The crucial part is the 'across three generations' bit - I can tell you from first hand experience that VMWare ESX has problems performing live migrations across CPUs with different steppings even within the same generation, so the fact that AMD pulled it off across 3 distinct generations without having to utilize cobbed-together solutions like EVC is a pretty big deal. Or, at least it is to me.
So is the Library. I don't see why I should put any more faith in the written words of an anonymous stranger simply because it is on paper rather than on my screen. Books can be (and often are) just as biased or distorted as anything found on the 'Interweb'.
Incidentally, the title and summary for this article suck...the OCO didn't fail, it was lost in a launch failure, and it didn't "fail its mission," it didn't get a chance to start. That's like saying your car broke down because someone ran a red light and T-boned it. No offense intended to the launch team.
Hmmm. I'd suspect a better car analogy would be "That's like saying your car broke down because the truck hauling it from the manufacturer to the dealership you just placed the order through fell off a bridge." But perhaps I'm just nitpicking.:)
The problem with that is that you'll still get bloat. Your smart-phone might not have exactly the same hardware as mine, so the package would need to acount for both. Otherwise the distro'ed packages will need to mandate hardware on a ridiculously specific level, vendor-locking people into one platform or there'd be so many different packages that easily finding the one I should be running is a near impossibility.
Do you really think you can drum up someone that will willingly admit that they are likely to delete their own critical files? Of course not. Only someone else would delete my files.
Must be nice to be perfect, eh? Back here in the really real world people can and do make mistakes. I myself have inadvertently typed rm some_file instead of mv some_file - stupid little things like this happen all the time.
If a file gets deleted and is not recoverable, then it is not backed up. Period. If you are just replicating data to another drive, you replicate errors and corruption as well. I don't know about where you work, but if one of my coworkers came in and said "hey, I seem to have lost file such-and-such, and need it back" and I told them it's gone for good since they didn't come to me within 24 hours, I'd lose my job.
Ok, maybe I did leap to my nearest conclusion, but I've seen it happen enough where people assume that because they have RAID, they don't need to think about anything else for backup - your solution is a valid one though, since you are periodically ensuring that the data is stored outside the live array.
If you don't have the data replicated periodically outside the RAID, it's not a backup solution at all - I could, within the OS & filesystem, delete files arbitrarily and the RAID will happily remove those bits from all the drives. Thus, it is not a backup solution. If those bits aren't anywhere else, they are gone for good.
Then you are just biding your time until something really subtle happens that screws the whole thing up. For example. what happens if the filesystem becomes corrupt? That corruption is propagated across your entire RAID array, leaving you with 4+ fully functional yet useless drives and you'll be left kicking yourself for not performing routine backups.
If I had mod points, they'd be yours. It boggles the mind how many people equate RAID to backup - RAID only ensures data availability, not integrity. If the bits get fubar'ed on one drive, that will propagate across the rest - leaving you with no data.
I think it's worth pointing out that what you suggest as a solution implies RAID 5, which is not a backup solution. There was an article on slashdot in the recent past about someone making that same mistake, and losing their entire hosted blog site (if someone could link the article that would be great - I'm having a heck of a time finding it).
OpenBSD is an OS designed from the ground up with security as its main focus. As such, the developers spend a lot of time reviewing 3rd party apps (such as KDE) to ensure that porting the software into the OS does not compromise that stance. As a result, many mainstream apps that make it into their port list are a few revisions out of date. If you want all the hot new whiz-bang apps, choose a different platform (FreeBSD, for example).
Also, roads? They are directly funded by drivers. If you don't drive, you don't pay the road tax which is collected at the gas pump.
Ah, but even those that do not drive or own a car still benefit from use of the roads (public transportation, sidewalks to walk on, stuff that gets delivered to their homes arrives via roadways, etc.). So by your initial point, roads should be provided for as part of general taxation, and not a 'fuel' tax at the pump.
I'm sure most of us here can come up with a dozen ways of making voting machines far more secure.
No one is arguing that point, but even if we follow all of your costly (both time and $$) security suggestions, they're all thrown out when one of your developers writes a stealth backdoor into the stupid thing. Nothing can be 100% secure, it's all about the trade offs - you make it as hard as economically possible to deter would-be attackers. But unfortunately if someone wants in, they'll get in - and in the case of vote tampering, there's a significant number of people who do.
I think what Google is trying to accomplish is free market at its finest. After being frustrated by trying to develop tools for other browsers, they feel that the tools they created would work better for the end-users if they made their own browser optimized for their code. If they then take a significant user base from Firefox, IE, and Opera then one can surmise that the devs (for Firefox and Opera at least) will then make efforts to make their browsers more compatible with Google's tools, since the market is clearly demanding it - which in the end will also give Google what they want, better support in browser for their software. I don't know why everyone's jumping to the conclusion that Google is trying to 'take over the Internet' or strong-arm competition away. In the end it sounds like we as end-users will all win, as long as the developers for your browser of choice listen to the demands of their users.
Too bad an infrastructure like that would never get off the ground as all the advertisers scream bloody murder, and the Washington cronies that are beholden to them will cave in yet again. Kind of like how they've tried time and again to work around PVRs skipping their commercials.
If by 'round features' you mean rolling hills, smooth rocks, etc. there is a good reason for that. Such features here on Earth were created over millions of years through the subtle movement of continent-spanning glaciers. Whether or not Mars ever had sustained liquid water, I doubt it ever had enough to have glacial features spanning a majority of its surface for any length of time. Now get one of those rovers up towards the poles and it could be a completely different story - there's plenty of ice-like material there that could be performing glacial-like erosion.
I'm not installing a new system until time and angle measurments get upgraded to base 10.
Unfortunately, time can never really be ported to a true base 10 system once you get to a large enough scale. A day is defined as the amount of time the Earth takes to complete 1 full rotation. No matter if you slice that day up into 10ths or 24ths to define an hour, it still takes 365.25 days to complete 1 trip around the Sun. Kind of throws a monkey-wrench into easy conversions, doesn't it?
Um, no. If cops were required to be visible when enforcing speed limits, then people would tend to drive more recklessly when there is no cop in site, because there is obviously no longer a risk of getting caught. The function of a speed trap is to get people slowing down even if there is no cop present - because there's a chance that there might be one there that you can't see. If we went with your idea of having them all out in the open, then there'd have to be a cop parked along every route every few hundred yards in order to ensure that people didn't just go back to driving too fast once the cop is out of view. Good luck with getting THAT enforced.
Slashdot Mirror
The crucial part is the 'across three generations' bit - I can tell you from first hand experience that VMWare ESX has problems performing live migrations across CPUs with different steppings even within the same generation, so the fact that AMD pulled it off across 3 distinct generations without having to utilize cobbed-together solutions like EVC is a pretty big deal. Or, at least it is to me.
the Interweb is full of lies and damned lies.
So is the Library. I don't see why I should put any more faith in the written words of an anonymous stranger simply because it is on paper rather than on my screen. Books can be (and often are) just as biased or distorted as anything found on the 'Interweb'.
Incidentally, the title and summary for this article suck...the OCO didn't fail, it was lost in a launch failure, and it didn't "fail its mission," it didn't get a chance to start. That's like saying your car broke down because someone ran a red light and T-boned it. No offense intended to the launch team.
Hmmm. I'd suspect a better car analogy would be "That's like saying your car broke down because the truck hauling it from the manufacturer to the dealership you just placed the order through fell off a bridge." But perhaps I'm just nitpicking. :)
The problem with that is that you'll still get bloat. Your smart-phone might not have exactly the same hardware as mine, so the package would need to acount for both. Otherwise the distro'ed packages will need to mandate hardware on a ridiculously specific level, vendor-locking people into one platform or there'd be so many different packages that easily finding the one I should be running is a near impossibility.
Do you really think you can drum up someone that will willingly admit that they are likely to delete their own critical files? Of course not. Only someone else would delete my files.
Must be nice to be perfect, eh? Back here in the really real world people can and do make mistakes. I myself have inadvertently typed rm some_file instead of mv some_file - stupid little things like this happen all the time.
If a file gets deleted and is not recoverable, then it is not backed up. Period. If you are just replicating data to another drive, you replicate errors and corruption as well. I don't know about where you work, but if one of my coworkers came in and said "hey, I seem to have lost file such-and-such, and need it back" and I told them it's gone for good since they didn't come to me within 24 hours, I'd lose my job.
Ok, maybe I did leap to my nearest conclusion, but I've seen it happen enough where people assume that because they have RAID, they don't need to think about anything else for backup - your solution is a valid one though, since you are periodically ensuring that the data is stored outside the live array.
If you don't have the data replicated periodically outside the RAID, it's not a backup solution at all - I could, within the OS & filesystem, delete files arbitrarily and the RAID will happily remove those bits from all the drives. Thus, it is not a backup solution. If those bits aren't anywhere else, they are gone for good.
Then you are just biding your time until something really subtle happens that screws the whole thing up. For example. what happens if the filesystem becomes corrupt? That corruption is propagated across your entire RAID array, leaving you with 4+ fully functional yet useless drives and you'll be left kicking yourself for not performing routine backups.
If I had mod points, they'd be yours. It boggles the mind how many people equate RAID to backup - RAID only ensures data availability, not integrity. If the bits get fubar'ed on one drive, that will propagate across the rest - leaving you with no data.
I think it's worth pointing out that what you suggest as a solution implies RAID 5, which is not a backup solution. There was an article on slashdot in the recent past about someone making that same mistake, and losing their entire hosted blog site (if someone could link the article that would be great - I'm having a heck of a time finding it).
There's a small loophole there - CCTV's do not use film!
Neither does the digital camera built into your phone.
OpenBSD is an OS designed from the ground up with security as its main focus. As such, the developers spend a lot of time reviewing 3rd party apps (such as KDE) to ensure that porting the software into the OS does not compromise that stance. As a result, many mainstream apps that make it into their port list are a few revisions out of date. If you want all the hot new whiz-bang apps, choose a different platform (FreeBSD, for example).
Also, roads? They are directly funded by drivers. If you don't drive, you don't pay the road tax which is collected at the gas pump.
Ah, but even those that do not drive or own a car still benefit from use of the roads (public transportation, sidewalks to walk on, stuff that gets delivered to their homes arrives via roadways, etc.). So by your initial point, roads should be provided for as part of general taxation, and not a 'fuel' tax at the pump.
I'm sure most of us here can come up with a dozen ways of making voting machines far more secure.
No one is arguing that point, but even if we follow all of your costly (both time and $$) security suggestions, they're all thrown out when one of your developers writes a stealth backdoor into the stupid thing. Nothing can be 100% secure, it's all about the trade offs - you make it as hard as economically possible to deter would-be attackers. But unfortunately if someone wants in, they'll get in - and in the case of vote tampering, there's a significant number of people who do.
You still have plenty of developers continuing to innovate. Steam, ID, Crytek to name a few.
Steam is just software. Valve is the developer.
I think what Google is trying to accomplish is free market at its finest. After being frustrated by trying to develop tools for other browsers, they feel that the tools they created would work better for the end-users if they made their own browser optimized for their code. If they then take a significant user base from Firefox, IE, and Opera then one can surmise that the devs (for Firefox and Opera at least) will then make efforts to make their browsers more compatible with Google's tools, since the market is clearly demanding it - which in the end will also give Google what they want, better support in browser for their software. I don't know why everyone's jumping to the conclusion that Google is trying to 'take over the Internet' or strong-arm competition away. In the end it sounds like we as end-users will all win, as long as the developers for your browser of choice listen to the demands of their users.
Too bad an infrastructure like that would never get off the ground as all the advertisers scream bloody murder, and the Washington cronies that are beholden to them will cave in yet again. Kind of like how they've tried time and again to work around PVRs skipping their commercials.
Anyone noticed the absence of round features?
If by 'round features' you mean rolling hills, smooth rocks, etc. there is a good reason for that. Such features here on Earth were created over millions of years through the subtle movement of continent-spanning glaciers. Whether or not Mars ever had sustained liquid water, I doubt it ever had enough to have glacial features spanning a majority of its surface for any length of time. Now get one of those rovers up towards the poles and it could be a completely different story - there's plenty of ice-like material there that could be performing glacial-like erosion.
A thousand dollars. He sold it six times, then... five times... six times... wait, $5,600?
One would think that the missing $400 went towards the overhead necessary to list and sell the app - Apple's gotta get their cut after all.
You're still hosed if your server's power supply goes titsup. Or if your hard drive crashes. Or if the building burns down.
Redundant power supplies, RAID, off-site replication.
But thanks for playing.
One would think if he were here, he'd be muttering something about brains, and his desire to eat them.
I'm not installing a new system until time and angle measurments get upgraded to base 10.
Unfortunately, time can never really be ported to a true base 10 system once you get to a large enough scale. A day is defined as the amount of time the Earth takes to complete 1 full rotation. No matter if you slice that day up into 10ths or 24ths to define an hour, it still takes 365.25 days to complete 1 trip around the Sun. Kind of throws a monkey-wrench into easy conversions, doesn't it?
Um, no. If cops were required to be visible when enforcing speed limits, then people would tend to drive more recklessly when there is no cop in site, because there is obviously no longer a risk of getting caught. The function of a speed trap is to get people slowing down even if there is no cop present - because there's a chance that there might be one there that you can't see. If we went with your idea of having them all out in the open, then there'd have to be a cop parked along every route every few hundred yards in order to ensure that people didn't just go back to driving too fast once the cop is out of view. Good luck with getting THAT enforced.
I'm not as concerned about getting black market body parts as I am about the consequences of missing a payment.