The media craziness would be solved if people just applied a simple rule: Don't assign a Torino rating to an object until you have observations covering 1% of the time between now and the first potential collision.
All these level 1 rated objects have been reclassified as level 0 as soon as a couple weeks of data have been obtained; why not wait those couple weeks before publising anything?
Why isn't the passenger area of a plane a Faraday cage yet? Keep any electronic interference on the inside, keep the flight electronics on the outside... problem solved.
As someone who relies upon medicine from Eli Lilly to keep me alive from day to day, I think it's a very good thing that they have background checks in place.
Blocking egress port 25 ought to be standard for all residential ISPs.
No. Monitoring port 25 should be standard for all residential ISPs. Redirect all outgoing connections to a transparent proxy; allow everything through, but keep logs.
When you get a spam or virus complaint, look at the log files, inform the customer, and block egress port TCP/25 for that IP address.
That's why database protection is not the same as copyright protection. Database right is normally more restricted in duration (eg, 15 years in the uk) and in scope.
Your hypothetical database "Names of Professional Writers of Manhattan and their Phone Numbers" would not infringe upon the "Names and Phone Numbers of all Manhattan residents" database; but other people would not be allowed to redistributed your database without your permission.
The log files are extremely useful. They're just not designed for human processing. Similarly, the configuration files aren't very human-friendly, but they are very easy to manipulate via scripts.
Anyway, if qmail's configuration makes you want to shoot yourself... what does *Sendmail's* configuration do?
Re:pity they can't have private namespaces
on
FreeBSD Jails
·
· Score: 3, Informative
DragonFlyBSD is supposed to be getting something like this; each process only sees its own version of shared libraries.
A bit over half the world's domain names are hosted on Apache servers. If you look at big targets (companies running https, for example), there isn't much difference between Apache and IIS.
This is, however, rather irrelevant to the question of worms; most of the machines hit by Code Red had IIS running (and weren't patched, of course) but weren't actually hosting any web sites.
Yes, and there are bugs which were in Sendmail for over a decade before they were discovered. Ditto for BIND. And BSD. And it would almost certainly be the same for linux, if linux were old enough.
My employer (who keeps up with security patches) was only halfway through the desktop update cycle.
For some value of "keeps up with security patches" meaning "is halfway through applying security patches which were released four weeks ago".
people using Windows are just about as insensitive to their peers as people who, say, smoke
No. People who don't apply security patches are about that insensitive. There are a lot of mismanaged Windows machines in the world; there are also a lot of mismanaged linux and BSD machines.
We see Windows worms because that's a big target; but let's not delude ourselves into thinking that our favourite operating systems are immune.
Don't forget that all this stuff called "air" gets in the way at times. Once you're 100 km above the surface of the earth, speeding up is just a matter of pushing yourself forward; starting from the surface, you need to worry about pushing all that air out of the way.
I don't know any sort of exact figures, but I'm sure the ratio is much less than 25:1 when you consider the energy lost to air resistance.
Why on earth arent most drivers included with windows anyway?
They aren't? I guess I'd better go and find some drivers then, because when I installed Windows 2000 I just accepted the drivers Windows already had.
Now, Windows does have the minor problem of not coming with drivers for hardware which hasn't been released yet; but I think you'll find the same problem with any operating system. (And with Windows, you can usually download drivers from Windows Update -- no need to search through piles of CDs.)
By my count, they'll have an R_peak of 8800 GFLOPS; unless they've got more efficient linpack code than anyone else, that will put them around 7th or 8th place.
Yes, you can *rebuild* the system without those; but that's really the point of packagizing code. If sendmail, bind, and other similar software was taken out of base and put into ports, people would be able to add or remove them as distinct packages. Right now, if you've installed a RELEASE and you decide you don't want sendmail around, you need to work out for yourself which files to delete.
The ports tree is good. Better than the base system; stuff in the ports tree is split into nice self-contained packages, while the base system is a single monolithic mess.
We need to hack parts of the base system off and put them into ports (like kerberos), not add more stuff into the base system.
Her name is Ceren Ercen; last I heard, she doesn't wear latex any more; and since this is a front page story, I'm not going to provide any links to pictures, but they can be easily found via google.
Slashdot Mirror
The media craziness would be solved if people just applied a simple rule: Don't assign a Torino rating to an object until you have observations covering 1% of the time between now and the first potential collision.
All these level 1 rated objects have been reclassified as level 0 as soon as a couple weeks of data have been obtained; why not wait those couple weeks before publising anything?
Why isn't the passenger area of a plane a Faraday cage yet? Keep any electronic interference on the inside, keep the flight electronics on the outside... problem solved.
If you didn't have insurance or medicaid, you'd be broke!
ITYM "If I didn't live in the UK, I'd be broke".
As someone who relies upon medicine from Eli Lilly to keep me alive from day to day, I think it's a very good thing that they have background checks in place.
Blocking egress port 25 ought to be standard for all residential ISPs.
No. Monitoring port 25 should be standard for all residential ISPs. Redirect all outgoing connections to a transparent proxy; allow everything through, but keep logs.
When you get a spam or virus complaint, look at the log files, inform the customer, and block egress port TCP/25 for that IP address.
The technical sessions start on September 10th, so there's still lots of time to register and book plane tickets.
Speaking of the 10th... I'd better start working on my slides.
tell me what moving part a mouse has? The photons from its LED? or the electrons in its ICs?
I was thinking of the mouse buttons, personally. Apple may still opt for one-button mice, but I haven't seen any zero-button mice recently.
it becomes possible to make a current model PC with zero moving parts perfect for terminals, employee workstations
Yep. Those keyboards and mice with zero moving parts are really nice.
Ok, I guess they could all use button-less touchpads, but I've yet to see a usable keyboard without any moving parts.
That's why database protection is not the same as copyright protection. Database right is normally more restricted in duration (eg, 15 years in the uk) and in scope.
Your hypothetical database "Names of Professional Writers of Manhattan and their Phone Numbers" would not infringe upon the "Names and Phone Numbers of all Manhattan residents" database; but other people would not be allowed to redistributed your database without your permission.
The log files are extremely useful. They're just not designed for human processing. Similarly, the configuration files aren't very human-friendly, but they are very easy to manipulate via scripts.
Anyway, if qmail's configuration makes you want to shoot yourself... what does *Sendmail's* configuration do?
DragonFlyBSD is supposed to be getting something like this; each process only sees its own version of shared libraries.
If it's broken - it's yours!
Oh, so *that's* why SCO isn't suing Microsoft.
A bit over half the world's domain names are hosted on Apache servers. If you look at big targets (companies running https, for example), there isn't much difference between Apache and IIS.
This is, however, rather irrelevant to the question of worms; most of the machines hit by Code Red had IIS running (and weren't patched, of course) but weren't actually hosting any web sites.
This bug has been in Windows for over a decade
Yes, and there are bugs which were in Sendmail for over a decade before they were discovered. Ditto for BIND. And BSD. And it would almost certainly be the same for linux, if linux were old enough.
My employer (who keeps up with security patches) was only halfway through the desktop update cycle.
For some value of "keeps up with security patches" meaning "is halfway through applying security patches which were released four weeks ago".
people using Windows are just about as insensitive to their peers as people who, say, smoke
No. People who don't apply security patches are about that insensitive. There are a lot of mismanaged Windows machines in the world; there are also a lot of mismanaged linux and BSD machines.
We see Windows worms because that's a big target; but let's not delude ourselves into thinking that our favourite operating systems are immune.
Once you're above, oh, say, 40,000ft or so (IE, and minute or two after launch) you're above 99% of the air.
Yes, and how much fuel do you burn in that first minute?
Don't forget that all this stuff called "air" gets in the way at times. Once you're 100 km above the surface of the earth, speeding up is just a matter of pushing yourself forward; starting from the surface, you need to worry about pushing all that air out of the way.
I don't know any sort of exact figures, but I'm sure the ratio is much less than 25:1 when you consider the energy lost to air resistance.
Why on earth arent most drivers included with windows anyway?
They aren't? I guess I'd better go and find some drivers then, because when I installed Windows 2000 I just accepted the drivers Windows already had.
Now, Windows does have the minor problem of not coming with drivers for hardware which hasn't been released yet; but I think you'll find the same problem with any operating system. (And with Windows, you can usually download drivers from Windows Update -- no need to search through piles of CDs.)
My guess is that Apple gave them a discount in exchange for the marketing stunt.
By my count, they'll have an R_peak of 8800 GFLOPS; unless they've got more efficient linpack code than anyone else, that will put them around 7th or 8th place.
Yes, you can *rebuild* the system without those; but that's really the point of packagizing code. If sendmail, bind, and other similar software was taken out of base and put into ports, people would be able to add or remove them as distinct packages. Right now, if you've installed a RELEASE and you decide you don't want sendmail around, you need to work out for yourself which files to delete.
The ports tree is good. Better than the base system; stuff in the ports tree is split into nice self-contained packages, while the base system is a single monolithic mess.
We need to hack parts of the base system off and put them into ports (like kerberos), not add more stuff into the base system.
You say there's generally less than 128MB of data which needs to be backed up; how much of that is new each day?
If most of that data isn't changing, you could use those antique things called phone lines to transmit the differences.
Will the authors of Sobig.G get it right next time?
Time for the BSD Babe post again
Her name is Ceren Ercen; last I heard, she doesn't wear latex any more; and since this is a front page story, I'm not going to provide any links to pictures, but they can be easily found via google.