the clothes you're currently wearing don't fit you well and don't reveal your body outline to the computer/video camera? I noticed the pictures in that story were of rather skinny customers wearing tight-fitting dresses -- not all of us dress like that. What about winter when one may be wearing several layers already?
Heh. Well, I still agree with your comments on FreeBSD, and I'm not being satirical -- it does feel a more well-put-together system, though linux is by no means a bad option. Now lets see where I get modded up to.
Read this post about using the Windows NDIS centrino driver (and other drivers) on FreeBSD, using the "NDISulator" (a.k.a. "Project Evil"). See this post for details on Project Evil. And unlike the linuxant thing, this is free.
Now let's take an email address like ed-slashdot@someplace.com. You still do a DNS lookup for someplace.com
Now, what exactly about commercial or throwaway accounts?
Hm - I misunderstood the scheme, I thought a list of valid addresses was being published, not just valid IPs.
But it seems the only check is that the transmitting IP belongs to the domain. As others have pointed out, it breaks forwarding, but it also does not prevent one user (foo@aol.com) impersonating another on the same domain (bar@aol.com). And a lot of spam is sent from home broadband connections, which may have legitimate use as email servers, but there's nothing to stop using them to send spam with fake earthlink.net addresses.
SPF is broken. It breaks forwarding, unless you want to rewrite the From header at every hop.
That seems to be by design. (Not offering an opinion, merely commenting. Seems to me all these schemes will cause much more pain for the small guys than for the big ones.)
Lots of e-businesses generate unique email addresses for different consumer requests, which can then be thrown away, and individuals and mailing list managers (like ezmlm for subscription confirmations) do this too. It works because often the part of the email address after a + sign (or for qmail, a -) is ignored by the mail delivery agent, but can still be used for filtering/sorting mail by the user. Seems to me any DNS-based email address registry has to be smart enough to deal with it.
I suspect that as the big commercial guys get more and more aggressive in breaking email standards in the name of combating spam, the internet will split into different incompatible email groups: the old-fashioned types (which include many university departments still) who use a text console and a program like pine or elm, and the AOL/Hotmail/Yahoo crowd. To some extent it's already happening: I can barely read some messages sent from MS Outlook, they're formatted so badly, and as a result I'm less likely to reply to them.
Anyway, I hope to start using the FreeBSD kernel soon.
Just curious, is there a reason you don't want to use it with the FreeBSD userland, as provided by the FreeBSD people? That would easily be the most stable way, best-performing, best-documented way to use the FreeBSD kernel. And it runs nearly all Linux software, either natively (with recompiling, which the ports system makes easy) or via binary compatibility.
Perhaps you're not comfortable unless your system has a GNU stuck on to the beginning.
Personally I'd sometimes like the linux kernel with the freebsd userland. FreeBSD's userland is just so much cleaner than any linux system I've come across, and the man pages are outstanding, while unfortunately there is some hardware that is better supported under linux than under freebsd (though it should be said that quite a lot of hardware also works better under freebsd than linux).
After running for less than 24 hours, 2 of them had
experienced kernel panics caused by Bind and Apache crashing!
That makes no sense. If Bind and Apache crash, why should the kernel panic? A kernel panic is typically a bug in the kernel, and a crashed program is a crashed program, it doesn't hurt the kernel particularly.
Not to mention the fact that
the Linux kernel itself lacks any support for any type of journaled
filesystem, memory protection, SMP support,
You mention using gcc 3.1; linux at that time most certainly had journalled filesystems and SMP support, and always had memory protection.
I wonder if the reverse could also be done (a qt engine that uses the gtk engine for its theme) or is gtk more flexible in this regard?
At one time this was indeed possible, and KDE's theme manager had an "import GTK theme" option. It worked well, but that option seems to have disappeared in newer releases. I don't know why: something to do with changes in GTK2 or Qt3 perhaps?
This is for identifying the owner of a passport. How is it worse or more error-prone or more insecure than eyeballing the passport photo and comparing with the guy in front of you? How many people actually look unmistakably like their passport photos? Are you arguing that if your passport photo is stolen, you'll have to get a new face?
A passport contains lots of personal information about you, by its very nature. Typically it's only used to enter a foreign country, or sometimes as identification once you're already inside a foreign country. What's "obtrusive" about including biometrics on it? Face recognition may not be perfect but iris recognition is pretty darn good and these things are better than what we have now. If it makes it harder for crooks to obtain a passport, or if it reduces queues at immigration counters, or if it prevents people being wrongly arrested or interrogated on suspicion of passport forgery, I'm all for it.
Slashdot Mirror
the clothes you're currently wearing don't fit you well and don't reveal your body outline to the computer/video camera? I noticed the pictures in that story were of rather skinny customers wearing tight-fitting dresses -- not all of us dress like that. What about winter when one may be wearing several layers already?
With humans, you not only have to get them there in one piece, you have to bring them back in one piece. It's a whole other problem.
Heh. Well, I still agree with your comments on FreeBSD, and I'm not being satirical -- it does feel a more well-put-together system, though linux is by no means a bad option. Now lets see where I get modded up to.
Is that you Ciaran O'Riordan? Complaining about licensing on Linux? And not calling it GNU/Linux, even?
Read this post about using the Windows NDIS centrino driver (and other drivers) on FreeBSD, using the "NDISulator" (a.k.a. "Project Evil"). See this post for details on Project Evil. And unlike the linuxant thing, this is free.
Now, what exactly about commercial or throwaway accounts?
Hm - I misunderstood the scheme, I thought a list of valid addresses was being published, not just valid IPs.
But it seems the only check is that the transmitting IP belongs to the domain. As others have pointed out, it breaks forwarding, but it also does not prevent one user (foo@aol.com) impersonating another on the same domain (bar@aol.com). And a lot of spam is sent from home broadband connections, which may have legitimate use as email servers, but there's nothing to stop using them to send spam with fake earthlink.net addresses.
That seems to be by design. (Not offering an opinion, merely commenting. Seems to me all these schemes will cause much more pain for the small guys than for the big ones.)
I suspect that as the big commercial guys get more and more aggressive in breaking email standards in the name of combating spam, the internet will split into different incompatible email groups: the old-fashioned types (which include many university departments still) who use a text console and a program like pine or elm, and the AOL/Hotmail/Yahoo crowd. To some extent it's already happening: I can barely read some messages sent from MS Outlook, they're formatted so badly, and as a result I'm less likely to reply to them.
Just curious, is there a reason you don't want to use it with the FreeBSD userland, as provided by the FreeBSD people? That would easily be the most stable way, best-performing, best-documented way to use the FreeBSD kernel. And it runs nearly all Linux software, either natively (with recompiling, which the ports system makes easy) or via binary compatibility.
Perhaps you're not comfortable unless your system has a GNU stuck on to the beginning.
Personally I'd sometimes like the linux kernel with the freebsd userland. FreeBSD's userland is just so much cleaner than any linux system I've come across, and the man pages are outstanding, while unfortunately there is some hardware that is better supported under linux than under freebsd (though it should be said that quite a lot of hardware also works better under freebsd than linux).
Sorry, that was meant to be a reply to this post (which is a troll).
That makes no sense. If Bind and Apache crash, why should the kernel panic? A kernel panic is typically a bug in the kernel, and a crashed program is a crashed program, it doesn't hurt the kernel particularly.
Not to mention the fact that the Linux kernel itself lacks any support for any type of journaled filesystem, memory protection, SMP support,
You mention using gcc 3.1; linux at that time most certainly had journalled filesystems and SMP support, and always had memory protection.
At one time this was indeed possible, and KDE's theme manager had an "import GTK theme" option. It worked well, but that option seems to have disappeared in newer releases. I don't know why: something to do with changes in GTK2 or Qt3 perhaps?
This is for identifying the owner of a passport. How is it worse or more error-prone or more insecure than eyeballing the passport photo and comparing with the guy in front of you? How many people actually look unmistakably like their passport photos? Are you arguing that if your passport photo is stolen, you'll have to get a new face?
A passport contains lots of personal information about you, by its very nature. Typically it's only used to enter a foreign country, or sometimes as identification once you're already inside a foreign country. What's "obtrusive" about including biometrics on it? Face recognition may not be perfect but iris recognition is pretty darn good and these things are better than what we have now. If it makes it harder for crooks to obtain a passport, or if it reduces queues at immigration counters, or if it prevents people being wrongly arrested or interrogated on suspicion of passport forgery, I'm all for it.