The one thing I've leared [sic] is that nothing beats the combination of an Intel CPU on an Intel Motherboard [for stability].
I agree entirely. I generally put this down to Intel's higher standard of documentation and errata notes, which means that good drivers can be developed more easily, and can work around any hardware bugs early on.
In response to the poster who suggests nVidia's nForce; I'm happy that their hardware quality and drivers are pretty good, but I'm not happy to be "held ransom" by a single vendor. If nVidia go bust/lose interest in Linux/cosy up to Microsoft in order to win the contract back from ATI for XBox 3, and the Linux kernel API changes, will their drivers carry on working?
Please describe the world situation where it would be in Japan's interest for the USA to collapse overnight. Or Germany's interest, or the UK's interest. If the USA 'collapsed' the entire world would plunge into economic ruin and chaos. It ain't going to happen.
They might want to trade in their stock of dollars for euros if oil becomes increasingly traded in euros, rather than US dollars. Of course, there will be a trade-off between that and the implications of the US economy collapsing, which each government will need to weigh up for itself.
Interestingly, Iraq under Saddam Hussein started trading its oil in Euros since November 2002 (and they did very well - financially speaking - out of doing so, too!) Iran and Venezuela have also reportedly considered the possibility.
Remember that Red Hat (wherever possible and sensible) backport fixes into older versions so as to minimize the risks in applying their updates. Thus the 8.12.9 errata that you installed may well have been just as secure as 8.12.10. You need to reed the Red Hat advisories to see what CVE/CERT vulnerabilities they fix. For more background on Red Hat's backporting policy, read this and this.
Incidentally, I gather Debian takes broadly the same approach. The alternative is to blindly supply the latest version, risking Microsoft-style breakage with errata/hotfixes.
...but honestly, Red Hat (and now Fedora) is light years (or about 6-12 months, at least!) ahead of Debian on the desktop. From what I hear, Gnome 2.4 might possibly make it into the next stable release of Debian. Fedora, released last week, already includes it, along with OpenOffice 1.1.0 and Mozilla 1.4.1. Anti-aliased fonts work out-of-the box too. And Fedora is no less Free than Debian. And Fedora has GUI system config tools (including for printing, samba and networking) that are well integrated and work reliably (Debian might well have equivalents too, there, of course).
With those things in mind, surely it would be better for pro-desktop hackers to concentrate on Fedora, stealing good stuff from Debian wherever that's appropriate?
What I was saying was that if the Tablet PC manufacturer had the tablet/pen portion of their hardware use the Wacom protocol (and I can't think of any reason why they couldn't), then drivers already exist.
There's some serious stuff in there, all the modern tablets have pressure sensitivity (the "button" is on the tip of the pen, not the screen like with palm pilots) and I'm pretty sure that Linux doesn't have drivers to support that last time I looked, so doodling or professional drawing won't work (Penny Arcade is drawn on a tablet, for example).
Wacom's graphics tablets are fully supported by Linux using these drivers. If a tablet PC manufacturer isn't arrogant enough to adopt a Not Invented Here attitude and instead uses the Wacom protocol, their products can work with Linux, today.
All that's needed is some handwriting-input software.
To make that analogy more accurate, the shop is a few weeks walk away and so you need a certain amount of energy from the pistachios in the room so as not to starve to death on the journey.
If we leave it too late to migrate to the next generation of (hopefully sustainable/renewable) energy sources, we'll find ourselves stranded without the energy to do so at all. Like running out of fuel on the motorway because we didn't want to take a 10 mile detour to refill the tank.
Another, less hypothetical example. Tomorrow, a hijacker takes over an airplane. There's a non-zero chance (though not absolutely certain) that he wants to pull a WTC. Given that the airplane has fifty people on board, and there are ten thousand in the building, do you, as the C-in-C, tell the F-16 driver to shoot it down?
First you tell the hijacker that if they're not planning on "pulling a WTC" they'll follow your instructions
Next, tell them to fly away from potential targets
Any deviation from the proscribed flight plan will result in whatever action is necessary (including shooting down) to prevent a WTC-style attack
If the hijacker follows instructions, they get to live and they might even get some of their demands met. Alternatively, you have the option of getting the plane on the ground and storming it. The ball's in your court.
Any smart hijacker should be aware of all of the above, post-9/11, and should know that playing along is the smart strategy if they're not planning on "pulling a WTC"
It wouldn't surprise me if that's a rough approximation of post-9/11 aircraft hijack response these days. As a result, I wouldn't be surprised if the crime rapidly disappears from terrorist groups' lists of tactics.
Remember that you still have the standard Fair Use rights allowed by your jurisdiction. These will probably allow you to study and write your own code based on GPLed code, as long as you don't include actual GPLed code.
The WD 'special edition' drives are a good example. Several tens of dollars more for $2 worth of semiconductors.
...and two more years warranty. For an extra ~11-14% compared to the plain -BB models, I reckon that's a bargain. Hardly "several tens of dollars more". More like "several dollars".
What are musical instruments, cameras, consumer electronics and cars made from? Lots of ingenuity and a (relatively) small amount of natural resources.
Indeed, the CIA World Factbook says this of Japan - "among world's largest and technologically advanced producers of motor vehicles, electronic equipment, machine tools, steel and nonferrous metals, ships, chemicals; textiles, processed foods". All of those things require large doses of ingenuity.
Let's look at their imports: "machinery and equipment, fuels, foodstuffs, chemicals, textiles, raw materials". Those imports, plus their ingenuity produces wealth.
If they had indigenous supplies of those things, they wouldn't need to import them. But they'd still need them to produce wealth.
Wealth is not a physical object. It doesn't have to be taken from one country to go to another. The fact the the US is the most successful country, doesn't prevent other countries from being extremely wealthly also.
Wealth is a physical object, at least partly. Wealth is derived from the combination of natural resources (land, mineral ores, oil, biomass) and human ingenuity. Whilst human ingenuity is (at least in theory) limitless, natural resources are not.
Sane FW-1 administrators won't be allowing any and all protocols through the firewall/VPN gateway just because they appear to come from an authorised SecureClient VPN client.
Having the remote security policy functionality does allow the firewall administrator to have a reasonable degree of trust in the VPN clients though, to the extent that they probably aren't 0wn3d and being actively used as gateways into the corporate network or whatever. Especially so if the clients (laptops, usually) are properly locked down to make it hard for their users (legitimate, or otherwise) to disable the real client software and install their own hacked version. And, to be honest, if your biggest risk is collusion (which installing a hacked VPN client is), you're screwed anyway.
That's one of the unique features I do actually like about CheckPoint's FireWall-1 suite; their SecureClient VPN client software allows the firewall administrator to push firewall policies to be enforced locally on hosts intended to be VPN clients.
It's not perfect of course, as a host could be compromised before SecureClient is installed, but in a controlled environment, that should never really be the case.
First of all, you can't just print out each vote in a long list. Each voter must be given their own voting slip so that they can manually, and privately, confirm that their vote has been registered correctly. (I'm not sure whether you were advocating a "long list" or not, I just figured I'd get this out in the open, regardless).
You're therefore going to have to make sure they can't photocopy/xerox the slip in some way, otherwise you fall into the "receipt for vote selling" trap I initially described.
You also need to make sure the voter doesn't modify their paper vote, otherwise they can, say, vote Green electronically, then change their paper vote to Democrat (which'll presumably get counted instead, if the election result is close). Whilst I'm strongly in favour of transferrable/proportional voting systems, allowing people to manipulate the system we currently have probably isn't a good idea. Incidentally, you can't use a graphical checksum or similar, because of the earlier requirement that the voter should be able to inspect their vote manually.
I'd say that the UK system of using a pencil to put a single X in a box opposite the candidate you want to vote for is pretty foolproof. The KISS principle applies. I've no idea why there's such an obsession with trying to use various contraptions in our voting processes.
Moving on to your second point, if you're still going to use the electronic count as the primary count, why would you resort to counting bits of paper? The best you could do is random audits to check that the paper count agrees with the electronic count. If you're going to use the paper vote as the primary count, why bother doing it electronically at all? You might as well use (much more reliable and easier to use) pencil and paper.
Finally, in the UK, spoiling your ballot paper is a common way of showing disapproval for all the candidates standing in an election. We don't have "write-in candidates" but some people feel better if they make the effort to show up to vote and spoil their ballot paper, rather than simply not voting at all. I don't agree, but I can understand why they feel that way.
As of 27 June, it would appear that the Crown Prosecution Service was still deciding whether or not to charge the employee, but she was fired.
If the story was made up, then no official secrets were leaked, and there would be no case for the arrest or any charges.
Of course, being hyper-paranoid, it's possible that the story was made up within some government agency, which would still make it an official secret...
1) I'm glad we don't use the hole-punch voting machines in the UK. They're flawed too, though not as badly as e-voting.
2) I thought the whole point of e-voting was that it was quicker to count votes, and easier to allow remote voting (i.e. by mobile phone, or interactive TV). If you're going to print the votes out and count them manually, you've lost both those advantages.
3) A vote-and-printout system won't allow me to deliberately spoil my vote in order to express disapproval of all candidates. ("Sorry, 'You're all bastards' is not a valid candidate in this election. Please try again!")
...because if a voter gets a printed receipt of their vote, then they can then use that to collect their cheque from anyone who offers to buy their vote.
If you're suggesting that the voter doesn't get a printed receipt, but instead the machine prints out their voting slip, they check it, and post it into a black box, then what's been won over manual voting? (i.e. bits of paper and pencils - thankfully still the way things are done in the UK, but our government is keen on doing the e-voting thing too...)
Like others, I've come to the conclusion that e-voting is a fundamentally less safe practice than manual voting.
Slashdot Mirror
I agree entirely. I generally put this down to Intel's higher standard of documentation and errata notes, which means that good drivers can be developed more easily, and can work around any hardware bugs early on.
In response to the poster who suggests nVidia's nForce; I'm happy that their hardware quality and drivers are pretty good, but I'm not happy to be "held ransom" by a single vendor. If nVidia go bust/lose interest in Linux/cosy up to Microsoft in order to win the contract back from ATI for XBox 3, and the Linux kernel API changes, will their drivers carry on working?
Each to their own, though.
--
They might want to trade in their stock of dollars for euros if oil becomes increasingly traded in euros, rather than US dollars. Of course, there will be a trade-off between that and the implications of the US economy collapsing, which each government will need to weigh up for itself.
Interestingly, Iraq under Saddam Hussein started trading its oil in Euros since November 2002 (and they did very well - financially speaking - out of doing so, too!) Iran and Venezuela have also reportedly considered the possibility.
--
Incidentally, I gather Debian takes broadly the same approach. The alternative is to blindly supply the latest version, risking Microsoft-style breakage with errata/hotfixes.
--
With those things in mind, surely it would be better for pro-desktop hackers to concentrate on Fedora, stealing good stuff from Debian wherever that's appropriate?
--
What I was saying was that if the Tablet PC manufacturer had the tablet/pen portion of their hardware use the Wacom protocol (and I can't think of any reason why they couldn't), then drivers already exist.
--
Wacom's graphics tablets are fully supported by Linux using these drivers. If a tablet PC manufacturer isn't arrogant enough to adopt a Not Invented Here attitude and instead uses the Wacom protocol, their products can work with Linux, today.
All that's needed is some handwriting-input software.
--
If we leave it too late to migrate to the next generation of (hopefully sustainable/renewable) energy sources, we'll find ourselves stranded without the energy to do so at all. Like running out of fuel on the motorway because we didn't want to take a 10 mile detour to refill the tank.
--
First you tell the hijacker that if they're not planning on "pulling a WTC" they'll follow your instructions
Next, tell them to fly away from potential targets
Any deviation from the proscribed flight plan will result in whatever action is necessary (including shooting down) to prevent a WTC-style attack
If the hijacker follows instructions, they get to live and they might even get some of their demands met. Alternatively, you have the option of getting the plane on the ground and storming it. The ball's in your court.
Any smart hijacker should be aware of all of the above, post-9/11, and should know that playing along is the smart strategy if they're not planning on "pulling a WTC"
It wouldn't surprise me if that's a rough approximation of post-9/11 aircraft hijack response these days. As a result, I wouldn't be surprised if the crime rapidly disappears from terrorist groups' lists of tactics.
--
--
I've (un)successfully used this to avoid learning both perl and python for some years. D'oh!
--
--
What are musical instruments, cameras, consumer electronics and cars made from? Lots of ingenuity and a (relatively) small amount of natural resources.
Indeed, the CIA World Factbook says this of Japan - "among world's largest and technologically advanced producers of motor vehicles, electronic equipment, machine tools, steel and nonferrous metals, ships, chemicals; textiles, processed foods". All of those things require large doses of ingenuity.
Let's look at their imports: "machinery and equipment, fuels, foodstuffs, chemicals, textiles, raw materials". Those imports, plus their ingenuity produces wealth.
If they had indigenous supplies of those things, they wouldn't need to import them. But they'd still need them to produce wealth.
--
That's in the kernel included in Fedora (formerly known as the Red Hat Linux Project) by default. It's called exec-shield.
--
Wealth is a physical object, at least partly. Wealth is derived from the combination of natural resources (land, mineral ores, oil, biomass) and human ingenuity. Whilst human ingenuity is (at least in theory) limitless, natural resources are not.
--
Having the remote security policy functionality does allow the firewall administrator to have a reasonable degree of trust in the VPN clients though, to the extent that they probably aren't 0wn3d and being actively used as gateways into the corporate network or whatever. Especially so if the clients (laptops, usually) are properly locked down to make it hard for their users (legitimate, or otherwise) to disable the real client software and install their own hacked version. And, to be honest, if your biggest risk is collusion (which installing a hacked VPN client is), you're screwed anyway.
--
It's not perfect of course, as a host could be compromised before SecureClient is installed, but in a controlled environment, that should never really be the case.
--
I'm not aware of a way to do this undectably with present technology, but it's something that'll need to be considered if printed slips are issued.
Plus, they don't leave the voting area with their slip, and there will not be photocopying machines present anyhow.
OK, cheap digicam instead of Xerox...
--
--
First of all, you can't just print out each vote in a long list. Each voter must be given their own voting slip so that they can manually, and privately, confirm that their vote has been registered correctly. (I'm not sure whether you were advocating a "long list" or not, I just figured I'd get this out in the open, regardless).
You're therefore going to have to make sure they can't photocopy/xerox the slip in some way, otherwise you fall into the "receipt for vote selling" trap I initially described.
You also need to make sure the voter doesn't modify their paper vote, otherwise they can, say, vote Green electronically, then change their paper vote to Democrat (which'll presumably get counted instead, if the election result is close). Whilst I'm strongly in favour of transferrable/proportional voting systems, allowing people to manipulate the system we currently have probably isn't a good idea. Incidentally, you can't use a graphical checksum or similar, because of the earlier requirement that the voter should be able to inspect their vote manually.
--
Moving on to your second point, if you're still going to use the electronic count as the primary count, why would you resort to counting bits of paper? The best you could do is random audits to check that the paper count agrees with the electronic count. If you're going to use the paper vote as the primary count, why bother doing it electronically at all? You might as well use (much more reliable and easier to use) pencil and paper.
Finally, in the UK, spoiling your ballot paper is a common way of showing disapproval for all the candidates standing in an election. We don't have "write-in candidates" but some people feel better if they make the effort to show up to vote and spoil their ballot paper, rather than simply not voting at all. I don't agree, but I can understand why they feel that way.
--
As of 27 June, it would appear that the Crown Prosecution Service was still deciding whether or not to charge the employee, but she was fired.
If the story was made up, then no official secrets were leaked, and there would be no case for the arrest or any charges.
Of course, being hyper-paranoid, it's possible that the story was made up within some government agency, which would still make it an official secret...
--
--
2) I thought the whole point of e-voting was that it was quicker to count votes, and easier to allow remote voting (i.e. by mobile phone, or interactive TV). If you're going to print the votes out and count them manually, you've lost both those advantages.
3) A vote-and-printout system won't allow me to deliberately spoil my vote in order to express disapproval of all candidates. ("Sorry, 'You're all bastards' is not a valid candidate in this election. Please try again!")
--
If you're suggesting that the voter doesn't get a printed receipt, but instead the machine prints out their voting slip, they check it, and post it into a black box, then what's been won over manual voting? (i.e. bits of paper and pencils - thankfully still the way things are done in the UK, but our government is keen on doing the e-voting thing too...)
Like others, I've come to the conclusion that e-voting is a fundamentally less safe practice than manual voting.
--
Well, that all depends on the quantities involved.
--