| 00:00.04 | Maquis | vontrapp: |rmlart |
| 00:00.17 | goozbach | because if somebody has done a "man in the middle attack" you'd be compromised as well. |
| 00:00.26 | TheWoozle | And presumably (2) IP addresses can also be spoofed -- perhaps not as easily, but easily enough that connecting to the same IP is no guarantee either. |
| 00:00.27 | vontrapp | |rmlart <action> forces $who to sit through an 8 hour VB class |
| 00:00.28 | brigham | can do! |
| 00:00.39 | goozbach | TheWoozle: exactly on both points |
| 00:00.50 | vontrapp | |addlart forces <who> to sit through an 8 hour VB class |
| 00:00.51 | brigham | okay then :) |
| 00:00.59 | vontrapp | Maquis: thar ya go |
| 00:01.01 | goozbach | that is the reason for the third party needed |
| 00:01.14 | Maquis | good... i think it may be one of the more brutal larts in the list... |
| 00:01.16 | goozbach | s/needed// |
| 00:01.44 | goozbach | now what you do is have somebody you trust, vouch for example.com's cert |
| 00:02.15 | TheWoozle | Right... well, given the premises, I accept the conclusion. |
| 00:02.44 | goozbach | you can do so by a public key infrastrure. aka I trust Maquis who trusts sjansen who trusts fozz who trusts example.com. |
| 00:03.04 | TheWoozle | I'm still chewing on the premises, though I'm not sure the spoofing scenario had really occurred to me. |
| 00:03.15 | goozbach | or we all defer to a third party( the certificate authority ) |
| 00:03.54 | goozbach | TheWoozle: it isn't very widespread, spoofing a dns entry is a localized problem, but it does occur. |
| 00:04.13 | TheWoozle | And we established that the cert keys are sent by snailmail? Or are they emailed? |
| 00:04.57 | Maquis | gvim a.out |
| 00:05.01 | Maquis | or not |
| 00:05.14 | goozbach | only people who use the compromised nameserver are at risk. but if that nameserver is at an ISP it's a large localized problem. |
| 00:05.21 | goozbach | brb |