>Attach a kernel debugger (automatically disables the feature). For _users_?
>Use a self-signed cert. Nope, won't work. MS requires CA's certificate to be signed by cross-certificate from MS.
>Use a commercial test-cert. >Use a WHQL test-cert. All problems of test certs, so nope.
>Use a self-signed cert and then configure the domain CA to trust and distribute the cert. Won't work, as I said.
>It's trivial and cheap to get a cert for distribution, and trivial to use a self-signed cert when you dont need to distribute publicly. Nope, you can't.
>Do you actually know anything about the subject you're so cavalierly making judgements about?
Yes. I developed several drivers.
>Have you ever acquired a code-signing cert such that your product can be installed in this way?
Yes. It required me to register my company and gather A LOT of documentation proving that I'm not a Nigerian scummer.
>Do you realize that Microsoft plays no part in the acquiring of such a cert, or the signing of your drivers?
Yeah, sure. And it wasn't Microsoft who thought about it. And it's DEFINITELY not Microsoft who makes TOS for certificates. And of course, Microsoft's cross certificate is not required to sign CA's certificates (so you can't set up your own local CA).
>Do you realize that your Windows Vista x64 can be configured to NOT require signed certs?
How, pray tell me? Of course, turning on test certificate doesn't count.
My computer can run Ubuntu, no problem. But I have a lot of critical Windows-only applications. I can use XP in virtual machine for now, but Vista disallows virtualization (except for most expensive versions) - so this loophole will be closed soon.
I don't care about userspace drivers (I'm porting FUSE to Windows in my spare time, for God's sake!).
But Vista requires _signed_ _kernel-mode_ _drivers_. It won't load unsigned drivers, and there's NO user override for this 'feature'. Let me repeat: Microsoft does not allow you to run some types of code on your computer.
You can turn on 'test certificate root' which allows to use self-signed certificate, but it is hard to do for a common user, causes DRMed content to stop playing and displays 'test mode' icon.
It's much easier to make autopilot for airborne vehicles than it's to create a self-driving car. So I hope we'll have fully automated flying machines with computer autopilots which CAN NOT be turned off in midair.
It's impractical. Losses are too large - you'd be better just transmitting power over the conventional power lines.
Electrolysis efficiency is about 10%, so it's out. Hydrocarbons synthesis is fairly efficient (but it is not easy) but then you'll have to burn them and convert to electricity with about %40 efficiency. Oh, and you need to start from a fairly 'concentrated' carbon for hydrocarbon synthesis - you take a piece of carbon, heat it with water and get CH4 (methane) and some CO/CO2 and some other hydrocarbons. And guess what is most often used as a carbon source?
The most efficient way to store energy now is hydro-accumulators: you pump water uphill at night and use it to drive generators during day. But it requires.
Read about 'baseload capacity'. It's not practical to use 100% wind power, because in this case you'll either have to transmit surplus of power over the long distances (inefficient!) during high winds or have energy deficit during low winds.
The most optimal ratio for wind power is about 40%/60% (60% - baseload capacity). In some places with good stable and windy weather it can be up to 60%/40%.
First of all, vanadium is NOWHERE close to carbon in abundance. And you need lots of energy to extract it. Calculate how many hundreds of millions (yes, that much) of tonnes of vanadium you'll need. And then calculate a small thing called 'environment footprint' of industry large enough to support vanadium battery infrastructure.
>We need to adjust our direction of civilisation away from more toys and gadgets to higher quality human interactions and more meaningful labour. >Sorry all you PR saps and admin assistants at hedge funds and nail salon operators. I would recommend you learn something useful, like FARMING. Or dismantling Las Vegas and Phoenix.
So, why are you using a computer and reading Slashdot (wasting untold number of man-hours which can be put to FARMING). Go and start digging (with spade, no complex tools for you!). I grew up on a farm (for several years), and believe me - it's not an easy task at all.
Why? Because I live in Russia. Believe me, popular support for Putin here in indisputable. It's REAL. I don't like it at all, but it's real.
To be fair, Putin restored at least some semblance of order. During Yeltsin's last days country was rapidly approaching another collapse (for example, several provinces added to their local constitutions ability to form independent militia), Chechens were openly kidnapping and torturing people (in the best traditions of Iraqi insurgents) and so on.
I still don't think that Russian special services killed Litvinenko. And it's not like anyone believed that the United Russia will not abuse his power...
The most interesting question: why have they done that? I live in Russia and nobody here really doubts that the ruling party ("United Russia" - "Edinaja Rossija") influenced elections.
The real approval rates of other opposition parties (communists excepted) were in single percents, anyway. And the real approval rate of United Russia was high enough - all manipulations possibly resulted in several extra seats in parliament for them. So it's not that Putin seriously risked losing his power.
The best way to authenticate a person is to use physical documents (ID card/passport with photo). Usually, passports have pretty good forgery protections. If you can't be present in person - use notarized documents sent by snail-mail. That should cover 'important' things like getting a mortgage or buying a car.
For less important things (like buying stuff on the Intertubes) the current system works pretty well. The occasional card thefts can be mostly eliminated by things like RSA tokens.
The current practice of authentication using SSN+date_of_birth is insane and should be stopped.
Slashdot Mirror
So? Microsoft can, for example, disable any copy of Windows if it detects virtualization (it's very much detectable), like they do with modded XBoxes.
In any case, that would be a copyright violation.
>F8 on boot and disable the feature.
Every time?
>Attach a kernel debugger (automatically disables the feature).
For _users_?
>Use a self-signed cert.
Nope, won't work. MS requires CA's certificate to be signed by cross-certificate from MS.
>Use a commercial test-cert.
>Use a WHQL test-cert.
All problems of test certs, so nope.
>Use a self-signed cert and then configure the domain CA to trust and distribute the cert.
Won't work, as I said.
>It's trivial and cheap to get a cert for distribution, and trivial to use a self-signed cert when you dont need to distribute publicly.
Nope, you can't.
>Do you actually know anything about the subject you're so cavalierly making judgements about?
Yes. I developed several drivers.
>Have you ever acquired a code-signing cert such that your product can be installed in this way?
Yes. It required me to register my company and gather A LOT of documentation proving that I'm not a Nigerian scummer.
>Do you realize that Microsoft plays no part in the acquiring of such a cert, or the signing of your drivers?
Yeah, sure. And it wasn't Microsoft who thought about it. And it's DEFINITELY not Microsoft who makes TOS for certificates. And of course, Microsoft's cross certificate is not required to sign CA's certificates (so you can't set up your own local CA).
>Do you realize that your Windows Vista x64 can be configured to NOT require signed certs?
How, pray tell me? Of course, turning on test certificate doesn't count.
Yes, that's why I mean 'no user override'.
My computer can run Ubuntu, no problem. But I have a lot of critical Windows-only applications. I can use XP in virtual machine for now, but Vista disallows virtualization (except for most expensive versions) - so this loophole will be closed soon.
I know I'm overreacting :)
But I really don't want to live with locked-down computers in 10 years from now.
I don't care about userspace drivers (I'm porting FUSE to Windows in my spare time, for God's sake!).
But Vista requires _signed_ _kernel-mode_ _drivers_. It won't load unsigned drivers, and there's NO user override for this 'feature'. Let me repeat: Microsoft does not allow you to run some types of code on your computer.
You can turn on 'test certificate root' which allows to use self-signed certificate, but it is hard to do for a common user, causes DRMed content to stop playing and displays 'test mode' icon.
Userspace drivers are very limited (that's why they are _userspace_) - they can't do anything that requires more than PASSIVE IRQL.
Vista x64 is the first step to locked-down systems, so it should be boycotted.
Vista x64 is a regression in functionality. For one thing, it _requires_ signed drivers.
It's much easier to make autopilot for airborne vehicles than it's to create a self-driving car. So I hope we'll have fully automated flying machines with computer autopilots which CAN NOT be turned off in midair.
How about loading unsigned drivers on Vista 64?
Oh, you'd say that it's insignificant. But it has _already_ bit me once.
Yes, and a pine is not a plant - it's a tree.
Representative Republic is _a_ _form_ _of_ _democracy_.
You can't get piston move faster than the speed of sound in water. It's about 1km/s, or 1/8 of orbital speed - not much.
And you'll need A LOT of energy to submerge the launch tube, so it won't be a win at all.
Ogg Theora sucks but Ogg Vorbis and Speex are, arguably, the best codecs for audio.
So what? You still need A LOT of vanadium.
Actually, the most practical way is to use lead-acid batteries - they have the best power/weight ratio (not the best energy/weight ration, mind).
It's impractical. Losses are too large - you'd be better just transmitting power over the conventional power lines.
Electrolysis efficiency is about 10%, so it's out. Hydrocarbons synthesis is fairly efficient (but it is not easy) but then you'll have to burn them and convert to electricity with about %40 efficiency. Oh, and you need to start from a fairly 'concentrated' carbon for hydrocarbon synthesis - you take a piece of carbon, heat it with water and get CH4 (methane) and some CO/CO2 and some other hydrocarbons. And guess what is most often used as a carbon source?
The most efficient way to store energy now is hydro-accumulators: you pump water uphill at night and use it to drive generators during day. But it requires.
Read about 'baseload capacity'. It's not practical to use 100% wind power, because in this case you'll either have to transmit surplus of power over the long distances (inefficient!) during high winds or have energy deficit during low winds.
The most optimal ratio for wind power is about 40%/60% (60% - baseload capacity). In some places with good stable and windy weather it can be up to 60%/40%.
Are you nuts?
First of all, vanadium is NOWHERE close to carbon in abundance. And you need lots of energy to extract it. Calculate how many hundreds of millions (yes, that much) of tonnes of vanadium you'll need. And then calculate a small thing called 'environment footprint' of industry large enough to support vanadium battery infrastructure.
>We need to adjust our direction of civilisation away from more toys and gadgets to higher quality human interactions and more meaningful labour.
>Sorry all you PR saps and admin assistants at hedge funds and nail salon operators. I would recommend you learn something useful, like FARMING. Or dismantling Las Vegas and Phoenix.
So, why are you using a computer and reading Slashdot (wasting untold number of man-hours which can be put to FARMING). Go and start digging (with spade, no complex tools for you!). I grew up on a farm (for several years), and believe me - it's not an easy task at all.
But of course, you eco-nuts can only preach...
Why? Because I live in Russia. Believe me, popular support for Putin here in indisputable. It's REAL. I don't like it at all, but it's real.
To be fair, Putin restored at least some semblance of order. During Yeltsin's last days country was rapidly approaching another collapse (for example, several provinces added to their local constitutions ability to form independent militia), Chechens were openly kidnapping and torturing people (in the best traditions of Iraqi insurgents) and so on.
Not a problem, "Just Russia" and "LDPR" parties are also Putin's lapdogs. Together they have a confident 80% of seats.
Nope, independent exit polls showed pretty much the same result. "Liberal Forces Union" ("Souz Pravih Sil") got 1%, "Jabloko" ("Apple" in Russian :) ) got another 1%.
Ok, maybe they could have got another 1-2% each without voting fraud. But it's still pitifully small.
I still don't think that Russian special services killed Litvinenko. And it's not like anyone believed that the United Russia will not abuse his power...
The most interesting question: why have they done that? I live in Russia and nobody here really doubts that the ruling party ("United Russia" - "Edinaja Rossija") influenced elections.
The real approval rates of other opposition parties (communists excepted) were in single percents, anyway. And the real approval rate of United Russia was high enough - all manipulations possibly resulted in several extra seats in parliament for them. So it's not that Putin seriously risked losing his power.
The best way to authenticate a person is to use physical documents (ID card/passport with photo). Usually, passports have pretty good forgery protections. If you can't be present in person - use notarized documents sent by snail-mail. That should cover 'important' things like getting a mortgage or buying a car.
For less important things (like buying stuff on the Intertubes) the current system works pretty well. The occasional card thefts can be mostly eliminated by things like RSA tokens.
The current practice of authentication using SSN+date_of_birth is insane and should be stopped.
Next step: require credit card payment to buy pre-paid phone.
You can have it, no problem. For example, MMS use general data channels.
But it's much less popular.