That's not a single entity, but thousands of individual copyright holders. The closed branch could only use the code from the developers who accept that, so it is highly improbable that this would ever happen, as even if just a few of them refuse to, major parts of the code would have to be rewritten (which would also be hard, as the code has to exactly fit in some structures and you have to expect some people to go to court to stop the closed source version)
> Basically if you put your eggs in several different baskets the number of eggs you can expect to lose will be greater than the number of eggs you can expect to lose by putting them in a single well designed basket.
Of course, if you put your eggs in several well designed baskets, you expect (in its mathematical sence) to lose the same number of eggs as in the single basket. But the probability of losing all eggs decreases exponentially with the number of baskets you put them in.
I once had to calculate the required length in a competition, and while I don't remember the result, it's pretty long. If you require it to be one long cable, so no very big counterbalance, it has to be several times longer than from the earth's surface to the geostationary orbit (ca. 35000 km). But even if you have a great mass available as a counterbalance, you better put it far beyond the geostationary orbit, otherwise it will have to be very big. And it's a lot easier to make it from one long cable, since it will be hard do make a connection between the cable and the mass because of the extreme forces.
Wrong: With RAID 5, if any two drives fail, all data is lost. Well, all data is not really exact, but the only thing you could hope is that you can still restore some text files.
Your description seems more to fit a RAID 1+0 which is something completely different.
And then, you don't seem to know anything about probabilities:
"If you have 15 drives, and two fail, the chances of them being consecutive are very low."
Correct. But the probability of two consecutive drives failing ist still just as high as with 3 drives! It is just much more probable that from 15 drives two fail at the same time (that's just 2/15) than from three drives (2/3). Still, it could be better to have more drives, just because you could have a better "feeling" of how many drives fail before it comes to the fatal crash.
But for RAID 5, this is irrelevant anyways, because any two drives failing will screw your data. And with 15 drives, the probability for that is much higher (and I would even say that 15 drives is a bit too much for RAID 5, use a RAID level where more than one drive can fail without data loss).
Why are "spam zombies" only clients? What's wrong with dial-up-hosts acting as servers, too?
IMHO, it is not the providers duty to prevent their users from trojans, because that is clearly impossible. You suggest that the providers should block the functions of some trojans, which does not solve the problem of the trojans. The users have to be aware of the problem, they have to ensure that their systems are secure.
I also have to admit that I haven't fully understood what the original poster meant:
Blocking outgoing connections to port 25: Would also block the current "accepted" way of sending e-mail through smarthosts. Very bad idea.
Blocking incoming connetions to port 25: Easily circumvented by setting the trojan to accept the mails on another port.
"ISPs (and any other business that gives a workstation a "real" IP address) need to block egress port 25. Comcast is going to be doing this soon, others should soon follow suit. This plugs the zombies."
I hate it when people like you try to split the internet in to parts, "clients" and "servers". The great thing is that everyone can be both client and server! Let's not change this!
Additionally, this measure achieves virtually nothing. Port numbers can be changed; and opening a connection to port 25 is still the normal way to send e-mail.
As I said, things will fail. So you better know what failed. Otherwise you have to rely on someone else to fix it. Of course, you are not able to fix everything yourself, but having a clue of _what_ went wrong can never be a failure. It's not enough to specifically know what is wrong. There may be more cases which seem to be perfecltly legitimate use but will cause severe damage. If you know what your microwave does (and the physics behind it), you can avoid such cases without being restricted to a very small white list of things that you definitely can do.
I don't know how everything works. I can't check every line of source code on my computer. But to know the basic principles and have some knowladge is enough to locate the source of most errors very quickly.
Maybe. But being technically correct is still important.
And by the way, I do care what is in virtually every product I buy. It's not enough that it "just works" - I know everything can fail, everything has its dangers. So I want to know where they are. I need to know what's in the product, and I hate it if everything is hidden to me and there are so many warning labels that I couldn't use the product if I really took care of all of them.
You may think everyone should be able to use products without understanding them. I want that everyone understands as much as possible of the technology behind them.
People should better understand _why_ not to put their cat into a microwave oven...
First, radiation is independant of low-g. Although they may of course occur together.
But have you thought about that bone "loss" (much more that the bones will grow less from the beginning) is maybe not even a problem under those conditions? That the human body simply adapts to the conditions of its environment? Bones we need on the earth would be overkill on the moon! Same is valid for muscles.
Probably you should. The GPL does not change anything about the ownership of the code. Who coded it does still have all his copyrights. Only, if you add code to GPL projects and publish it you have to release your code under GPL.
It seems you do not really know what "exponential" means. I bet you can map everything that exists today in digital storage with 64 bits. 16 Exabytes. Even your "molecular computer" "just around the corner" (that means some decades I think) will probably not have more.
I don't think 64 bits will be the last word, but 1024 bit is definitely too much. That's about 10300 - the number of particles in the universe we know is about 100, so more than 256 bit should hardly ever be needed. I think there is even a long time to go until we'll see 128 bit computers.
I'm a Linux user, but I do consider trying FreeBSD. The only thing that turns me off is people like you.
Only to tell you that you won't get any Linux user to use BSD by telling them "they don't understand it" (as the article does) or by ranting as you do. (Yes, I know there are Linux zealots, too)
One message per minute is not acceptable. I sometimes do send much more e-mail. It can be legitimate to send mail to more than one person.
For me, there is _nothing_ worse than a legitimate e-mail not coming through - especially if it's without notice. And in my experience, blocklists are not effective against spam, and it was more than one time that I could not send someone a mail because of some blocklist. And what's the point - Spammers have circumvented every anti-spam measure so far. Statistical analysis is the only prevention method that yields acceptable results.
That _is_ worse than spam itself. You may be annoyed by spam, but I find it worse if my mail gets dropped because of some overzealous anti-spam-measures. First time it happened to me was when the SMTP-server of my provider was blocked by some list. So what should I do? Abandon my e-mail address and hope that it won't happen again? Yeah, thanks. And now, it is not even possible to send mail from dialup addresses anymore - it is blocked. What happened to the beauty of the anarchistic system??? Mails not having "ASMTP" in the headers will now get blocked, too... Soon we'll have whitelists? That could stil be acceptable if it yielded any results. But it didn't. As far as I can see, more than 90% of the spam I get would not be blocked by any of those blocklists.
IP blocking is pointless. If you want less spam, use spamassassin or another decent and intelligent filter. I configured it very moderately (9 points to be marked as spam), it blocks 95% of all spam and I have not seen false positives yet.
Slashdot Mirror
Well, in fact, if you define smallness=1/size, there is no problem.
That's not a single entity, but thousands of individual copyright holders. The closed branch could only use the code from the developers who accept that, so it is highly improbable that this would ever happen, as even if just a few of them refuse to, major parts of the code would have to be rewritten (which would also be hard, as the code has to exactly fit in some structures and you have to expect some people to go to court to stop the closed source version)
> Basically if you put your eggs in several different baskets the number of eggs you can expect to lose will be greater than the number of eggs you can expect to lose by putting them in a single well designed basket.
Of course, if you put your eggs in several well designed baskets, you expect (in its mathematical sence) to lose the same number of eggs as in the single basket. But the probability of losing all eggs decreases exponentially with the number of baskets you put them in.
I once had to calculate the required length in a competition, and while I don't remember the result, it's pretty long. If you require it to be one long cable, so no very big counterbalance, it has to be several times longer than from the earth's surface to the geostationary orbit (ca. 35000 km). But even if you have a great mass available as a counterbalance, you better put it far beyond the geostationary orbit, otherwise it will have to be very big. And it's a lot easier to make it from one long cable, since it will be hard do make a connection between the cable and the mass because of the extreme forces.
Wrong. DOSKEY (supplied with all later DOS versions) did. And then, he could have meant having a word.bat in an appropriate place.
Wrong: With RAID 5, if any two drives fail, all data is lost. Well, all data is not really exact, but the only thing you could hope is that you can still restore some text files.
Your description seems more to fit a RAID 1+0 which is something completely different.
And then, you don't seem to know anything about probabilities:
"If you have 15 drives, and two fail, the chances of them being consecutive are very low."
Correct. But the probability of two consecutive drives failing ist still just as high as with 3 drives! It is just much more probable that from 15 drives two fail at the same time (that's just 2/15) than from three drives (2/3). Still, it could be better to have more drives, just because you could have a better "feeling" of how many drives fail before it comes to the fatal crash.
But for RAID 5, this is irrelevant anyways, because any two drives failing will screw your data. And with 15 drives, the probability for that is much higher (and I would even say that 15 drives is a bit too much for RAID 5, use a RAID level where more than one drive can fail without data loss).
Why are "spam zombies" only clients? What's wrong with dial-up-hosts acting as servers, too?
IMHO, it is not the providers duty to prevent their users from trojans, because that is clearly impossible. You suggest that the providers should block the functions of some trojans, which does not solve the problem of the trojans. The users have to be aware of the problem, they have to ensure that their systems are secure.
I also have to admit that I haven't fully understood what the original poster meant:
Blocking outgoing connections to port 25: Would also block the current "accepted" way of sending e-mail through smarthosts. Very bad idea.
Blocking incoming connetions to port 25: Easily circumvented by setting the trojan to accept the mails on another port.
"ISPs (and any other business that gives a workstation a "real" IP address) need to block egress port 25. Comcast is going to be doing this soon, others should soon follow suit. This plugs the zombies."
I hate it when people like you try to split the internet in to parts, "clients" and "servers". The great thing is that everyone can be both client and server! Let's not change this!
Additionally, this measure achieves virtually nothing. Port numbers can be changed; and opening a connection to port 25 is still the normal way to send e-mail.
And it would be impossible with GPL software (unless you get all authors/copyright holders to agree to the new terms...
Of course, if someone says reasonable things like this on /., it won't be modded up. That should be +10 Insightful.
As I said, things will fail. So you better know what failed. Otherwise you have to rely on someone else to fix it.
Of course, you are not able to fix everything yourself, but having a clue of _what_ went wrong can never be a failure. It's not enough to specifically know what is wrong. There may be more cases which seem to be perfecltly legitimate use but will cause severe damage. If you know what your microwave does (and the physics behind it), you can avoid such cases without being restricted to a very small white list of things that you definitely can do.
I don't know how everything works. I can't check every line of source code on my computer. But to know the basic principles and have some knowladge is enough to locate the source of most errors very quickly.
Maybe. But being technically correct is still important.
And by the way, I do care what is in virtually every product I buy. It's not enough that it "just works" - I know everything can fail, everything has its dangers. So I want to know where they are. I need to know what's in the product, and I hate it if everything is hidden to me and there are so many warning labels that I couldn't use the product if I really took care of all of them.
You may think everyone should be able to use products without understanding them. I want that everyone understands as much as possible of the technology behind them.
People should better understand _why_ not to put their cat into a microwave oven...
Using gentoo? Just type
emerge screensaver
Debian?
apt-get install screensaver
Impossible. Win95 will crash after 497 days.
First, radiation is independant of low-g. Although they may of course occur together.
But have you thought about that bone "loss" (much more that the bones will grow less from the beginning) is maybe not even a problem under those conditions? That the human body simply adapts to the conditions of its environment? Bones we need on the earth would be overkill on the moon! Same is valid for muscles.
Wrong. In an inertial system, there is no gravity; Einstein demonstrated it is a pseudo-force.
Probably you should. The GPL does not change anything about the ownership of the code. Who coded it does still have all his copyrights. Only, if you add code to GPL projects and publish it you have to release your code under GPL.
It seems you do not really know what "exponential" means. I bet you can map everything that exists today in digital storage with 64 bits. 16 Exabytes.
Even your "molecular computer" "just around the corner" (that means some decades I think) will probably not have more.
I don't think 64 bits will be the last word, but 1024 bit is definitely too much. That's about 10300 - the number of particles in the universe we know is about 100, so more than 256 bit should hardly ever be needed. I think there is even a long time to go until we'll see 128 bit computers.
do crappy drivers keep you from owning a car?
Yes.
But as I said above, I will try FreeBSD. I only think that people like "him" are bad for BSD...
Only thing that I didn't say anything about Windows users?
I didn't want to imply that the "BSD zealots" would stop me from trying it. I definitely will. (At the moment I use Debian and Gentoo)
I'm a Linux user, but I do consider trying FreeBSD. The only thing that turns me off is people like you.
Only to tell you that you won't get any Linux user to use BSD by telling them "they don't understand it" (as the article does) or by ranting as you do.
(Yes, I know there are Linux zealots, too)
One message per minute is not acceptable. I sometimes do send much more e-mail. It can be legitimate to send mail to more than one person.
For me, there is _nothing_ worse than a legitimate e-mail not coming through - especially if it's without notice. And in my experience, blocklists are not effective against spam, and it was more than one time that I could not send someone a mail because of some blocklist.
And what's the point - Spammers have circumvented every anti-spam measure so far. Statistical analysis is the only prevention method that yields acceptable results.
That _is_ worse than spam itself. You may be annoyed by spam, but I find it worse if my mail gets dropped because of some overzealous anti-spam-measures. First time it happened to me was when the SMTP-server of my provider was blocked by some list. So what should I do? Abandon my e-mail address and hope that it won't happen again? Yeah, thanks. ... Soon we'll have whitelists?
And now, it is not even possible to send mail from dialup addresses anymore - it is blocked. What happened to the beauty of the anarchistic system??? Mails not having "ASMTP" in the headers will now get blocked, too
That could stil be acceptable if it yielded any results. But it didn't. As far as I can see, more than 90% of the spam I get would not be blocked by any of those blocklists.
IP blocking is pointless. If you want less spam, use spamassassin or another decent and intelligent filter. I configured it very moderately (9 points to be marked as spam), it blocks 95% of all spam and I have not seen false positives yet.
Not here: /opt/intel/compiler70/ia32/bin/icc -O3 primzahl.c ./a.out
./a.out
bash-2.05b$
bash-2.05b$ time
real 0m0.004s
user 0m0.004s
sys 0m0.000s
bash-2.05b$ gcc primzahl.c -march=athlon-xp -O3
bash-2.05b$ time
real 0m0.003s
user 0m0.003s
sys 0m0.000s
(Filtering prime numbers up to 1M)