied by zero is zero. For years, these companies were willing to donate lots of code to the 'community' so that the pie would get nice and large. Then they would add all kinds of value to an 'enterprise' version and convince a small piece of the pie (e.g. 10%) to buy their very expensive solution. They gave away the milk as long as they could scrape the cream off the top and sell it.
Now they are discovering that with cloud services, other companies can come in and scrape away all that cream and leave them with nothing. Those other companies have contributed little or nothing to the actual development of the code so they have no costs to recoup. The open source companies are realizing that the open source model contributes to this whole freeloading situation and want to put a stop to it. I like free software as much as the next guy, but somebody has to pay the bills.
...for certain tasks. If you need to rapidly prototype something; if you are dealing with a small amount of data; or if you need to run something just once a week or month then scripting is a quick way to accomplish this. But if you are doing anything many times each hour, dealing with huge data sets, or if otherwise performance is really important you don't use scripting! There is a reason why things like file system drivers, network controllers, or enterprise databases are not written in a scripting language.
I tend to be skeptical about things like QC because we have heard so many hyped up stories over the years. Cold fusion, fusion generators, room-temperature super-conduction, 100 mpg engines, etc. are just a few examples. That does even begin to address things like WinFS, the decentralized web, mainstream crypto-currencies, or 100 TB hard drives that were supposed to be here by now. But plenty of breakthrough technologies have come about in spite of all the skepticism around them (I am working on one of my own), so I try not to be too skeptical. Let's give it a few years and reserve judgement until we see tangible evidence. Just don't throw a lot of money at it (bitcoin, cough, cough) until you see some.
Slashdot Mirror
...and also the Germans. If I remember my history, Finland was fighting multiple enemies throughout WW2. Sometimes walls work. Even if it just slows the invasion down, that is at least SOMETHING.
If the interview was in 1991 instead of 1999, then that would really be something. Bowie's side would have been much more prophetic and Paxman would have seemed much less of an idiot.
Maybe I misread the summary, but it made it sound like QKD has nothing to do with encryption. It sounds like a way to determine if someone is listening or tapped into the line. Or did I misunderstand this statement: "If any of the pulses' paths are interrupted and they don't arrive at the endpoint at the expected nanosecond, the sender and receiver know their communication has been compromised."
How about encrypting the data so that you don't care how many people capture the 1s and 0s going over the wire (be it electrical or optical) since none of them can make any sense of them without the decryption keys?
Exactly my thoughts. I really like my SSDs for data I use every day, but for the TBs of stuff I want to keep around but only access once in a while (or backups); nothing beats the good old HDD. SSD storage is still about 10x the $ per TB as HDD. In spite of all the predictions that they would cross over by now, they are still a big premium. Sure SDD prices have fallen, but HDD prices have as well so the premium has largely remained unchanged. Even if they somehow managed to get SDD or XPoint technology down to where it was just double the price of HDD, I doubt I would abandon my hard drives entirely. Data always seems to expand to fill the space we have to store it.
...their 'Assistant' can detect that it is a scam artist or robo-caller on the line and can try and waste as much of their time as possible. They simply will not go away until it becomes very expensive to find those 'suckers' that P.T. Barnum told us about. If it costs them 10 minutes of a real person's time (even if it is some poor guy in India) every single time they robo-call someone and they get nothing in return (because you were just pretending to look for your 'lost' credit card), then they will eventually stop trying. Call blockers or Do Not Call lists or laws will do nothing to stop them.
A lot of government sites currently publish data that can be easily downloaded, but the average user has trouble making sense of it. For example, the city of Chicago has a website where you can download crime data https://data.cityofchicago.org... for the last 18 years. You can pick from formats like CSV or XML (but they don't have Json yet) to download. Their website visualization tools are getting better, but I wish they were much more flexible.
I am building an analytic tool that makes it really easy to create relational tables from CSV or Json files and do all kinds of analytics using it. See a quick 4 minute demo video at https://www.youtube.com/watch?... to see a few things we can do with that Chicago data. It is just as easy to do similar things against any data you can download. In this instance, trying to load the 6.5 million row table into Excel is not very practical.
Let's see....they paid $10 million for it two years ago. Since they are the government, they probably paid 3 times what it was worth by going through 'approved vendors' who are used to soaking it to the taxpayer. Computer equipment tends to depreciate at a quick pace, so it could easily lose half its value in a couple of years. Still, it should be worth a couple million dollars to somebody if there was any decent hardware in there that wasn't so specialized that it doesn't have any practical use outside of its original intent (e.g. anything built by NASA).
'Computer equipment' could mean anything. If it were practical things like disk drives, SSD, or tons of memory that could be easily used in other hardware then I think they could have gotten some decent bids on it. The fact that they couldn't attract any bids tells me either it is all junk, or they are not giving out proper information so they can unload it for pennies on the dollar to some crony friend who will make a killing on the deal at taxpayer expense.
You know that system that is behaving badly and just seems to work again the minute you reboot? That might not work if engineers get non-volatile RAM to work really well. Everything will be in memory with no need to load it off disk. But those slow memory leaks, weird data corruption bugs, or software that stops working when it gets into a certain state; will not just magically disappear when you reboot!
Everyone knows that switching DB vendors can be a real pain, even if you are just moving between mainstream, similar systems. A company I once worked for spent a couple years (and lots of money) to move from MS SQL Server to PostgreSQL because of licensing costs. Cost savings are certainly one consideration, but there are others as well so I would like to know what has made you switch.
I am building a whole new kind of data management system that also does some relational DB functions. Getting early adopters can really be a challenge so I want to focus on things that are the most painful for people today. Our system can handle some pretty big tables (tested to 300 Million rows with 50 columns), and most queries have been faster than other systems (benchmarked against SQL Server, MySQL, and Postgres). We don't need to index our tables to get really fast queries either! In short, it is an entirely new kind of system that uses a completely different architecture than other systems. It is a columnar store with a flexible schema so we can add or delete entire columns on big tables without needing to reload data. It can do both transactional operations and analytics very quickly. See a 4 minute demo video at https://www.youtube.com/watch?... if you want to see it in action.
When I said it was 'solved' I didn't mean that somebody magically put a bunch of money into your 401(k) account or that everybody who qualifies for it takes advantage of it. You have to be disciplined to contribute regularly to it over the coarse of your career. If the median balance is only 75k then obviously few people are actually doing that. If you are talking about average balances, that is different than only counting balances of people ready to retire.
Double check what numbers? The numbers I pulled out of my hat to give as an example? Nobody claimed they were actual, valid numbers. The only real number to consider is 'Did overall tax revenues (you know, the amount the government takes in) increase or decrease after the tax rate cuts?' I believe the answer is INCREASE.
Unfortunately, I am working on my own startup that doesn't have the revenues yet where I can pay myself anything. I can't wait until I can say that where the number I am multiplying it by is not zero.
You won't do anything personally to try and fix this situation (learn a new skill, find a different job, make yourself more valuable to your current employer, etc.) but you probably want some politician to 'Fix it', by punishing your employer or making them form a union or something. Big government advocates love to have more people like that.
Yes it really sucks when your retirement and health care is tied to your employer. But it is silly to think that the only solution to that is a nationalized health care system. We basically solved the retirement problem with the 401(k) system. Now you don't have to work for GM for 40 years to get a pension. You can bounce from job to job and have each employer contribute something to your private retirement fund that you control. I have worked for 7 different companies over my career and put some money into a 401(k) from each of them.
We could do the same thing with our healthcare. Each person/family could have a personal health plan where your current employer pays some or all of its costs. You can control what coverage you want (instead of some government bureaucracy) or what deductibles you are comfortable with (gasp, the same as you currently do with life, home, and auto insurance). If you leave your current employer, those contributions may stop but the plan is still in force. You can pay them yourself, or get your next employer to chip in as a condition for getting you to work for them.
You should never measure tax cuts against deficits and debt. That is just a political game that gets played every time and propagandized in the media. They should only be measured against TAX REVENUES. If you cut taxes and overall revenues go up, (because the economy is stimulated and growth drives it) then the tax cut is a benefit to the country as a whole. Just because the politicians decided to spend an extra $500 Billion during the same year that tax revenues only went up by $200 Billion, does not mean the tax cuts causes an extra $300 Billion deficit.
This is precisely the kind of project I am working on. It is a new data management system that performs many operations in parallel. Some databases, for example are multi-threaded but only in the sense that they can run multiple, separate queries in parallel. They can't break up a single query and run parts of it on separate cores for faster processing. If you run a big query on a quad-core CPU it will take the same amount of time as running it on a 10-core CPU (assuming similar clock speeds).
My system on the other hand can run parts of the query in separate threads so I get about a 50% speed improvement when processing a query against a relational table on a hex-core vs a quad-core CPU. It does the same thing when dealing with things like lists, bitmaps, or key-value stores. All is implemented in the library, so applications get the speed improvement without needing to create the threads themselves.
There is a sucker born every minute. Unfortunately, the rest of us make it too easy for these scammers to find them. As long as there are a healthy number of suckers out there (just .01% = 700,000 people), these scammers will never quit. Laws won't stop them. Blockers won't stop them. Do Not Call lists will not stop them. As long as they get more money out of the scam than they have to invest in it, they will do it (kind of like Bitcoin miners). The rest of us have to make it extremely expensive for the scammers to find those suckers who will fall for it. Once it costs the scammers even $1.01 for every $1 they get out of it, they will go the way of the Dodo bird and the rest of us won't be bugged anymore. So waste their time! Get a real person on the line and keep them talking. Robocalls are almost free. Paying these foreign call centers may be cheap, but they are far from free.
The problem is almost nobody is building high quality PC software these days. Most new development is out in the 'cloud' with the PC being treated like the old 'dumb terminals' from the Main Frame era. The stuff running on the PC now is mostly front end client software that just sends commands to the servers in the cloud.
Once upon a time, all the cool new software was running on a PC that was not connected at all to the internet. Local databases, spreadsheets, word processors, etc. were all designed to work on a standalone PC with you controlling all your data. I think the pendulum can swing back and bring demand for more HEDT machines if compelling PC software once again is created. The backlash against big cloud providers that either abuse your data for their own profit, or fail to protect it from hackers will be a major factor in the pendulum swing if it happens.
This is precisely why I do buy IoT devices and do not subscribe to 'services' that send my personal information to somebody's 'cloud'. They will sell you out the first chance they get to make a buck at your expense. If you can control it with your phone, then you don't own the data it creates. It is sitting on somebody else's server and they can do whatever they want with it. Even if they promise (in writing no less) that they will never share it, they will. This goes for your video doorbell, your alarm system, your smart lighting system, your sprinkling system, etc., etc..
I fully realize that is not according to FOSS requirements, but that does not stop companies from wanting now to put restrictions on it. They will simply have a hard time doing that after the genie was let out of the bottle. Once you release your source code as open source, you have given up your rights to all kinds of things. Now these same companies are having 'donor's remorse' when they find out they did it with one business model in mind, but now the rules have changed and they want to turn back the clock.
I face this dilemma myself. I have been working on a new data management system (kind of a database/file system hybrid that is way better than WinFS was supposed to be) and have been advised by many to release it as open source. That would undoubtedly increase its exposure and might promote wide adoption, but once you do that there is no going back. If I actually want to be paid for all that hard work I have done, then I may never be able to do that, so I am very hesitant. Until I see a clear path to actual money, I am not going to chase after some mirage.
What these companies are realizing is that it doesn't matter at all if you have a big slice or a small slice if the pie itself is worth ZERO. Anything multiplied by zero is zero. For years, these companies were willing to donate lots of code to the 'community' so that the pie would get nice and large. Then they would add all kinds of value to an 'enterprise' version and convince a small piece of the pie (e.g. 10%) to buy their very expensive solution. They gave away the milk as long as they could scrape the cream off the top and sell it.
Now they are discovering that with cloud services, other companies can come in and scrape away all that cream and leave them with nothing. Those other companies have contributed little or nothing to the actual development of the code so they have no costs to recoup. The open source companies are realizing that the open source model contributes to this whole freeloading situation and want to put a stop to it. I like free software as much as the next guy, but somebody has to pay the bills.
...for certain tasks. If you need to rapidly prototype something; if you are dealing with a small amount of data; or if you need to run something just once a week or month then scripting is a quick way to accomplish this. But if you are doing anything many times each hour, dealing with huge data sets, or if otherwise performance is really important you don't use scripting! There is a reason why things like file system drivers, network controllers, or enterprise databases are not written in a scripting language.
I tend to be skeptical about things like QC because we have heard so many hyped up stories over the years. Cold fusion, fusion generators, room-temperature super-conduction, 100 mpg engines, etc. are just a few examples. That does even begin to address things like WinFS, the decentralized web, mainstream crypto-currencies, or 100 TB hard drives that were supposed to be here by now. But plenty of breakthrough technologies have come about in spite of all the skepticism around them (I am working on one of my own), so I try not to be too skeptical. Let's give it a few years and reserve judgement until we see tangible evidence. Just don't throw a lot of money at it (bitcoin, cough, cough) until you see some.