Anyway, what I meant is that as long as the energy required for D-He3 is sufficiently less than that required for D-D, then you have a chance to have a clean reactor.
I'm not sure I understood correctly, but it seems like D-He3 requires less energy than D-T. If that's the case then it may be possible to have enough energy for the D-He3 reaction to occur, but not enough for the D-D reaction (which probably requires even more than D-T otherwise, it would be used instead).
This "new" method of spreading by scalling all kinds of documents for e-mails makes it look like it might be yet another test for "new ways to spam even more people by being even more annoying".
Hey Mr. ass-backward. Stop the gnome/KDE flamewars and take my comment for what it is: commenting on the philosophical difference between KDE and gnome. Saying Qt is a new or an existing toolkit is just a matter of when you look at it. The main idea remains that the idea of gnome was to to reuse as much stuff as possible (even when it shouldn't have), while KDE wrote much of these "from scratch" and has its stuff "more integrated" (just think about window managers).
If people on slashdot want to be taken seriously they really ought to make use of the freedom they are given and actually use some of the source code we donate.
Development time is reduced because you can re-use existing packages. That was mostly at the beginning, when gnome was using xscreensaver, sawfish/icewm,... As for why it's lagging behind KDE (though some may think that's not the case), I'd say it has to do with the use of C instead of C++ (though that's also part of the philosophy).
I wasn't attacking KDE/Qt or saying it's bloated. As you said, in your last paragraph, I'm mostly talking about development and packages. I think the gnome way makes development simpler (faster given equal resources, compared to a "from the ground up, monolithic development) at the expense of simplicity of use (application development). I think gnome's now a bit better by reducing th enumber of packages, but there are AFAIK still more packages required for gnome than KDE.
Actually, I started using KDE at 1.0beta2. Qt was released under a license that was even more restrictive than the QPL. I later switched to gnome (when it was at 0.13) mostly for esthetic reasons.
Also, what I meant in my original comment is that Qt was designed originally do to what it does today (Qt was "new" at some point). On the other hand, Gtk was taken from Gimp and some people added stuff on top to it (pango,...). I don't want to start a "which is better" debate. Just say that I see more than two competing projects. I see two different development philosophies.
In that sense, yes. But what I was mostly refering to is the philosophy of aggregating stuff. Qt is a big, monolithic thing (not saying is bad), while gnome/gtk is composed of tons of small packages. The advantate of the gnome/gtk approach is reusability, but the cost is increased complexity.
Don't you think it's a bit more complicated than that? There are two completely opposite methodologies. One was to write a new toolkit (Qt, at that time) to do everything, while the other was to reuse everything that was available (gtk from Gimp,...) and plug different things together. In some way, I'd say KDE/Qt is closer to the Windows idea (integrated stuff), while Gnome/Gtk is closer to the unix philosophy (put lots of small packages together). I'm not saying one is better/worse, but KDE and gnome really different in terms of development philosophy.
Note that the GPL is also very clear on another point: You don't have to agree to the GPL since you have not signed it, but if you don't agree to it, nothing allows you distribute the work. In other words, you are free to not accept the GPL. In this case, it's exactly as if the code was released with no license (only the copyright), so you're just guilty of blatant copyright violation. I don't think a judge can force GPL release of code for that offense. However, the fine can be so large that the company will prefer to make an arrangement and release the source.
Your programs just be pretty small or your country's mail service must be pretty cheap. In hear, my 100,000-line-of-code OSS program would likely cost more than that to mail.
You can't force them to release anything that is whole-cloth theirs, but you anything that's a clear derivative of yours, their legal choice is to release the source, or face a judge.
Even with a trial, there's no way the judge can force a company to release its code. Only fine them and prevent them from using the code again.
Sure, that was implied in my post. Yet, a company incorporating GPL code in its main product (and playing it nice) will likely choose to say 1) We won't release our code as GPL 2) We ackledge we've been violating your copyright. At that point, depending on the negotiations: 1) Buys non-GPL license from the author 2) Appologize and remove offending code. 3) Remove code and pay some compensation 4) Go to court and get fined (*and remove code*).
Do NOT tell them they must now releaes all their source code to the public.
Besides, the GPL cannot force them to do that. They always have the choice between 1) comply with the GPL and release their source or 2) plead guilty of copyright violation and pay a fine. For example, you can bet that if MS, Adobe, whatever, is caught with GPL code in a major product, they'll likely choose the second option.
...hey, under the DMCA you can probably send a CaD to anybody you want for anything, without necessarily having cause...
Not only the DMCA. You can send anyone any cease-and-desist about anything. They're just free to ignore it if they like. It has little legal value, a bit like "do this or I'll sue".
That's the new terrorist laws. Suspected terrorists must prove that the are terrorists. If they fail to provide proof, the are jailed for obstruction to justice.
In here (southern Quebec), the real temp (3:00pm) is almost that, with windchill at -39 (-50 forcast for tonight). I think this also happens in Finland, so you get the idea...
Now, removing Saddam from power allows a more humane govt. to be put in place in Iraq, which allows all the ethical and practical benefits for the people of Iraq that you claim we need domestically.
Come on. Except for allowing Al-qaeda to finally get a foot inside Iraq, I don't see what this was accomplished. With all the billions the war cost, you could probably feed all of Africa for a while.
So either you're arguing the US should remain an isolationist nation that ignores world problems, or your arguements contradict each other to some extent.
So the US is "joining the world" (not being isolationist) by fighting a war that the whole world opposed to. To be more precise, the only two countries where more than 50% of the *population* was in favor of the war were the US and Israel (and even then I'm not 100% sure). Sure you might want to blame the big bad peace lobby that pours billions of $ to make the people think they don't need a war...
Now let's talk about isolationim. What's the most important country that hasn't signed Kyoto? What about the "International Court" (don't know the exact Engligh name for that one)? Or maybe the ban on antipersonnel landmines? Or the last negotiations about access to drugs for 3rd world country?
They can craft a highly hammy message and append it to their spam.
Sorry to be replying twice. There maybe another (non-technical) reason why this might not work. The problem is that if they put the spam after some ham (enough to cover the spam), then the person receiving it will see the ham first and discard it before reading the spam part. The main target for the spam is naive people. If such person gets an e-mail from the python mailing list, it'll likely be discarded, so the spammer makes no money even if the e-mail goes through. Now if the spam part is at the beginning, the filters can be adapted to look mainly at the start. Even without that, there might be ways to look for very different scores in different parts of the message or things like that.
It may not be easy, but I don't think the bayesian battle is lost yet.
Actually, I'm not saying bayesian filters are perfect, but I was pointing out that it's probably not *that* easy to get past them.
Well, I sure didn't have any trouble finding ham for my training!
Actually, that's not exactly what I meant. The real "training data" for a spammer would be probability tables (or whatever you call what you train) for many different users. Of course stuff available on the net may help them, but it's probably sub-optimal.
Regarding your test, I didn't get through all your methodology, but to be sure, you'd need to "train" on a very different set of "ham" to be sure. Also, consider that personal mail is not likely to be found online. If you like I can offer my spambayes prob. table if you want to test how your messages go through.
Well, creating a highly hammy message and appending any short spam to it they want ought to work. That's not too expensive.
I think at that point, what might happen is that after a while no word will have a really hammy probability, but there will still be words with very spammy probabilities. Have you tried taking e.g. 10,000 ham messages, creating another 10,000 by appending a (different) spam message to each of them and then training on the 20,000. I'm curious how good/bad a bayesian filter would do.
Even if your filter stop it, it poisons the filter.
I agree that even if a filter manages to adapt, there might be an increase in false positive, simply because there will be no "really hammy" tokens.
But I now realize that I have overestimated them by a significant margin.
Well, it's true that some are still using frontpage:)
Slashdot Mirror
Let's see...
The optimistic interpretation: The pentagon is full of idiots.
Even idiots would have changed their mind after the results.
The pessimistic interpretation: The pentagon is full of corrupt people.
Non-idiot corrupt people wouldn't have asked for a security analysis like that (makes them look bad).
My interpretation: The pentagon is full of corrupt idiots.
I have to agree with this interpretation, since it's the only one that's consistent.
Anyway, what I meant is that as long as the energy required for D-He3 is sufficiently less than that required for D-D, then you have a chance to have a clean reactor.
I'm not sure I understood correctly, but it seems like D-He3 requires less energy than D-T. If that's the case then it may be possible to have enough energy for the D-He3 reaction to occur, but not enough for the D-D reaction (which probably requires even more than D-T otherwise, it would be used instead).
...you could be working for SCO.
This "new" method of spreading by scalling all kinds of documents for e-mails makes it look like it might be yet another test for "new ways to spam even more people by being even more annoying".
Wrong
OK, I'm trying to figure out which of my original statement is wrong? Let's see.
I wasn't attacking KDE/Qt or saying it's bloated.
You think KDE is bloated?
I think the gnome way makes development simpler at the expense of simplicity of use
You think not writing a real gnome window manager made development longer or that gnome is too simple to use?
I think gnome's now a bit better by reducing the number of packages...
Gnome should have kept tons of packages?
there are AFAIK still more packages required for gnome than KDE.
There are more dependencies/packages required for KDE than gnome?
Hey Mr. ass-backward. Stop the gnome/KDE flamewars and take my comment for what it is: commenting on the philosophical difference between KDE and gnome. Saying Qt is a new or an existing toolkit is just a matter of when you look at it. The main idea remains that the idea of gnome was to to reuse as much stuff as possible (even when it shouldn't have), while KDE wrote much of these "from scratch" and has its stuff "more integrated" (just think about window managers).
If people on slashdot want to be taken seriously they really ought to make use of the freedom they are given and actually use some of the source code we donate.
Maybe you want to have a look at the stuff I donated:
Speex
FlowDesigner (previously Overflow
GLPlot
Development time is reduced because you can re-use existing packages. That was mostly at the beginning, when gnome was using xscreensaver, sawfish/icewm, ... As for why it's lagging behind KDE (though some may think that's not the case), I'd say it has to do with the use of C instead of C++ (though that's also part of the philosophy).
I wasn't attacking KDE/Qt or saying it's bloated. As you said, in your last paragraph, I'm mostly talking about development and packages. I think the gnome way makes development simpler (faster given equal resources, compared to a "from the ground up, monolithic development) at the expense of simplicity of use (application development). I think gnome's now a bit better by reducing th enumber of packages, but there are AFAIK still more packages required for gnome than KDE.
Actually, I started using KDE at 1.0beta2. Qt was released under a license that was even more restrictive than the QPL. I later switched to gnome (when it was at 0.13) mostly for esthetic reasons.
...). I don't want to start a "which is better" debate. Just say that I see more than two competing projects. I see two different development philosophies.
Also, what I meant in my original comment is that Qt was designed originally do to what it does today (Qt was "new" at some point). On the other hand, Gtk was taken from Gimp and some people added stuff on top to it (pango,
In that sense, yes. But what I was mostly refering to is the philosophy of aggregating stuff. Qt is a big, monolithic thing (not saying is bad), while gnome/gtk is composed of tons of small packages. The advantate of the gnome/gtk approach is reusability, but the cost is increased complexity.
Don't you think it's a bit more complicated than that? There are two completely opposite methodologies. One was to write a new toolkit (Qt, at that time) to do everything, while the other was to reuse everything that was available (gtk from Gimp, ...) and plug different things together. In some way, I'd say KDE/Qt is closer to the Windows idea (integrated stuff), while Gnome/Gtk is closer to the unix philosophy (put lots of small packages together). I'm not saying one is better/worse, but KDE and gnome really different in terms of development philosophy.
Note that the GPL is also very clear on another point: You don't have to agree to the GPL since you have not signed it, but if you don't agree to it, nothing allows you distribute the work. In other words, you are free to not accept the GPL. In this case, it's exactly as if the code was released with no license (only the copyright), so you're just guilty of blatant copyright violation. I don't think a judge can force GPL release of code for that offense. However, the fine can be so large that the company will prefer to make an arrangement and release the source.
BTW 2a is 37cents
Your programs just be pretty small or your country's mail service must be pretty cheap. In hear, my 100,000-line-of-code OSS program would likely cost more than that to mail.
You can't force them to release anything that is whole-cloth theirs, but you anything that's a clear derivative of yours, their legal choice is to release the source, or face a judge.
Even with a trial, there's no way the judge can force a company to release its code. Only fine them and prevent them from using the code again.
Sure, that was implied in my post. Yet, a company incorporating GPL code in its main product (and playing it nice) will likely choose to say 1) We won't release our code as GPL 2) We ackledge we've been violating your copyright. At that point, depending on the negotiations:
1) Buys non-GPL license from the author
2) Appologize and remove offending code.
3) Remove code and pay some compensation
4) Go to court and get fined (*and remove code*).
Do NOT tell them they must now releaes all their source code to the public.
Besides, the GPL cannot force them to do that. They always have the choice between 1) comply with the GPL and release their source or 2) plead guilty of copyright violation and pay a fine. For example, you can bet that if MS, Adobe, whatever, is caught with GPL code in a major product, they'll likely choose the second option.
...hey, under the DMCA you can probably send a CaD to anybody you want for anything, without necessarily having cause...
Not only the DMCA. You can send anyone any cease-and-desist about anything. They're just free to ignore it if they like. It has little legal value, a bit like "do this or I'll sue".
That's the new terrorist laws. Suspected terrorists must prove that the are terrorists. If they fail to provide proof, the are jailed for obstruction to justice.
In here (southern Quebec), the real temp (3:00pm) is almost that, with windchill at -39 (-50 forcast for tonight). I think this also happens in Finland, so you get the idea...
Whats wrong with Finland?
Weather. Personally I don't mind the cold (unless it's -30 like today), but Linus doesn't seem to like it that much.
Just a question: if all the holes are free to move, how is it different from a liquid?
Now, removing Saddam from power allows a more humane govt. to be put in place in Iraq, which allows all the ethical and practical benefits for the people of Iraq that you claim we need domestically.
Come on. Except for allowing Al-qaeda to finally get a foot inside Iraq, I don't see what this was accomplished. With all the billions the war cost, you could probably feed all of Africa for a while.
So either you're arguing the US should remain an isolationist nation that ignores world problems, or your arguements contradict each other to some extent.
So the US is "joining the world" (not being isolationist) by fighting a war that the whole world opposed to. To be more precise, the only two countries where more than 50% of the *population* was in favor of the war were the US and Israel (and even then I'm not 100% sure). Sure you might want to blame the big bad peace lobby that pours billions of $ to make the people think they don't need a war...
Now let's talk about isolationim. What's the most important country that hasn't signed Kyoto? What about the "International Court" (don't know the exact Engligh name for that one)? Or maybe the ban on antipersonnel landmines? Or the last negotiations about access to drugs for 3rd world country?
They can craft a highly hammy message and append it to their spam.
Sorry to be replying twice. There maybe another (non-technical) reason why this might not work. The problem is that if they put the spam after some ham (enough to cover the spam), then the person receiving it will see the ham first and discard it before reading the spam part. The main target for the spam is naive people. If such person gets an e-mail from the python mailing list, it'll likely be discarded, so the spammer makes no money even if the e-mail goes through. Now if the spam part is at the beginning, the filters can be adapted to look mainly at the start. Even without that, there might be ways to look for very different scores in different parts of the message or things like that.
It may not be easy, but I don't think the bayesian battle is lost yet.
Actually, I'm not saying bayesian filters are perfect, but I was pointing out that it's probably not *that* easy to get past them.
:)
Well, I sure didn't have any trouble finding ham for my training!
Actually, that's not exactly what I meant. The real "training data" for a spammer would be probability tables (or whatever you call what you train) for many different users. Of course stuff available on the net may help them, but it's probably sub-optimal.
Regarding your test, I didn't get through all your methodology, but to be sure, you'd need to "train" on a very different set of "ham" to be sure. Also, consider that personal mail is not likely to be found online. If you like I can offer my spambayes prob. table if you want to test how your messages go through.
Well, creating a highly hammy message and appending any short spam to it they want ought to work. That's not too expensive.
I think at that point, what might happen is that after a while no word will have a really hammy probability, but there will still be words with very spammy probabilities. Have you tried taking e.g. 10,000 ham messages, creating another 10,000 by appending a (different) spam message to each of them and then training on the 20,000. I'm curious how good/bad a bayesian filter would do.
Even if your filter stop it, it poisons the filter.
I agree that even if a filter manages to adapt, there might be an increase in false positive, simply because there will be no "really hammy" tokens.
But I now realize that I have overestimated them by a significant margin.
Well, it's true that some are still using frontpage