Actually, iirc there's some rfc that states no part of a domain (not toplevel, not second level, not 3rd level etc) may start with a number. Not that anybody sticks to it ofcourse...
I think you're confusing it with multicast, which appears to do some promiscoid stuff... However, If you could give some evidence of your claim... (source linenumbers would be nice).
I've never heard of 'mac translation', but there's no performance loss when using a mac address other then the one in your cards rom. At initialisation time, the driver basicly loads the mac out of the rom (from the networkcard) and gives it to the little controler-chippy thing on the network card. It can just as easily give it another mac, the procedure is exactly the same. Try:
ifconfig iface ip netmask netmask hw ether mac-addr
This is for linux, I'm not sure freebsd's ifconfig supports setting the mac, it doesn't appear to know the hw param.
What you might be thinking of is a not-so-subtle hack to pretend to have multiple nic's on a single network, where you put the nic in promisc mode, and then do the filtering of incoming packets in software. This is quite a bit slower than doing it in the hardware on the nic, but has the nice sideeffect that you can have as many mac's (and thus give them each an ip, and thus have them appear to be different interfaces) as you want.
Unix OOM handling could be better too you know.. there are several cases (linux, to name one), where it's: OOM->Kill a random process (Well, not really random, but close enough). Ofcourse you can ulimit, but it's still not nice.
I think he was referring to the recent outlook vulnerability which had a bufferoverflow in the date parsing code, which would run on the mail as it came in.
The DNS (which is over UDP) is only transmitting IP packets, and it doesn't matter if some get lost. But in those IP packets (that are going over DNSUDPIP), there are tcp headers. And they give the computers on both ends the chance to make the transmission reliable. So the original poster is right, it doesn't have to be reliable...
Atmosphere doesn't matter. To be safe you have to be so high that the angular momentum (is that the right word?) of the earths rotation will keep you in orbit. And that's fscking high. And you don't want to be further from earth either, since you'd fly off into space. (entertaining, but ultimately annoying)
Very simple. I'll publish (or claim to publish) my private key on a site somewhere for ~1 sec. That way, it won't be my private key but a private key known to me and possiblly a number of others... Anyway, just because my email addy is in the key name, doesn't mean I automatically want that key to be legally binding. I haven't got much confidence in the legal system in my country (.nl), but I'm pretty sure they haven't got their heads _that_ far up their asses.
slightly OT passphrase minirant
on
GPG vs. PGP?
·
· Score: 2
Am I the only one who thinks passphrases have only slowed down pgp/gpg acceptance? Without passphrases, decryption could be completely handled by the MUA without user intervention, thus making the encryption totally transparant. With passphrases it's a royal pain in the ass to have to retype your (secure, and thus long/using weird chars) password every time you restart your MUA. (Which I do a lot, I just kill it and relaunch if I need it. mutt all the way.). You may say 'But that would mean if someone cracks my box they've got my key!'. YES. tough luck. You won't detect a sufficiently sophisticated attacker anyway, so he might as well just snag your passphrase as you type it in next time you use your key.
I think I'm going to remove the passphrase from my key now, unless anyone has any good reasons why not...
What does the site not being accessable have to do with safety? Think of it as losing your wallet behind the couch. Your money isn't in danger, you just can't get to it for a while. Just because the site is down doesn't mean someone has access they shouldn't have.
Can't look it up atm as I somehow I can't reach www.e-gold.com:-(, but there are no storage fees, and the transfer fees are suprisingly small. All costs are listed in their faq. If you can still reach the site (it might just be my #%^*#$^%@ ISP), there's a link to the faq on their frontpage.
Slashdot Mirror
Actually, iirc there's some rfc that states no part of a domain (not toplevel, not second level, not 3rd level etc) may start with a number. Not that anybody sticks to it ofcourse...
Do you have any idea how much it costs to get a patent approved? Sufficing to say it's cheaper to dispute the patent...
I think you're confusing it with multicast, which appears to do some promiscoid stuff... However, If you could give some evidence of your claim... (source linenumbers would be nice).
Actually, if you want to describe the area I think you'd be better off with 'rijndal'...
I've never heard of 'mac translation', but there's no performance loss when using a mac address other then the one in your cards rom. At initialisation time, the driver basicly loads the mac out of the rom (from the networkcard) and gives it to the little controler-chippy thing on the network card. It can just as easily give it another mac, the procedure is exactly the same. Try:
ifconfig iface ip netmask netmask hw ether mac-addr
This is for linux, I'm not sure freebsd's ifconfig supports setting the mac, it doesn't appear to know the hw param.
What you might be thinking of is a not-so-subtle hack to pretend to have multiple nic's on a single network, where you put the nic in promisc mode, and then do the filtering of incoming packets in software. This is quite a bit slower than doing it in the hardware on the nic, but has the nice sideeffect that you can have as many mac's (and thus give them each an ip, and thus have them appear to be different interfaces) as you want.
Me 4.
Btw, has what happened to the 'unpartial observer' that was promised? I hardly think elections.com qualifies, them being paid by icann and all...
bofh lives on at the register. They also have a link to his old site somewhere on the site.
Your standard CD-ROM? Wouldn't you need a DVD drive?
Unix OOM handling could be better too you know.. there are several cases (linux, to name one), where it's: OOM->Kill a random process (Well, not really random, but close enough). Ofcourse you can ulimit, but it's still not nice.
Actually, they'll prolly tell you "C can warship knowledge..."
I think he was referring to the recent outlook vulnerability which had a bufferoverflow in the date parsing code, which would run on the mail as it came in.
rot13'd for spoilerness, and I'm not 100% sure it's correct, It's been a long time...
Whfg jnvg n srj gheaf
Of course not, it's obviously a regexp. So it'd be fckedcompany.com, ffckedcompany.com, fffckedcompany.com etc.
Nice hack, sadly most newer netscapes won't allow access to port 79 :-(.
The DNS (which is over UDP) is only transmitting IP packets, and it doesn't matter if some get lost. But in those IP packets (that are going over DNSUDPIP), there are tcp headers. And they give the computers on both ends the chance to make the transmission reliable. So the original poster is right, it doesn't have to be reliable...
s/to/too/
search for http tunnel on freshmeat, can't be to hard...
Atmosphere doesn't matter. To be safe you have to be so high that the angular momentum (is that the right word?) of the earths rotation will keep you in orbit. And that's fscking high. And you don't want to be further from earth either, since you'd fly off into space. (entertaining, but ultimately annoying)
Very simple. I'll publish (or claim to publish) my private key on a site somewhere for ~1 sec. That way, it won't be my private key but a private key known to me and possiblly a number of others... Anyway, just because my email addy is in the key name, doesn't mean I automatically want that key to be legally binding. I haven't got much confidence in the legal system in my country (.nl), but I'm pretty sure they haven't got their heads _that_ far up their asses.
Am I the only one who thinks passphrases have only slowed down pgp/gpg acceptance? Without passphrases, decryption could be completely handled by the MUA without user intervention, thus making the encryption totally transparant. With passphrases it's a royal pain in the ass to have to retype your (secure, and thus long/using weird chars) password every time you restart your MUA. (Which I do a lot, I just kill it and relaunch if I need it. mutt all the way.). You may say 'But that would mean if someone cracks my box they've got my key!'. YES. tough luck. You won't detect a sufficiently sophisticated attacker anyway, so he might as well just snag your passphrase as you type it in next time you use your key.
I think I'm going to remove the passphrase from my key now, unless anyone has any good reasons why not...
YIKES! I have a daughter?!?!?
Oh, n/m.
Hmm, my bad. from some other comments posted here there is a storage cost, but it can't be much or I'd have remembered ;-).
What does the site not being accessable have to do with safety? Think of it as losing your wallet behind the couch. Your money isn't in danger, you just can't get to it for a while. Just because the site is down doesn't mean someone has access they shouldn't have.
Can't look it up atm as I somehow I can't reach www.e-gold.com :-(, but there are no storage fees, and the transfer fees are suprisingly small. All costs are listed in their faq. If you can still reach the site (it might just be my #%^*#$^%@ ISP), there's a link to the faq on their frontpage.
e-gold is still alive and kicking. (There's a refer(r)er-id in that link, if you don't want it chop it off. You have been warned.
But on the subject: e-gold is managing fine, and at least they don't have problems with people not from the US.