Keeping an email address current is not a big deal. There are anonymous remailers that will produce an email address that you can use once. This provides the benefits of contact in the case of problems.
Keeping a *name* current, however, has no justification other than wanting to track down domain owners.
Also, I would expect a domain registrar to consider information in their database confidential, something that they never release or sell.
Of course, then they just find loopholes like merging with other companies that want to use data, like the financial services companies did -- sigh.
If US legal cases had capped damages, companies wouldn't be so hung up on avoiding the slightest hint of liability, willing to lose customers, even, to avoid the faintest trace.
A friend from Norway once told me that the reason that Norway doesn't have the problems with ridiculous worries over liability that the US does isn't that it's harder to prosecute a case in Norway. It's just that multimillion dollar awards are unheard of. You get restitution, but not scads of money above and beyond.
Cut down on the amount of money floating around in the legal system, and you return sanity to the consumer world.
You said that you were worry about proxies -- I have secondhand that GANDI is a good registrar -- prices slightly higer than GoDaddy, but significantly better (as in favoring the user versus the registrar) policies. Not sure if they provide a proxy service.
GAIM is nice, but *boy* would I love to see GPG support being available (and in the base distribution, instead of in a plugin that doesn't always build with the current main GAIM release, as the existing encryption support is).
You are wrong -- this is determining whether or not this is covered by the Consitution.
However, from my reading of just the summary (didn't read the article), the judge is probably still out of line, just because federal courts (Might be US Supreme Court, don't remember) have said that the length of copyright is not an issue.
I personally agree with the judge, but that's irrelevant -- according to the rules of the system, he probably should have ruled that the case has already been ruled on.
(a) By "terrorists", I assume you're talking about al Qaeda. How does al Qaeda gain from the presidential election? Neither Bush nor Kerry is likely to stop hunting them down.
(b) There are lots of groups with more stake in who becomes president and who are better equipped to screw with the election -- either political party, for instance. An activist programmer. A state official involved in the machines. I'm worried about *them* mucking with the election, not with terrorism.
(c) It'd hardly be terrorism to hack a system (producing political influence by inflicting terror on a populace), so from a simple, stupid, logical standpoint, unless someone had already engaged in terrorism, they wouldn't be a terrorist.:-)
Why is this a FOX News issue when all they state the obvious?
Because they're being deliberately misleading. Terrorists "hacking the election" is just not a big concern, but they keep trying to keep terrrorism in people's heads. Terrorism has never been a real top national problem, not on 9/11 and certainly not now. Smoking, car crashes, alcohol -- all of these kill more people and cause vastly more economic damage, and do so on a recurring basis. The only reason people care so much about 9/11 is because of the steady and constant media coverage.
I, for one, would like to hear not at all about Bush and Kerry's war records, little about their stupid "war on terror" initiatives, and more about issues that actually affect American citizens.
Also, it wasn't just that the chimp could be taught to *hack* the voting system. That, apparently, is beneath even a monkey. No, what's impressive is that he then covered up his tracks, removing all trace of his presence.
If it wasn't really obvious that this was a Fox News story, the incessant repetition that terrorists will attack the voting booths in the text should have really driven things home.
I started poking at implementing this a few years back, but at the time there weren't solid and free public satellite data archives.
It is kind of sad that the first implementation of Earth is Windows-only, though it is open source, so hopefully some enterprising hacker will improve the situation.
Remember when Bush was getting voted in against Gore? And everyone said "yeah, Bush is inexperienced and maybe won't be so great, but he'll have great advisors and he's going to listen to them"?
How does that same logic not apply to Kerry? You think he won't have technology advisors? That he personally makes all policy? Just because Kerry's following a technology platform that he personally didn't develop doesn't mean that one should vote against him.
Sure, but if you make the data opaque to the program -- say, "this is array of raw data and this is the length" and "this is your pixel array after I've decoded it", then the content of the data should never matter. All these exploits aren't due to just basic bugs in programs, but due to some sort of specially-crafted content that isn't handled causing problems.
The code that needs to be secured is the logic where operation depends on the content of the data.
It depends on the software. Games and other "disposable" software demonstrate exactly what you say here. Other software has a longer useage. Database servers, for example, can be in production for many years (sure, the version numbers may change and you get newer products, but the patents that lie at the base of the work can be in effect and applicable for many years).
Sure, but the main IP affecting a DBMS is in the optimizer, something that is not a one-off project and can be steadily improved. It's quite feasible to work on a DBMS and keep improving it and keep ahead of the curve.
In order for you to argue that software patents are economically worthwhile, you'd have to demonstrate that there is a (worthwhile) specific process involved in a system that costs so much to produce that without patent protection, without a seventeen year monopoly on distribution of any related system to provide funding to produce this process, it could not be developed. And I just can't think of many features along those lines. Perhaps the mechanisms that we currently use in public key cryptography, but even the RSA folks didn't really do their work with direct profit in mind.
Sure, copyrights mean that someone else can't use your code, but doesn't prevent someone else from doing the same thing as you with different code. So, the "little guy" spends a year writing code and releases it into GPL because he's a nice guy. He intends to make money supporting his project. A big company sees it, throws a bunch of resources on it, develops a product in six months, and offers it as a product including support with more features. "Little guy" sees his revenue dry up and he can no longer have the dream of supporting himself through support contracts on his work.
He can't GPL it, because it has patents on it. Patented code (well, without a public release for all derivative works) is incompatible with the GPL. Most such folks don't open source their code. I'm much happier with someone closed sourcing their code and not patenting it (sure, it'll take a while to get reverse engineered, but someone will manage it eventually).
Sure, you say that this fosters "innovation" and "competition", but what it does is makes software a non-money making process.
It makes it more competitive. Which is usually (though, I will admit, not always) a good thing for the system as a whole.
I could, if I wanted, try to set up shop offering services and support for MySQL and/or Apache and prices less than the folks who wrote and maintain that software. I can fork off their code and take their business away, if I were aggressive enough (includes doing it all for free, subsidizing it off my savings, until I drive them out of the game).
A few people have tried (esp. with Apache), and it hasn't happened.
What they rarely take into account is that they might be on the receiving end of that sometime and will need to feed themselves and their families somehow. It's very easy to take the stance of "survival of the fittest" when you ignore the fact that it will also apply to yourself. A person's ego will rarely let them even think that they will be the one to suffer in such a situtation, they will always be the winner in such contests. For every "winner", there is at least one "loser", many times there are many "losers".
Sure, but it's still necessary, if the economy as a whole really is better off. A lot of people have been badly screwed over by shifts in industry (take, for instance, steelworkers in the United States -- the plight of them and the number of them is much more significant than that of software engineers).
When there was no concept of copyright (and later, for a long time, no concept of international copyright), content producers still made content. The rules change a bit, but the system still works.
Funny, that. The people that I know most rabidly against software patents are those who would, by your way of thinking, stand to gain the most from them -- computer science academics.
Having software patents has benefits, not having them has benefits as well. Both cases benefit the larger companies.
That's logically nonsensical: Given A or ~A, A benefits large companies and ~A benefits large companies.
Without software patents, the instant anybody comes up with an idea, they've generated a new revenue stream for a large company. The large company simply sees and likes the idea so develops a product around it and sells it.
Except that:
(a) Implementation time in software relative to product cycle time is very long, making the value of being the first person to implement something still valuable without the need for an artificial monopoly. Traditional patents were designed for systems where ideas were pretty simple ("use a new sort of gear here") and the product lifecycle was long ("Yes, we've been making this type of plow for seven years now"). This vastly decreases the benefits of patents in the software field.
(b) Nobody uses patents as intended, in a manner that benefits the population as a whole. Large companies just maintain patent portfolios to keep people from entering and cross-license with their competitors. Little incentive to produce better products. Lots of companies have hard caps on what they'll pay for a license due to outrageous software patent litigation. For example, IIRC (and this may be out of date), Intel has a hard limit on a one-time $100K fee per patent, though they are willing to cross-license with other holders. That's not a very conducive environment to protect that little independent researcher that you're thinking of.
Sure, products may become cheaper because not having software patents allows every Tom, Dick, and Harry to have their own versions of the software, but the large companies who can employ a lot of folks can have more resources to develop the application further.
Software patents have absolutely nothing to do with encouraging companies to do research. In a corporate lab environment, real advancements don't get patented, because then all your competitors have cross-licensed access to your research. They remain secret.
Along those lines, we should do away with copyrights on software as well for the same reason. Once someone writes some code, why should it be limited in how you can use it? Why have to have someone rewrite more code to do the same thing? All code written should be released under the BSD license, then.
Of course not. Copyright doesn't have the problems of software patents (you don't have "obvious copyrights"). Copyright does a better job than patents of dealing with the long-implementation short-lifecycle approach of software.
You can't regulate software. A large software project (say, Windows) is vastly more complex than any other sort of engineering project (say, a bridge).
Regulation relies on establishing some set of standard actions that guarantee reliable output. You can't standardize software development -- it just varies too much.
Most of that is due to PR fallout from Enron. Now, you can't blame Bush for Enron (which some have tried to do ), but on the other hand, it's pretty hard to credit him for doing anything special when the public got outraged.
One is setting up a broadband router that prioritizes outbound traffic from each host based on ToS field. Depends on your router how you do this. I hacked up a bunch of scripts myself with a buddy to turn a Linux 2.6 box into a nice broadband router with QoS. I guess you could take a look at the LARTC. I should really post 'em up somewhere. Conceptually, it's not that complicated, but Linux's networking documentation could be better.
The other part is telling the client to use low-priority traffic. That bit's easy.
groupadd mldonkey chgrp mldonkey mlnet (assuming mlnet is your binary name) chmod g+s mlnet iptables -t mangle -A OUTPUT -m owner -p tcp --gid-owner mldonkey -j TOS --set-tos Maximize-Throughput
This will tell the client to send out all packets from mlnet set to IPTOS_MAXIMIZE_THROUGHPUT.
A few clients, like gtk-gnutella, already have support for ToS and don't need to run EGID anything. ToS support is *huge* -- with an appropriate broadband router, it means users can saturate the outbound line without impacting other usage of the line.
You know, it might be worthwhile to write things like libjpeg in safe languages.
Ocaml is pretty fast, but I realize that not everyone wants the runtime. How about cyclone? It's an extended version of C that's backwards compatible with C, but can pick up unsafe errors at compile time -- sounds pretty much like what folks might want.
Set up your broadband router to prioritize regular or ToS MINIMIZE_DELAY packets above MAXIMIZE_THROUGHPUT packets, run mldonkey EGID mldonkey, and set your box to reclassify stuff from EGID mldonkey programs as MAXIMIZE_THROUGHPUT.
You can use your full outbound connection, keep it constantly saturated, and it won't affect web browsing or gaming performance at all.
Slashdot Mirror
His food comparison is BS, sure.
However, the comparison to the $65 set is reasonable.
The universe is deterministic.
:-)
The question is just in how large of a system you define as encompassing the universe.
Keeping an email address current is not a big deal. There are anonymous remailers that will produce an email address that you can use once. This provides the benefits of contact in the case of problems.
Keeping a *name* current, however, has no justification other than wanting to track down domain owners.
Also, I would expect a domain registrar to consider information in their database confidential, something that they never release or sell.
Of course, then they just find loopholes like merging with other companies that want to use data, like the financial services companies did -- sigh.
Of course, the reason why is fear of liability.
If US legal cases had capped damages, companies wouldn't be so hung up on avoiding the slightest hint of liability, willing to lose customers, even, to avoid the faintest trace.
A friend from Norway once told me that the reason that Norway doesn't have the problems with ridiculous worries over liability that the US does isn't that it's harder to prosecute a case in Norway. It's just that multimillion dollar awards are unheard of. You get restitution, but not scads of money above and beyond.
Cut down on the amount of money floating around in the legal system, and you return sanity to the consumer world.
You said that you were worry about proxies -- I have secondhand that GANDI is a good registrar -- prices slightly higer than GoDaddy, but significantly better (as in favoring the user versus the registrar) policies. Not sure if they provide a proxy service.
GAIM is nice, but *boy* would I love to see GPG support being available (and in the base distribution, instead of in a plugin that doesn't always build with the current main GAIM release, as the existing encryption support is).
You are wrong -- this is determining whether or not this is covered by the Consitution.
However, from my reading of just the summary (didn't read the article), the judge is probably still out of line, just because federal courts (Might be US Supreme Court, don't remember) have said that the length of copyright is not an issue.
I personally agree with the judge, but that's irrelevant -- according to the rules of the system, he probably should have ruled that the case has already been ruled on.
We trust election officials in so far as a redundant set of them counting votes for a particular area can be trusted.
One malicious computer programmer on a Diebold system can cause vastly more damage.
I'm not very concerned because
:-)
(a) By "terrorists", I assume you're talking about al Qaeda. How does al Qaeda gain from the presidential election? Neither Bush nor Kerry is likely to stop hunting them down.
(b) There are lots of groups with more stake in who becomes president and who are better equipped to screw with the election -- either political party, for instance. An activist programmer. A state official involved in the machines. I'm worried about *them* mucking with the election, not with terrorism.
(c) It'd hardly be terrorism to hack a system (producing political influence by inflicting terror on a populace), so from a simple, stupid, logical standpoint, unless someone had already engaged in terrorism, they wouldn't be a terrorist.
Why is this a FOX News issue when all they state the obvious?
Because they're being deliberately misleading. Terrorists "hacking the election" is just not a big concern, but they keep trying to keep terrrorism in people's heads. Terrorism has never been a real top national problem, not on 9/11 and certainly not now. Smoking, car crashes, alcohol -- all of these kill more people and cause vastly more economic damage, and do so on a recurring basis. The only reason people care so much about 9/11 is because of the steady and constant media coverage.
I, for one, would like to hear not at all about Bush and Kerry's war records, little about their stupid "war on terror" initiatives, and more about issues that actually affect American citizens.
Also, it wasn't just that the chimp could be taught to *hack* the voting system. That, apparently, is beneath even a monkey. No, what's impressive is that he then covered up his tracks, removing all trace of his presence.
If it wasn't really obvious that this was a Fox News story, the incessant repetition that terrorists will attack the voting booths in the text should have really driven things home.
I like how our government neatly identifies all valuable and potentially damaging targets with large white blocks.
Seriously, what the is a set of satellite photos supposed to do WRT security? That's just idiotic.
No. Might even have been inspired by Snow Crash.
I started poking at implementing this a few years back, but at the time there weren't solid and free public satellite data archives.
It is kind of sad that the first implementation of Earth is Windows-only, though it is open source, so hopefully some enterprising hacker will improve the situation.
Democrats are also notoriously pro-censorship.
And hopefully, after Lieberman scared a bunch of people off and cost them the last election, they learned from this.
Remember when Bush was getting voted in against Gore? And everyone said "yeah, Bush is inexperienced and maybe won't be so great, but he'll have great advisors and he's going to listen to them"?
How does that same logic not apply to Kerry? You think he won't have technology advisors? That he personally makes all policy? Just because Kerry's following a technology platform that he personally didn't develop doesn't mean that one should vote against him.
Sure, but if you make the data opaque to the program -- say, "this is array of raw data and this is the length" and "this is your pixel array after I've decoded it", then the content of the data should never matter. All these exploits aren't due to just basic bugs in programs, but due to some sort of specially-crafted content that isn't handled causing problems.
The code that needs to be secured is the logic where operation depends on the content of the data.
It depends on the software. Games and other "disposable" software demonstrate exactly what you say here. Other software has a longer useage. Database servers, for example, can be in production for many years (sure, the version numbers may change and you get newer products, but the patents that lie at the base of the work can be in effect and applicable for many years).
Sure, but the main IP affecting a DBMS is in the optimizer, something that is not a one-off project and can be steadily improved. It's quite feasible to work on a DBMS and keep improving it and keep ahead of the curve.
In order for you to argue that software patents are economically worthwhile, you'd have to demonstrate that there is a (worthwhile) specific process involved in a system that costs so much to produce that without patent protection, without a seventeen year monopoly on distribution of any related system to provide funding to produce this process, it could not be developed. And I just can't think of many features along those lines. Perhaps the mechanisms that we currently use in public key cryptography, but even the RSA folks didn't really do their work with direct profit in mind.
Sure, copyrights mean that someone else can't use your code, but doesn't prevent someone else from doing the same thing as you with different code. So, the "little guy" spends a year writing code and releases it into GPL because he's a nice guy. He intends to make money supporting his project. A big company sees it, throws a bunch of resources on it, develops a product in six months, and offers it as a product including support with more features. "Little guy" sees his revenue dry up and he can no longer have the dream of supporting himself through support contracts on his work.
He can't GPL it, because it has patents on it. Patented code (well, without a public release for all derivative works) is incompatible with the GPL. Most such folks don't open source their code. I'm much happier with someone closed sourcing their code and not patenting it (sure, it'll take a while to get reverse engineered, but someone will manage it eventually).
Sure, you say that this fosters "innovation" and "competition", but what it does is makes software a non-money making process.
It makes it more competitive. Which is usually (though, I will admit, not always) a good thing for the system as a whole.
I could, if I wanted, try to set up shop offering services and support for MySQL and/or Apache and prices less than the folks who wrote and maintain that software. I can fork off their code and take their business away, if I were aggressive enough (includes doing it all for free, subsidizing it off my savings, until I drive them out of the game).
A few people have tried (esp. with Apache), and it hasn't happened.
What they rarely take into account is that they might be on the receiving end of that sometime and will need to feed themselves and their families somehow. It's very easy to take the stance of "survival of the fittest" when you ignore the fact that it will also apply to yourself. A person's ego will rarely let them even think that they will be the one to suffer in such a situtation, they will always be the winner in such contests. For every "winner", there is at least one "loser", many times there are many "losers".
Sure, but it's still necessary, if the economy as a whole really is better off. A lot of people have been badly screwed over by shifts in industry (take, for instance, steelworkers in the United States -- the plight of them and the number of them is much more significant than that of software engineers).
When there was no concept of copyright (and later, for a long time, no concept of international copyright), content producers still made content. The rules change a bit, but the system still works.
Funny, that. The people that I know most rabidly against software patents are those who would, by your way of thinking, stand to gain the most from them -- computer science academics.
Having software patents has benefits, not having them has benefits as well. Both cases benefit the larger companies.
That's logically nonsensical: Given A or ~A, A benefits large companies and ~A benefits large companies.
Without software patents, the instant anybody comes up with an idea, they've generated a new revenue stream for a large company. The large company simply sees and likes the idea so develops a product around it and sells it.
Except that:
(a) Implementation time in software relative to product cycle time is very long, making the value of being the first person to implement something still valuable without the need for an artificial monopoly. Traditional patents were designed for systems where ideas were pretty simple ("use a new sort of gear here") and the product lifecycle was long ("Yes, we've been making this type of plow for seven years now"). This vastly decreases the benefits of patents in the software field.
(b) Nobody uses patents as intended, in a manner that benefits the population as a whole. Large companies just maintain patent portfolios to keep people from entering and cross-license with their competitors. Little incentive to produce better products. Lots of companies have hard caps on what they'll pay for a license due to outrageous software patent litigation. For example, IIRC (and this may be out of date), Intel has a hard limit on a one-time $100K fee per patent, though they are willing to cross-license with other holders. That's not a very conducive environment to protect that little independent researcher that you're thinking of.
Sure, products may become cheaper because not having software patents allows every Tom, Dick, and Harry to have their own versions of the software, but the large companies who can employ a lot of folks can have more resources to develop the application further.
Software patents have absolutely nothing to do with encouraging companies to do research. In a corporate lab environment, real advancements don't get patented, because then all your competitors have cross-licensed access to your research. They remain secret.
Along those lines, we should do away with copyrights on software as well for the same reason. Once someone writes some code, why should it be limited in how you can use it? Why have to have someone rewrite more code to do the same thing? All code written should be released under the BSD license, then.
Of course not. Copyright doesn't have the problems of software patents (you don't have "obvious copyrights"). Copyright does a better job than patents of dealing with the long-implementation short-lifecycle approach of software.
You can't regulate software. A large software project (say, Windows) is vastly more complex than any other sort of engineering project (say, a bridge).
Regulation relies on establishing some set of standard actions that guarantee reliable output. You can't standardize software development -- it just varies too much.
Most of that is due to PR fallout from Enron. Now, you can't blame Bush for Enron (which some have tried to do ), but on the other hand, it's pretty hard to credit him for doing anything special when the public got outraged.
There's too much economic inefficiency associated with the existence of software patents to allow them to exist.
Well, there's two parts there.
One is setting up a broadband router that prioritizes outbound traffic from each host based on ToS field. Depends on your router how you do this. I hacked up a bunch of scripts myself with a buddy to turn a Linux 2.6 box into a nice broadband router with QoS. I guess you could take a look at the LARTC. I should really post 'em up somewhere. Conceptually, it's not that complicated, but Linux's networking documentation could be better.
The other part is telling the client to use low-priority traffic. That bit's easy.
groupadd mldonkey
chgrp mldonkey mlnet (assuming mlnet is your binary name)
chmod g+s mlnet
iptables -t mangle -A OUTPUT -m owner -p tcp --gid-owner mldonkey -j TOS --set-tos Maximize-Throughput
This will tell the client to send out all packets from mlnet set to IPTOS_MAXIMIZE_THROUGHPUT.
A few clients, like gtk-gnutella, already have support for ToS and don't need to run EGID anything. ToS support is *huge* -- with an appropriate broadband router, it means users can saturate the outbound line without impacting other usage of the line.
You know, it might be worthwhile to write things like libjpeg in safe languages.
Ocaml is pretty fast, but I realize that not everyone wants the runtime. How about cyclone? It's an extended version of C that's backwards compatible with C, but can pick up unsafe errors at compile time -- sounds pretty much like what folks might want.
Set up your broadband router to prioritize regular or ToS MINIMIZE_DELAY packets above MAXIMIZE_THROUGHPUT packets, run mldonkey EGID mldonkey, and set your box to reclassify stuff from EGID mldonkey programs as MAXIMIZE_THROUGHPUT.
You can use your full outbound connection, keep it constantly saturated, and it won't affect web browsing or gaming performance at all.