'rm -Rf star' is much preferrable, or '/bin/yes | rm -R star'. Otherwise you would be there all day pressing 'y'. You could always do it the Homer Simspon water bird way...
You address a massive issue. Consumerism, industrialization, modernization, digitalization, all fine ideas. But do they provide any sort of happiness. Fundamentally, a professor of mine once described our desires as the desires of the aristocracy, from a Foucaultian perspective - to own a castle (house), segregate our personal lives from that of those around us, and choose our fate to a tea or tee time. What has happened, perhaps, is that we did not in our individualization ask whether or not the aristocracy was happy.
From a paleolithic perspective, we are certainly anything but in our element - packs animals that do seasonal hunting and foraging and farming. We turn nowadays to television, and in cases internet, for the elements of our physio-psychology in derth. Yet we are supplementing artificial inventions for natural (presumably natural) environments.
One case in point is a family of refugees where I live here in Canada, a man from Berundi and his wife. The man's ex-wife was killed, and some of his brothers, sisters, cousins, were killed in war, he cannot see his children and cannot afford a vehicle and relies upon donations for essentials. Yet he is the happiest person I have met.
"works with any Socketed Athlon/Duron cpu"
You never tried the Athlon 1.4 Ghz T-Bird with the K7S5a.On a message board with, on average, 44 posts per topic, there were 14,000 posts on the Athlon 1.4 + K7S5a. Someone did solve the problem, that being total system instability, by putting a 200 ohm resistor in parallel with something underneath the chip (soldered onto the motherboard), but I wasn't brave enough for this and settled with upgrading to an Athlon XP which works fine. Strangely enough, this issue really only reared itself en masse with revision 4 of the board, which constituted the most shipped by far. Revisions 1-3 were flakey, and oddly enough revision 0 was rock solid, from what I read (so this is hearsay), and I stopped paying attention by the time revision 5 was out.
A point of interest: If all the intellectually affluent people know how to, and indeed do, uninstall spyware, and this margin is not taken into account by the people that are recepients of the spyware data, would this not lead to a sponsoring of a dumber internet by promoting the sites that attract, well, the less technically fortunate?
Suppose HP (who is advertising here right now, by the looks of it) is looking to advertise on the net - if the spyware data they buy shows that Slashdot, for example, is hardly even notable on the top spyware list, would this not be detrimental to Slashdot's (or rather VA's) efforts to make a buck off advertising, and in particular directed advertising? Advertisements that are possibly better directed to Slashdot may go to PC Magazine (for lack of a more appropriate choice) or other "mainstream" service.
Of course, when advertising a car, Slashdot is hardly well-directed advertising and is oft notably a selection of people most fortunate technically, but there is probably a clear area where the technically inclined can find better content on any topic over the internet that spyware would never reveal statistically.
Re:Evolution and the corporate lifeform
on
Monsanto and PCBs
·
· Score: 2
The health and life of humans is inconsequential - so long as there is a consumer. Remember: The optimal pharmacutical consumer is the recurrant necessity consumer, as in Aids, diabetes, etc., etc. The only incentive to actually cure someone arises when a competitor releases a cure where you merely have a recurrant solution.
Extrapolating this a-rights, a-humanism, a-life incentive structure reveals that the death of every human on the planet very likely could be directly a result of laissez faire western economic incentives.
For public/private key encryption, a.NAP file serving node would have to 1. decrypt the.NAP, and 2. encrypt it with the public key. Somewhere in the middle, it must be decrypted. It is possible that this is done stream-wise (meaning that temporary files are not necessary), but nevertheless a binary unencrypted format is necessary - maybe not a playable format, but certainly one that contains all the necessary information. Nevertheless, public key encryption is not tenable given the complexity (computational) of this scenario - it is processor prohibitive.
It is possible that the checksums are hashed, or some such, but that will, as noted, quickly be undone. Theoretically the issue is simple:
To share peer-to-peer requires a communication format. A communication format that is encrypted requires key sharing. Key sharing from a central server can undermine the issue, but extensions to Napster such that keys are shared alongside files or simply publishing of keys (since they need never change, or the files themselves would have to change in which case p2p sharing is again impossible) would undermine the "protection".
Who decides the cost. Will it be a competitive rate, or a monopolistic rate. If the latter, then will it be a cost regulated by the government? Not likely.
How could this possibly be a competitive rate, given that most of the music is given a corporate monopoly through copyright in the first place?
To assume that $15/mo will satisfy a given monopoly would be naive, but then, there is no alternative, is there? Monopolies charge prohibitive prices, to maximize profit by appealing to the margin buyer.
If we go down this road, then $15 per song will be considered lucky, get mass produced cookie-cutter music, and lose the artistic freedoms that inspire dyonisian (for lack of a better word, folk) works.
Multicast itself requires multicasting routers from end to end, or at least, it stops at the first non-compliant router the multicast stream encounters. Not a big deal if the big-backbones feed multicasting through. (I actually do not remember if the major backbones have multicasting; I am certain everything.com would have some notes)
The on-demand streaming problem you present is solved with smart-caching, at least partially. It involves things like SQUID or the commercial counterparts, and the concept is very popular with some major broadband ISP's since internal broadband is still cheaper than external broadband. This is effective for repetitive data, not so much for live data, but then, live data falls under the multicasting category of network reduction. For a.ogg or what-have-you file format that streams the exact same data, all that the caching mechanism need check is that the data has not changed, which is provisioned for in the HTTP protocol.
In the case of interim stream requests, where you want a live stream, but the existing streams are already multicast (and you need a syn/ack), a destination can be added (er, might want to check that, but the notion that it cannot is almost offensive;) - check the RFC) and a unicast stream can preempt the multicast data so the multicast stream is not entirely out of order (and firewalls do not inherently nix them). In most relevent cases, ie. streaming video, the image will just rebuild itself with aggregate data + discrete error correction. Hence you can just plug into an existing stream happily, at least for semi-smart streamed formats.
I'd say it's more like 1/30th will want to kill you because you're doing to much, 1/30th will want to kill you because you haven't done enough, and 28/30ths won't care.
I gathered in brief synopsis on xenogenic (foreign species) transplants that the biggest risk is that of 'dormant' virus DNA, as all the species that were once ravaged by pandemics survived but retained considerable portions of the pandemic virus DNA in their own, and as such it remains dormant. Xenogenic (do not believe that's quite the same as transgenic) transplants would therein be a source of pandemic virus DNA, which could be potentially reanimated. That and Murphy's Law seem should stimulate us to believe that that this is a risky proposition.
Re:This is excellent news
on
KDE 2.2.2
·
· Score: 2
Consider a distribution optimized for a modern processor, such as Mandrake. Or consider using a BSD ports collection to compile it optimized for your machine (The LinuxFromScratch.org page has some good tips on optimized compiling).
I recommend getting and installing OpenBSD or NetBSD for the educational experience, but as for a desktop it's fairly obtuse. But for a learning experience, it's a fabulous way to find out what exactly your OS is doing, particularly if you are uncomfortable with it. FreeBSD is a larger target and frankly is just plain faster than most of the BSD's, but a more cluttered install (but still generally simpler/cleaner than the SystemV systems' design).
For KDE, if you do get a BSD (or Gentoo Linux) you can download and compile everything pretty much transparently, plus you can optimize quite a bit by appealing to a modern processor's optimizations. You, if so inclined, can even get the Intel compiler (which optimizes quite well for Athlon's too, I might add), which has numerous significant gains over GCC, but it does break hefty things like glibc and the kernel. (but so does optimizing glibc and the kernel with gcc).
If what I've read is correct, you are the youngest maintainer for this kernel. Do you have any feelings on a social level, regarding much of the peer review and critism will come from people who are older? (and very possibly set in their ways, and potentially intimidating)
* which course will lead to huge American public backlash (clue: look at polls for desire to bomb)
* which course will produce huge economic turnaround by producing a post-war upturning
* which course will produce even more arms exports that can be used against the US. (The Afghans are using American weapons, after all.)
Just pointing those out. I agree with your points more than anyone, I'd say. I'm quite openly hostile towards people ignorant enough to believe that bombing a nation prone to generating suicidal bombers will make them less of a threat (short of a genocidal solution). But the people making the decisions are not the ones looking at the big picture; they're fighting their own bubble-world discourse.
SSL can provide certificate authentication or key based authentication. If you have a hardware SSL mechanism (or even better, dedicated IPSec box), you can VPN multiple clients relatively easilly. It all depends upon the application at hand. For simply tunneling HTTP, HTTPS tends to be 'good enough', although without a certified key against a secure authority it's subject to Man in the Middle attacks. In the case of hardware SSL boxen, all boxen behind the SSL box can assume secure connections. This is made much simpler with diagrams...
In gist the notions of sessions are subject to scrutiny:
SSL sessions are a result of asymetric random key transfer, the random key being a session key, at the end of which (as decided by the server) it becomes useless and void. Tying this into the web server is not possible in some universes, such as IIS, without expensive people and software. (read: commerce server) Note that there are certain rules pertaining to SSL sessions that make them 'user session' prohibitive, such as timeouts and key regeneration policies.
HTTP sessions are often the direct result of cookie transfers, which is often tunneled in an SSL session. The notion of a session here is somewhat moot, since HTTP is by definition stateless and it's merely a pseudo-state that's maintained. This pseudo-state, unless cross examined against random key of the SSL connection, can be spoofed or hijacked.
An alternative pseudo-session is the passing of the username and password around in hidden form variables. The problem with this is that all subsequent requests from the browser to the server must be of the POST form (or insecure GET forms), and worse, javascript 'spoons' can be used to retrieve and disseminate the password and username to... whereever the spoon'er decides.
The final alternative is certificates, which is a scary one since it gives all control to a central certificate authority. IMHO this authority should be the government since it is essentially the mandate of government to provide this sort of identification to the services of the people (birth certificate, drivers license, etc.). Thus the only 'public' certificate authority should really be the government. A scary proposition if you're American or French right now, but not so bad an idea for the rest of the civilized world. The notion of corporate controlled certifications is much worse.
let me clarify, because I'm bitchy due to a fried athlon, I'll be brief.
cookies provide state. certificates provide state. (hidden form elements also provide state). cookies are not anonymous; useful cookies from banks, microsoft, et al., online stores require you to enter personal data. at one point, a good deal of that personal information was stored in cookes; that is no longer the case since the ns4.x and ie3.x cookie exploits permitting you to access all cookies regardless of their domains. that is no longer the case and cookies now reflect an identity for (1) sessions and (2) identification.
anyway, ranting. the point is that the clear alternative to cookie-session states is certificate based session states (by enabling a random key passed over the asymetric cypher); since certificates are verified against a 3rd person, no MiM or hijacking is possible, if done properly and mathematically sound.
there is a great deal of depth to the cert vs cookie debacle; for one on iis the change from cookie sessions to cert sessions is a single click (as is nt auth, with the lanman2/3 password problem noted), therein requiring virtually no code work.
it's pretty clear that either I didn't write what I wanted to say very well or you didn't understand the gist. perhaps a combination. doesn't matter. it's slashdot.
I'm not so sure. Given that those organizations prone to using cookies are prone to keeping track of your personal information (msft,banks,insurance,advertisers,etc) to profitable ends, perhaps the EU really does understand the problem, and will force corporations to find an alternative solution.
Mind you, with luck, that solution will be free certificates (as opposed to verisign et al. certs), so that cookies are no longer necessary to identify a user. Mind you, certificates will provide another point of failure in the identification schema. What we need is an certified anonymous user with the browser, but I doubt corporations sponsoring certification will go for that.
The inherent problem with certificate idenfification is that most browsers now just send it implicitly, without asking you if you actually want to be identified to this system. (This is similar to NT/lanman hacks that give you the NT password of everyone who connects to your web in a nice, easilly decryptable form.)
The problem of privacy is that it fights against personalization of the internet. Corporations will fight for personalization since personalization provides avenues of revenue and control. Cookies are a method of personalization. Banning them may not be the wosrt thing in the world; certificates could be worse (or much better, if done properly:/ ), or the alternative.
Mind you, banning cookies somewhat stifles all existing infrastructure on the internet and attacks what should be a harmless technology of properties.
To help minimize heat and noise, consider these tips:
Enermax has a good rep for p/s.
Get a 4500-5k RPM CPU fan.
Cut out those grills where the computer case fans go; they cut airflow down by up to 50%.
Use a silver thermal transfer compound on your heat sink (wash the heat sink with acetone, then isopropyl solution, then a silk/lens cloth (no lint) before applying the compound). Use plastic to rub the compound around clockwise and counter-clockwise, then clean off excess compound with the cloth.
Get a video card without a fan, like a Geforce2 MX. Video card fans are small and wear down faster, so you get ball bearing failures more often and faster, so they end up being the noisiest part of the machine.
If you've got the money, get an aluminum case.
Tie back your cabling
Direct airflow towards the CPU.
Slashdot Mirror
What is this "store" you speak of?
Actually, the 570V, analog only connection mind you, is exceptionally good among the 15' LCD monitors. In the lesat, I have no complaints.
'rm -Rf star' is much preferrable, or '/bin/yes | rm -R star'. Otherwise you would be there all day pressing 'y'. You could always do it the Homer Simspon water bird way...
... offtopic and frivilous...)
(I know
You address a massive issue. Consumerism, industrialization, modernization, digitalization, all fine ideas. But do they provide any sort of happiness. Fundamentally, a professor of mine once described our desires as the desires of the aristocracy, from a Foucaultian perspective - to own a castle (house), segregate our personal lives from that of those around us, and choose our fate to a tea or tee time. What has happened, perhaps, is that we did not in our individualization ask whether or not the aristocracy was happy.
From a paleolithic perspective, we are certainly anything but in our element - packs animals that do seasonal hunting and foraging and farming. We turn nowadays to television, and in cases internet, for the elements of our physio-psychology in derth. Yet we are supplementing artificial inventions for natural (presumably natural) environments.
One case in point is a family of refugees where I live here in Canada, a man from Berundi and his wife. The man's ex-wife was killed, and some of his brothers, sisters, cousins, were killed in war, he cannot see his children and cannot afford a vehicle and relies upon donations for essentials. Yet he is the happiest person I have met.
I enjoyed reading your note on the matter.
"works with any Socketed Athlon/Duron cpu"
You never tried the Athlon 1.4 Ghz T-Bird with the K7S5a.On a message board with, on average, 44 posts per topic, there were 14,000 posts on the Athlon 1.4 + K7S5a. Someone did solve the problem, that being total system instability, by putting a 200 ohm resistor in parallel with something underneath the chip (soldered onto the motherboard), but I wasn't brave enough for this and settled with upgrading to an Athlon XP which works fine. Strangely enough, this issue really only reared itself en masse with revision 4 of the board, which constituted the most shipped by far. Revisions 1-3 were flakey, and oddly enough revision 0 was rock solid, from what I read (so this is hearsay), and I stopped paying attention by the time revision 5 was out.
A point of interest: If all the intellectually affluent people know how to, and indeed do, uninstall spyware, and this margin is not taken into account by the people that are recepients of the spyware data, would this not lead to a sponsoring of a dumber internet by promoting the sites that attract, well, the less technically fortunate?
Suppose HP (who is advertising here right now, by the looks of it) is looking to advertise on the net - if the spyware data they buy shows that Slashdot, for example, is hardly even notable on the top spyware list, would this not be detrimental to Slashdot's (or rather VA's) efforts to make a buck off advertising, and in particular directed advertising? Advertisements that are possibly better directed to Slashdot may go to PC Magazine (for lack of a more appropriate choice) or other "mainstream" service.
Of course, when advertising a car, Slashdot is hardly well-directed advertising and is oft notably a selection of people most fortunate technically, but there is probably a clear area where the technically inclined can find better content on any topic over the internet that spyware would never reveal statistically.
The health and life of humans is inconsequential - so long as there is a consumer. Remember: The optimal pharmacutical consumer is the recurrant necessity consumer, as in Aids, diabetes, etc., etc. The only incentive to actually cure someone arises when a competitor releases a cure where you merely have a recurrant solution.
Extrapolating this a-rights, a-humanism, a-life incentive structure reveals that the death of every human on the planet very likely could be directly a result of laissez faire western economic incentives.
For public/private key encryption, a .NAP file serving node would have to 1. decrypt the .NAP, and 2. encrypt it with the public key. Somewhere in the middle, it must be decrypted. It is possible that this is done stream-wise (meaning that temporary files are not necessary), but nevertheless a binary unencrypted format is necessary - maybe not a playable format, but certainly one that contains all the necessary information. Nevertheless, public key encryption is not tenable given the complexity (computational) of this scenario - it is processor prohibitive.
It is possible that the checksums are hashed, or some such, but that will, as noted, quickly be undone. Theoretically the issue is simple:
To share peer-to-peer requires a communication format. A communication format that is encrypted requires key sharing. Key sharing from a central server can undermine the issue, but extensions to Napster such that keys are shared alongside files or simply publishing of keys (since they need never change, or the files themselves would have to change in which case p2p sharing is again impossible) would undermine the "protection".
The question is:
Who decides the cost. Will it be a competitive rate, or a monopolistic rate. If the latter, then will it be a cost regulated by the government? Not likely.
How could this possibly be a competitive rate, given that most of the music is given a corporate monopoly through copyright in the first place?
To assume that $15/mo will satisfy a given monopoly would be naive, but then, there is no alternative, is there? Monopolies charge prohibitive prices, to maximize profit by appealing to the margin buyer.
If we go down this road, then $15 per song will be considered lucky, get mass produced cookie-cutter music, and lose the artistic freedoms that inspire dyonisian (for lack of a better word, folk) works.
Brian
Is this not Adam Smith's (or Robert Nozick's) Invisible Hand explanation quantified?
Wherein social forces invisible to the market act.
Multicast itself requires multicasting routers from end to end, or at least, it stops at the first non-compliant router the multicast stream encounters. Not a big deal if the big-backbones feed multicasting through. (I actually do not remember if the major backbones have multicasting; I am certain everything.com would have some notes)
.ogg or what-have-you file format that streams the exact same data, all that the caching mechanism need check is that the data has not changed, which is provisioned for in the HTTP protocol.
;) - check the RFC) and a unicast stream can preempt the multicast data so the multicast stream is not entirely out of order (and firewalls do not inherently nix them). In most relevent cases, ie. streaming video, the image will just rebuild itself with aggregate data + discrete error correction. Hence you can just plug into an existing stream happily, at least for semi-smart streamed formats.
The on-demand streaming problem you present is solved with smart-caching, at least partially. It involves things like SQUID or the commercial counterparts, and the concept is very popular with some major broadband ISP's since internal broadband is still cheaper than external broadband. This is effective for repetitive data, not so much for live data, but then, live data falls under the multicasting category of network reduction. For a
In the case of interim stream requests, where you want a live stream, but the existing streams are already multicast (and you need a syn/ack), a destination can be added (er, might want to check that, but the notion that it cannot is almost offensive
Brian
Does moderating Slashdot count as work?
I'd say it's more like 1/30th will want to kill you because you're doing to much, 1/30th will want to kill you because you haven't done enough, and 28/30ths won't care.
I gathered in brief synopsis on xenogenic (foreign species) transplants that the biggest risk is that of 'dormant' virus DNA, as all the species that were once ravaged by pandemics survived but retained considerable portions of the pandemic virus DNA in their own, and as such it remains dormant. Xenogenic (do not believe that's quite the same as transgenic) transplants would therein be a source of pandemic virus DNA, which could be potentially reanimated. That and Murphy's Law seem should stimulate us to believe that that this is a risky proposition.
Consider a distribution optimized for a modern processor, such as Mandrake. Or consider using a BSD ports collection to compile it optimized for your machine (The LinuxFromScratch.org page has some good tips on optimized compiling).
I recommend getting and installing OpenBSD or NetBSD for the educational experience, but as for a desktop it's fairly obtuse. But for a learning experience, it's a fabulous way to find out what exactly your OS is doing, particularly if you are uncomfortable with it. FreeBSD is a larger target and frankly is just plain faster than most of the BSD's, but a more cluttered install (but still generally simpler/cleaner than the SystemV systems' design).
For KDE, if you do get a BSD (or Gentoo Linux) you can download and compile everything pretty much transparently, plus you can optimize quite a bit by appealing to a modern processor's optimizations. You, if so inclined, can even get the Intel compiler (which optimizes quite well for Athlon's too, I might add), which has numerous significant gains over GCC, but it does break hefty things like glibc and the kernel. (but so does optimizing glibc and the kernel with gcc).
If what I've read is correct, you are the youngest maintainer for this kernel. Do you have any feelings on a social level, regarding much of the peer review and critism will come from people who are older? (and very possibly set in their ways, and potentially intimidating)
And another nail into the coffin of American civil liberties.
* which course will lead to huge American public backlash (clue: look at polls for desire to bomb)
* which course will produce huge economic turnaround by producing a post-war upturning
* which course will produce even more arms exports that can be used against the US. (The Afghans are using American weapons, after all.)
Just pointing those out. I agree with your points more than anyone, I'd say. I'm quite openly hostile towards people ignorant enough to believe that bombing a nation prone to generating suicidal bombers will make them less of a threat (short of a genocidal solution). But the people making the decisions are not the ones looking at the big picture; they're fighting their own bubble-world discourse.
Good question. :) I don't know, actually. I'm inclined to believe that they have negligable differences.
Try blackbox. You don't get smaller or faster than blackbox.
Does this mean people at Warner Bros can be jailed for years under the new terrorism act?
SSL can provide certificate authentication or key based authentication. If you have a hardware SSL mechanism (or even better, dedicated IPSec box), you can VPN multiple clients relatively easilly. It all depends upon the application at hand. For simply tunneling HTTP, HTTPS tends to be 'good enough', although without a certified key against a secure authority it's subject to Man in the Middle attacks. In the case of hardware SSL boxen, all boxen behind the SSL box can assume secure connections. This is made much simpler with diagrams ...
... whereever the spoon'er decides.
In gist the notions of sessions are subject to scrutiny:
SSL sessions are a result of asymetric random key transfer, the random key being a session key, at the end of which (as decided by the server) it becomes useless and void. Tying this into the web server is not possible in some universes, such as IIS, without expensive people and software. (read: commerce server) Note that there are certain rules pertaining to SSL sessions that make them 'user session' prohibitive, such as timeouts and key regeneration policies.
HTTP sessions are often the direct result of cookie transfers, which is often tunneled in an SSL session. The notion of a session here is somewhat moot, since HTTP is by definition stateless and it's merely a pseudo-state that's maintained. This pseudo-state, unless cross examined against random key of the SSL connection, can be spoofed or hijacked.
An alternative pseudo-session is the passing of the username and password around in hidden form variables. The problem with this is that all subsequent requests from the browser to the server must be of the POST form (or insecure GET forms), and worse, javascript 'spoons' can be used to retrieve and disseminate the password and username to
The final alternative is certificates, which is a scary one since it gives all control to a central certificate authority. IMHO this authority should be the government since it is essentially the mandate of government to provide this sort of identification to the services of the people (birth certificate, drivers license, etc.). Thus the only 'public' certificate authority should really be the government. A scary proposition if you're American or French right now, but not so bad an idea for the rest of the civilized world. The notion of corporate controlled certifications is much worse.
Anyway, that's my speal.
let me clarify, because I'm bitchy due to a fried athlon, I'll be brief.
cookies provide state. certificates provide state. (hidden form elements also provide state). cookies are not anonymous; useful cookies from banks, microsoft, et al., online stores require you to enter personal data. at one point, a good deal of that personal information was stored in cookes; that is no longer the case since the ns4.x and ie3.x cookie exploits permitting you to access all cookies regardless of their domains. that is no longer the case and cookies now reflect an identity for (1) sessions and (2) identification.
anyway, ranting. the point is that the clear alternative to cookie-session states is certificate based session states (by enabling a random key passed over the asymetric cypher); since certificates are verified against a 3rd person, no MiM or hijacking is possible, if done properly and mathematically sound.
there is a great deal of depth to the cert vs cookie debacle; for one on iis the change from cookie sessions to cert sessions is a single click (as is nt auth, with the lanman2/3 password problem noted), therein requiring virtually no code work.
it's pretty clear that either I didn't write what I wanted to say very well or you didn't understand the gist. perhaps a combination. doesn't matter. it's slashdot.
Clearly?
:/ ), or the alternative.
I'm not so sure. Given that those organizations prone to using cookies are prone to keeping track of your personal information (msft,banks,insurance,advertisers,etc) to profitable ends, perhaps the EU really does understand the problem, and will force corporations to find an alternative solution.
Mind you, with luck, that solution will be free certificates (as opposed to verisign et al. certs), so that cookies are no longer necessary to identify a user. Mind you, certificates will provide another point of failure in the identification schema. What we need is an certified anonymous user with the browser, but I doubt corporations sponsoring certification will go for that.
The inherent problem with certificate idenfification is that most browsers now just send it implicitly, without asking you if you actually want to be identified to this system. (This is similar to NT/lanman hacks that give you the NT password of everyone who connects to your web in a nice, easilly decryptable form.)
The problem of privacy is that it fights against personalization of the internet. Corporations will fight for personalization since personalization provides avenues of revenue and control. Cookies are a method of personalization. Banning them may not be the wosrt thing in the world; certificates could be worse (or much better, if done properly
Mind you, banning cookies somewhat stifles all existing infrastructure on the internet and attacks what should be a harmless technology of properties.
To help minimize heat and noise, consider these tips:
Enermax has a good rep for p/s.
Get a 4500-5k RPM CPU fan.
Cut out those grills where the computer case fans go; they cut airflow down by up to 50%.
Use a silver thermal transfer compound on your heat sink (wash the heat sink with acetone, then isopropyl solution, then a silk/lens cloth (no lint) before applying the compound). Use plastic to rub the compound around clockwise and counter-clockwise, then clean off excess compound with the cloth.
Get a video card without a fan, like a Geforce2 MX. Video card fans are small and wear down faster, so you get ball bearing failures more often and faster, so they end up being the noisiest part of the machine.
If you've got the money, get an aluminum case.
Tie back your cabling
Direct airflow towards the CPU.