First - I like the innovative approach MySQL is taking with dual-licensing/etc. I think that sustainable approaches to Open Source are something that are underexplored.
The only downside to this is what I might call the Norton-factor from the days of DOS/Win95/etc:
Microsoft releases product lacking key critical features. Microsoft works with 3rd-party partners to overcome these lacks of features (undelete, defrag, disk repair, registry cleanup, etc). 3rd party becomes entrenched and Microsoft doesn't want to upset them, so they license stripped-down versions of the 3rd party code and never change their software design to avoid needing the 3rd party tools in the first place (journaled filesystems, robust undelete, no rats-nest registries, self-defragging filesystems, etc). Microsoft has tended to see the error of its ways and WinXP is generally free of stuff like Stacker/ RamDoubler/ PC-Tools/ Norton/ etc.
When a new feature truly is a significant value-add I think that treating it as a separate closed-source product is appropriate. When a new feature is really just making something work correctly that ought to just work, I think we're starting to drift into some mistakes that others have made before.
In the post you just made the one that stands out to me is the Monitor tool. I've sat and looked at PHPMyadmin at all the red fields that tell me that something non-ideal is going on. I'm told that I have queries that aren't using indexed joins. Unfortunately, I have thousands of queries running every day from about 5 different programs, and figuring out which ones are the culprits aren't easy. I can look at the slow-query log and find out what some real monsters are, but often those can't be improved significantly (tons of functions in the SQL - short of a significant app design change not much would happen and this isn't my app), and they don't run often anyway. I don't mind taking the time to run EXPLAIN on my queries, but sometimes it is hard to tell if there is some frequent query running that could be optimized. I think that MySQL monitoring is perhaps a little deficient, and the existence of a separate monitoring tool obviously doesn't help there. Perhaps there should be some improvements to the community monitoring tool, or maybe just better documentation around how to go about doing monitoring manually. I'm fine with the paying customers getting first attention - just try to avoid repeating MS's past mistakes (from which I think even they have learned).
I agree with the issue you raise. The big difficulty is to how to best achieve the following goals for the benefit of all:
1. Encourage development of technology. 2. Make the results of this research available to everybody for the cheapest price possible.
Obviously the simplest solution for this is government R&D. This achieves both goals, but neither is optimal. Some avenues of research might not get studied due to politics/etc, and the cost could end up being high (just in the form of taxes and not license fees).
Patents encourage private development, and generally tend to result in highly efficient applied R&D. The problem is that they don't work if people can just copy the results without paying for them. Even a model where everybody pays once doesn't always work, as the up-front cost could be prohibitive - especially to smaller farmers. If only mega-conglomerates could afford efficient crops then smaller farmers would just go under and you'd have the Walmart effect.
I agree that some of these activities should be reigned in. On the other hand, you can't just allow people to freely copy patented seeds otherwise there is no incentive to make them in the first place. I'd be curious as to what Monsanto detractors would recommend as an actual compromise solution.
Clearly farmers should not be penalized for undesired contamination of their crops. On the other hand, farmers should not get a free ride to be able to essentially deliberately cultivate GM crops and claim ignorance of what is going on. How can your prove intent?
Just deploy the thing armed with rubber bullets or paintballs or something (or just blanks). Use it for a year or two in actual operations, but without lethal armment, just to see what it would do.
Sure, in the meantime you're not getting any benefits of the unit and you'd need to make sure you had enough real troops to do the job, but at least you get a good feel for what the machine is capable of.
Granted, there will be those who raise the issue of how much testing is enough. I think that you need to look at this versus a human soldier. Human soldiers shoot the wrong people sometimes - so the question isn't whether the robot kills the wrong people, but whether it does it less often than humans. In my book that would be an improvement, but of course the way courts run the company that makes the thing is in for real trouble the first time it kills a friendly.
I suspect that this is the same reason we won't see cars driving themselves anytime soon. Nobody minds thousands of people being killed a year by human drivers. However, if you automated every car and only three people were killed every year everybody would call the machines death traps and sue the manufacturers into poverty. There is just no sense of perspective, and the only thing that matters is that the 1000 human killers didn't have deep pockets...
I'd be shocked if the insurance terms didn't cover stuff like this. Granted, I don't know what satellite policies typically look like, but for most general policies there is a duty of the insured to work to protect against future loss.
So, if a tree falls and smashes my roof, that might be covered. However, if I don't do anything about it for a month and it rains 3 times and destroys half the contents of my home, that would probably not be covered. If I did a makeshift repair and a really big windstorm tore it apart and the rain still got in, then I'd probably be covered, since I did what I could under the circumstances.
You don't need to run into a tornado to patch your roof or anything, but you can't just be negligent in doing what any homeowner would do if they were paying out of their own pocket.
If the policy doesn't discourage satellite owners from just sitting back and watching their satellites plummet, then I'm sure future policies will be adjusted accordingly...
Here is the easiest explanation to follow that I've heard. Extend the game to 1000 doors.
One door has a car, 990 have nothing.
You pick a door. Monty opens 998 of the other doors showing nothing. Which door would you pick? He essentially gave away the location of the car - you only had a 0.1% chance of winning, but he eliminated 998 incorrect choices. The chance of the car being behind the last remaining door is 99.9%. This way it actually is somewhat intuitive.
The 3-door game is just the same game but reduced in size.
Honestly, I'm not sure that any transition would need this level of central planning.
I'd have some blue-sky research funding (the kind that never gets funded industrially), and some general tax incentives. And I'd use tariffs effectively:
1. Ramping up over three years a tariff on imported oil would be imposed to cover 100% of all external costs associated with oil imports (such as maintaining an army in the Middle East and safe sea lanes to and from). 2. Ramping up over 15 years a tariff on undesirable non-renewable energy of all kinds. This will be to ensure the strategic strength of the nation and to cover hard-to-measure externalities like global warming. I used the term undesirable since even nuclear power is non-renewable but it is preferable as an interim source to most things we are burning today. 3. Along with both of those tariffs there would be import tariffs on any goods from nations that don't take similar measures - this is to avoid punishing industry that doesn't relocate to places where they can freely pollute. Treaties would be sought with other industrialized nations to take similar steps, and any nation that does not participate equally would be sanctioned in trade. 4. I would publish well in advance the schedule for all the tariffs, so that industry can plan appropriately. If somebody is building a factory that will last 30 years, they'll be proactive about choosing how it will be powered.
Maybe some incentives for early adopters would be pursued as well. A market based approach should be very effective if given time to work, and as long as it is followed rigorously. Disaster would ensue if major industry decided to gamble on the government losing its nerve. I'd advocate increasing the tariffs every quarter just so that the momentum is continuous and steady. If you keep everything low and then have a huge hike then there will be a lot of pressure to keep delaying the hike.
Well, you don't have to have all that money can buy to benefit from the US health care system. Anybody who is upper-middle class with half-decent insurance benefits quite a bit from the standard of care.
I agree that for those who are poor the care available is far lower than in socialized nations.
The issue isn't really quality of care so much as the philosophy behind how it is apportioned. In Europe it is based on "fairness" - mostly FIFO, and in the US it is based on how much it is worth to you (more cash -> more quality). However, I'd question whether a typical MP in Europe waits as long for a medical procedure as the average working-class citizen. In most socialist nations they're really quite capitalist, but instead of paying with money you pay with political power/access/etc.
Personally I think that better care for the poor is important, but if you don't provide any benefits to actually working, then nobody will bother to work. I'm not sure the US has the right balance, but I know that Europe has issues of its own. And in the US when the evil insurance company puts the brakes on care at least there is a government to appeal to - with socialized care the government is your insurance and if you don't like it you can go live elsewhere...
How about this - send each mailing address a form with 1 question - how many people live at this address. It goes on a postcard-sized business reply card that you just drop in a mailbox.
Why do we need a huge questionaire to determine congressional representation?
Or better still, why not just ask a credit reporting agency how many people there are at what address - they probably know what house I'm going to buy next better than I do!
Well, clearly you don't have Intuit/etc charging $15 to upload tax data. I paper file every year for this reason. A fancy computer program generates the return and then I mail it, because I refuse to pay money for something that is otherwise free...
Most US states on the other hand allow you to file online. I'm not asking for fancy tax software - just a simple web-based form where I can type in the numbers and hit submit. Congress keeps saying stuff like "the US ought not to be in the tax preparation software business." I say two things:
1. Fine, don't write fancy software, just create simple web forms that let you type in the numbers that you'd otherwise hand-write onto a paper form.
2. If Congress didn't make the tax code so darn complicated maybe we wouldn't need software to do tax preparation. Most state/local tax forms take all of 2 minutes to fill out.
Ah, yes, instead of paying police to catch people who are an actual danger to society, we can all take weeks off of work to fund our own effort to police the police.
Maybe elected officials should be taken to task to uphold the law in the first place...
Sure, but the dispute is that on the lights with shortened yellows it is simply not possible to safely stop in time. If you're in the wrong spot when that light turns yellow you're either jamming on the brakes or getting a ticket. That's about all there is to it...
Just google for yellow light timings and traffic cameras...
Alternatively, the laws could be set at a reasonable level. When 90% of the population violates a law, perhaps the law is inappropriate?
That is actually one thing I'd like about speed cameras - it would force the laws to adjust at least a little to reality. When the average citizen gets 3 speeding tickets a month there will be a riot...
Is there any reason that limits couldn't be set just a little higher? The argument that is always made is safety. So, why not just ban cars altogether, or make the limits 15mph - a lot fewer people would be killed by cars that way, after all! The issue is that productivity does matter too - and most people seem to be voting with their actions that the current balance is off-kilter.
The military would most likely need to reverse engineer Storm to take it down, so they would be brushing up on some much-needed reverse engineering skills, as well as creating new tools to help automate the process of reverse engineering programs and looking for exploits. THAT is what they would need to keep secret, and it wouldn't be exposed by releasing a program they wrote to take advantage of a specific exploit they found.
But, they could brush up on the same skills by developing something capable of taking down Storm, and then holding it in reserve for some time when they would need to take down Storm, and not before. Then, when some attacker makes a huge strategy around using Storm to take out us infrastructure the US can just take it down at that time and the attacker might not have a fallback plan. If Storm were gone something else would rise up to take its place - perhaps something harder to take down.
In warfare you want to keep your enemy complacent, and the best way to do that is to not attack them while amassing your armies...
Only issue with brushing up is that they would also be advertising their tactics.
Back during the cold war it wasn't uncommon for US forces to intentionally run their radars in a mode good enough for general navigation, but limiting their true capability. That way, in the event of war the enemy wouldn't have all kinds of intel on the frequencies used and general technology.
The same would apply in warfare. If the US started taking out botnets all kinds of folks would be capturing packet dumps and discussing new zero-day exploits and methodologies. Then everybody would engineer defenses against them. Generally speaking, you don't want to use weapons that you don't want the enemy to know about.
The issue is that having a ton of money in the bank lets you "take risk" while in reality being quite safe.
For example, I work a standard salaried job. I could probably make quite a bit more if I became self-employed - I'm quite competent and I think I could market my services. However, there would be a substantial risk of periods of non-employment - especially while getting started off. As the sole significant income for my family I can't just neglect my responsibilities and shoot for the moon, so I accept a comfortable salary that isn't likely to have huge swings in it.
On the other hand, if I had $50M in the bank aside from the obvious investment income my own personal income would probably be much higher, as I could afford to take bigger risks since a year without work wouldn't be a big deal. I'd probably be a lot happier for it as well.
Sure, that makes me less of a risk-taker than some billionare I guess, but only on an absolute scale of measurement. On a relative scale I take bigger risks than Bill Gates any time I take my family on vacation. In the same way I'm not super-eager to shower some billionare with accolades every time they donate $10k to some charity - sure, for the average person that is a lot of money, and on some level it is to be commended even for the rich, but the fact is that Gates/etc could easily lose $10k in change in his sofa and not notice it except for the huge bulge in the cushions...
I don't really buy into the whole tax-shelter-due-to-risk argument. I agree that higher risk needs to have higher reward to encourage investment. However, I think that the current ratio is a bit skewed.
Oh, I know it happens. But something more standardized would be useful.
I'd say the capability is there when you can shut down your webserver, boot it back up, and then touch absolutely NOTHING and have users able to download anything on the site. Oh, and all this on a server that doesn't have X11 installed.
GUI clients really shouldn't be part of server infrastructure. The whole thing should run from init.d and config files.
A nice PHP frontend like you say would be nice, but maybe not completely essential. What is essential is automated operation.
Yeah - it would be nice if somebody created a server-oriented bittorrent distribution system.
The website would post a torrent, and would also seed the torrent. If nobody else seeded it then the website would end up uploading the file to anybody who asked for it - which is no worse than what they'd otherwise end up doing. However, as soon as more than one person starts downloading at the same time you get automatic load-distribution, and if anybody sticks around and seeds then you get even more bandwidth.
All you need is a decent daemon that will take a file and create a torrent and track it and seed it, and restore all this stuff anytime the daemon is restarted. I can't find anything that does exactly that...
Uh, he was talking about USB. The post he replied to suggested that scsi wasn't any worse than USB in numbers of connectors.
USB has 4. SCSI has some number that might be countable using the set of integers - I'm not sure if that has been proven yet.
Even so, USB probably should only have 1 connector - or maybe 1 big and one small. Yes, I know the others have their reasons for existing, but I'm sure there is another way to do things...
That and the fact that software inevitably uses the mysql runtime libraries and you'd need to patch half an application to make it work in postgresql as a result.
For all that MS does wrong, one thing they did right is getting everybody to use ODBC for everything (well, almost everything). If you stick to ODBC and ANSI SQL your app will work with just about any backend. Sure, similar solutions might exist for unix, but nobody uses them.
If mythtv officially supports anything other than mysql then maybe I can consider switching. However, I really don't want to run two databases for everything!
The issue wasn't so much partnering with a for-profit, but not being smart about how they did it.
If you just take a for-profit's product and retool it, then don't be upset if they discontinue the model and leave you stuck - they aren't under any obligation to sell it forever.
On the other hand, writing up a contract between an open-source and for-profit organization where each benefits and has defined responsibilities should be safe. Perhaps the open-source develops the design/code, and the for-profit can make a killing selling it to companies with deep pockets but has to allow cheap/free use for certain purposes. Or you could just develop full blueprints and make them open-source, and pay per unit for actual production.
For-profit companies are often more efficient and can achieve higher economies of scale. On the other hand, they don't share your mission, so you're an idiot if you just lock yourself in without any legal protection.
In this particular case things are a bit more sleezy since the company didn't just stop supplying new equipment, but they went out of their way to break already-sold stuff as well.
Regardless - even if the console did display the error I wouldn't be able to report it short of taking a photo of the monitor...
Don't get me wrong - I'd be happy to go open-source with video drivers. My understanding is that the competitors haven't really been comparable in performance - although perhaps that has changed. When my current $15 video card dies maybe I'll try a $15 AMD card instead...
And the box actually is a firewall already. It is by general-purpose backend server (and linux desktop)... I don't actually watch TV on it.:)
Slashdot Mirror
First - I like the innovative approach MySQL is taking with dual-licensing/etc. I think that sustainable approaches to Open Source are something that are underexplored.
The only downside to this is what I might call the Norton-factor from the days of DOS/Win95/etc:
Microsoft releases product lacking key critical features. Microsoft works with 3rd-party partners to overcome these lacks of features (undelete, defrag, disk repair, registry cleanup, etc). 3rd party becomes entrenched and Microsoft doesn't want to upset them, so they license stripped-down versions of the 3rd party code and never change their software design to avoid needing the 3rd party tools in the first place (journaled filesystems, robust undelete, no rats-nest registries, self-defragging filesystems, etc). Microsoft has tended to see the error of its ways and WinXP is generally free of stuff like Stacker/ RamDoubler/ PC-Tools/ Norton/ etc.
When a new feature truly is a significant value-add I think that treating it as a separate closed-source product is appropriate. When a new feature is really just making something work correctly that ought to just work, I think we're starting to drift into some mistakes that others have made before.
In the post you just made the one that stands out to me is the Monitor tool. I've sat and looked at PHPMyadmin at all the red fields that tell me that something non-ideal is going on. I'm told that I have queries that aren't using indexed joins. Unfortunately, I have thousands of queries running every day from about 5 different programs, and figuring out which ones are the culprits aren't easy. I can look at the slow-query log and find out what some real monsters are, but often those can't be improved significantly (tons of functions in the SQL - short of a significant app design change not much would happen and this isn't my app), and they don't run often anyway. I don't mind taking the time to run EXPLAIN on my queries, but sometimes it is hard to tell if there is some frequent query running that could be optimized. I think that MySQL monitoring is perhaps a little deficient, and the existence of a separate monitoring tool obviously doesn't help there. Perhaps there should be some improvements to the community monitoring tool, or maybe just better documentation around how to go about doing monitoring manually. I'm fine with the paying customers getting first attention - just try to avoid repeating MS's past mistakes (from which I think even they have learned).
I agree with the issue you raise. The big difficulty is to how to best achieve the following goals for the benefit of all:
1. Encourage development of technology.
2. Make the results of this research available to everybody for the cheapest price possible.
Obviously the simplest solution for this is government R&D. This achieves both goals, but neither is optimal. Some avenues of research might not get studied due to politics/etc, and the cost could end up being high (just in the form of taxes and not license fees).
Patents encourage private development, and generally tend to result in highly efficient applied R&D. The problem is that they don't work if people can just copy the results without paying for them. Even a model where everybody pays once doesn't always work, as the up-front cost could be prohibitive - especially to smaller farmers. If only mega-conglomerates could afford efficient crops then smaller farmers would just go under and you'd have the Walmart effect.
I agree that some of these activities should be reigned in. On the other hand, you can't just allow people to freely copy patented seeds otherwise there is no incentive to make them in the first place. I'd be curious as to what Monsanto detractors would recommend as an actual compromise solution.
Clearly farmers should not be penalized for undesired contamination of their crops. On the other hand, farmers should not get a free ride to be able to essentially deliberately cultivate GM crops and claim ignorance of what is going on. How can your prove intent?
Just deploy the thing armed with rubber bullets or paintballs or something (or just blanks). Use it for a year or two in actual operations, but without lethal armment, just to see what it would do.
Sure, in the meantime you're not getting any benefits of the unit and you'd need to make sure you had enough real troops to do the job, but at least you get a good feel for what the machine is capable of.
Granted, there will be those who raise the issue of how much testing is enough. I think that you need to look at this versus a human soldier. Human soldiers shoot the wrong people sometimes - so the question isn't whether the robot kills the wrong people, but whether it does it less often than humans. In my book that would be an improvement, but of course the way courts run the company that makes the thing is in for real trouble the first time it kills a friendly.
I suspect that this is the same reason we won't see cars driving themselves anytime soon. Nobody minds thousands of people being killed a year by human drivers. However, if you automated every car and only three people were killed every year everybody would call the machines death traps and sue the manufacturers into poverty. There is just no sense of perspective, and the only thing that matters is that the 1000 human killers didn't have deep pockets...
I'd be shocked if the insurance terms didn't cover stuff like this. Granted, I don't know what satellite policies typically look like, but for most general policies there is a duty of the insured to work to protect against future loss.
So, if a tree falls and smashes my roof, that might be covered. However, if I don't do anything about it for a month and it rains 3 times and destroys half the contents of my home, that would probably not be covered. If I did a makeshift repair and a really big windstorm tore it apart and the rain still got in, then I'd probably be covered, since I did what I could under the circumstances.
You don't need to run into a tornado to patch your roof or anything, but you can't just be negligent in doing what any homeowner would do if they were paying out of their own pocket.
If the policy doesn't discourage satellite owners from just sitting back and watching their satellites plummet, then I'm sure future policies will be adjusted accordingly...
Doh - make that 999 have nothing!
Here is the easiest explanation to follow that I've heard. Extend the game to 1000 doors.
One door has a car, 990 have nothing.
You pick a door. Monty opens 998 of the other doors showing nothing. Which door would you pick? He essentially gave away the location of the car - you only had a 0.1% chance of winning, but he eliminated 998 incorrect choices. The chance of the car being behind the last remaining door is 99.9%. This way it actually is somewhat intuitive.
The 3-door game is just the same game but reduced in size.
Honestly, I'm not sure that any transition would need this level of central planning.
I'd have some blue-sky research funding (the kind that never gets funded industrially), and some general tax incentives. And I'd use tariffs effectively:
1. Ramping up over three years a tariff on imported oil would be imposed to cover 100% of all external costs associated with oil imports (such as maintaining an army in the Middle East and safe sea lanes to and from).
2. Ramping up over 15 years a tariff on undesirable non-renewable energy of all kinds. This will be to ensure the strategic strength of the nation and to cover hard-to-measure externalities like global warming. I used the term undesirable since even nuclear power is non-renewable but it is preferable as an interim source to most things we are burning today.
3. Along with both of those tariffs there would be import tariffs on any goods from nations that don't take similar measures - this is to avoid punishing industry that doesn't relocate to places where they can freely pollute. Treaties would be sought with other industrialized nations to take similar steps, and any nation that does not participate equally would be sanctioned in trade.
4. I would publish well in advance the schedule for all the tariffs, so that industry can plan appropriately. If somebody is building a factory that will last 30 years, they'll be proactive about choosing how it will be powered.
Maybe some incentives for early adopters would be pursued as well. A market based approach should be very effective if given time to work, and as long as it is followed rigorously. Disaster would ensue if major industry decided to gamble on the government losing its nerve. I'd advocate increasing the tariffs every quarter just so that the momentum is continuous and steady. If you keep everything low and then have a huge hike then there will be a lot of pressure to keep delaying the hike.
Well, you don't have to have all that money can buy to benefit from the US health care system. Anybody who is upper-middle class with half-decent insurance benefits quite a bit from the standard of care.
I agree that for those who are poor the care available is far lower than in socialized nations.
The issue isn't really quality of care so much as the philosophy behind how it is apportioned. In Europe it is based on "fairness" - mostly FIFO, and in the US it is based on how much it is worth to you (more cash -> more quality). However, I'd question whether a typical MP in Europe waits as long for a medical procedure as the average working-class citizen. In most socialist nations they're really quite capitalist, but instead of paying with money you pay with political power/access/etc.
Personally I think that better care for the poor is important, but if you don't provide any benefits to actually working, then nobody will bother to work. I'm not sure the US has the right balance, but I know that Europe has issues of its own. And in the US when the evil insurance company puts the brakes on care at least there is a government to appeal to - with socialized care the government is your insurance and if you don't like it you can go live elsewhere...
How about this - send each mailing address a form with 1 question - how many people live at this address. It goes on a postcard-sized business reply card that you just drop in a mailbox.
Why do we need a huge questionaire to determine congressional representation?
Or better still, why not just ask a credit reporting agency how many people there are at what address - they probably know what house I'm going to buy next better than I do!
Well, clearly you don't have Intuit/etc charging $15 to upload tax data. I paper file every year for this reason. A fancy computer program generates the return and then I mail it, because I refuse to pay money for something that is otherwise free...
Most US states on the other hand allow you to file online. I'm not asking for fancy tax software - just a simple web-based form where I can type in the numbers and hit submit. Congress keeps saying stuff like "the US ought not to be in the tax preparation software business." I say two things:
1. Fine, don't write fancy software, just create simple web forms that let you type in the numbers that you'd otherwise hand-write onto a paper form.
2. If Congress didn't make the tax code so darn complicated maybe we wouldn't need software to do tax preparation. Most state/local tax forms take all of 2 minutes to fill out.
Ah, yes, instead of paying police to catch people who are an actual danger to society, we can all take weeks off of work to fund our own effort to police the police.
Maybe elected officials should be taken to task to uphold the law in the first place...
Sure, but the dispute is that on the lights with shortened yellows it is simply not possible to safely stop in time. If you're in the wrong spot when that light turns yellow you're either jamming on the brakes or getting a ticket. That's about all there is to it...
Just google for yellow light timings and traffic cameras...
Alternatively, the laws could be set at a reasonable level. When 90% of the population violates a law, perhaps the law is inappropriate?
That is actually one thing I'd like about speed cameras - it would force the laws to adjust at least a little to reality. When the average citizen gets 3 speeding tickets a month there will be a riot...
Is there any reason that limits couldn't be set just a little higher? The argument that is always made is safety. So, why not just ban cars altogether, or make the limits 15mph - a lot fewer people would be killed by cars that way, after all! The issue is that productivity does matter too - and most people seem to be voting with their actions that the current balance is off-kilter.
The military would most likely need to reverse engineer Storm to take it down, so they would be brushing up on some much-needed reverse engineering skills, as well as creating new tools to help automate the process of reverse engineering programs and looking for exploits. THAT is what they would need to keep secret, and it wouldn't be exposed by releasing a program they wrote to take advantage of a specific exploit they found.
But, they could brush up on the same skills by developing something capable of taking down Storm, and then holding it in reserve for some time when they would need to take down Storm, and not before. Then, when some attacker makes a huge strategy around using Storm to take out us infrastructure the US can just take it down at that time and the attacker might not have a fallback plan. If Storm were gone something else would rise up to take its place - perhaps something harder to take down.
In warfare you want to keep your enemy complacent, and the best way to do that is to not attack them while amassing your armies...
Only issue with brushing up is that they would also be advertising their tactics.
Back during the cold war it wasn't uncommon for US forces to intentionally run their radars in a mode good enough for general navigation, but limiting their true capability. That way, in the event of war the enemy wouldn't have all kinds of intel on the frequencies used and general technology.
The same would apply in warfare. If the US started taking out botnets all kinds of folks would be capturing packet dumps and discussing new zero-day exploits and methodologies. Then everybody would engineer defenses against them. Generally speaking, you don't want to use weapons that you don't want the enemy to know about.
The issue is that having a ton of money in the bank lets you "take risk" while in reality being quite safe.
For example, I work a standard salaried job. I could probably make quite a bit more if I became self-employed - I'm quite competent and I think I could market my services. However, there would be a substantial risk of periods of non-employment - especially while getting started off. As the sole significant income for my family I can't just neglect my responsibilities and shoot for the moon, so I accept a comfortable salary that isn't likely to have huge swings in it.
On the other hand, if I had $50M in the bank aside from the obvious investment income my own personal income would probably be much higher, as I could afford to take bigger risks since a year without work wouldn't be a big deal. I'd probably be a lot happier for it as well.
Sure, that makes me less of a risk-taker than some billionare I guess, but only on an absolute scale of measurement. On a relative scale I take bigger risks than Bill Gates any time I take my family on vacation. In the same way I'm not super-eager to shower some billionare with accolades every time they donate $10k to some charity - sure, for the average person that is a lot of money, and on some level it is to be commended even for the rich, but the fact is that Gates/etc could easily lose $10k in change in his sofa and not notice it except for the huge bulge in the cushions...
I don't really buy into the whole tax-shelter-due-to-risk argument. I agree that higher risk needs to have higher reward to encourage investment. However, I think that the current ratio is a bit skewed.
Oh, I know it happens. But something more standardized would be useful.
I'd say the capability is there when you can shut down your webserver, boot it back up, and then touch absolutely NOTHING and have users able to download anything on the site. Oh, and all this on a server that doesn't have X11 installed.
GUI clients really shouldn't be part of server infrastructure. The whole thing should run from init.d and config files.
A nice PHP frontend like you say would be nice, but maybe not completely essential. What is essential is automated operation.
Uh, I didn't post that anonymously. Feel free to look at my posting history. I hadn't even read that article until you posted it.
Before you accuse somebody of astroturfing you might want to see if your accusation has any grounding in reality...
Yeah - it would be nice if somebody created a server-oriented bittorrent distribution system.
The website would post a torrent, and would also seed the torrent. If nobody else seeded it then the website would end up uploading the file to anybody who asked for it - which is no worse than what they'd otherwise end up doing. However, as soon as more than one person starts downloading at the same time you get automatic load-distribution, and if anybody sticks around and seeds then you get even more bandwidth.
All you need is a decent daemon that will take a file and create a torrent and track it and seed it, and restore all this stuff anytime the daemon is restarted. I can't find anything that does exactly that...
Uh, he was talking about USB. The post he replied to suggested that scsi wasn't any worse than USB in numbers of connectors.
USB has 4. SCSI has some number that might be countable using the set of integers - I'm not sure if that has been proven yet.
Even so, USB probably should only have 1 connector - or maybe 1 big and one small. Yes, I know the others have their reasons for existing, but I'm sure there is another way to do things...
Uh, Java? Sure, it has limitations as well but at least the language is well-documented and you can get a VM for just about anything...
Forget the factories - a certain block of concrete holding back a flood of water the weight of the British Isles comes to mind...
That and the fact that software inevitably uses the mysql runtime libraries and you'd need to patch half an application to make it work in postgresql as a result.
For all that MS does wrong, one thing they did right is getting everybody to use ODBC for everything (well, almost everything). If you stick to ODBC and ANSI SQL your app will work with just about any backend. Sure, similar solutions might exist for unix, but nobody uses them.
If mythtv officially supports anything other than mysql then maybe I can consider switching. However, I really don't want to run two databases for everything!
The issue wasn't so much partnering with a for-profit, but not being smart about how they did it.
If you just take a for-profit's product and retool it, then don't be upset if they discontinue the model and leave you stuck - they aren't under any obligation to sell it forever.
On the other hand, writing up a contract between an open-source and for-profit organization where each benefits and has defined responsibilities should be safe. Perhaps the open-source develops the design/code, and the for-profit can make a killing selling it to companies with deep pockets but has to allow cheap/free use for certain purposes. Or you could just develop full blueprints and make them open-source, and pay per unit for actual production.
For-profit companies are often more efficient and can achieve higher economies of scale. On the other hand, they don't share your mission, so you're an idiot if you just lock yourself in without any legal protection.
In this particular case things are a bit more sleezy since the company didn't just stop supplying new equipment, but they went out of their way to break already-sold stuff as well.
Regardless - even if the console did display the error I wouldn't be able to report it short of taking a photo of the monitor...
:)
Don't get me wrong - I'd be happy to go open-source with video drivers. My understanding is that the competitors haven't really been comparable in performance - although perhaps that has changed. When my current $15 video card dies maybe I'll try a $15 AMD card instead...
And the box actually is a firewall already. It is by general-purpose backend server (and linux desktop)... I don't actually watch TV on it.