QUOTE(Kelly Martin @ Sun 6th September 2009, 5:43pm)
My problem is that you people keep spouting off about the infalliability of checkuser evidence, when in fact a lot of the time the IP evidence is shot through with holes, largely because the people interpreting it have no real idea what they're doing. However, since you refuse to make the evidence available for public review, all we have to go on is your word, which (given Wikipedia's track record) we cannot fairly trust. I don't think you're lying about the two IPs, but I have no reason to trust your conclusion as to one of those IPs being "residential" and the other not, and your comments indicate to me that you have serious deficiencies in your understanding how the Internet works. If you'd care to share those IPs, I could review your conclusion for reasonability. We can do that privately if you prefer.
Here's the thing. The second IP is clearly non-residential with semi-public access, and is clearly labeled as such. Let's assume that the first IP is not, in fact, residential, and is a coffee shop, cafe, library or other small business. Then we have at least 4 editors (172, Cognition, Mrs. Breedlove and Tha-HGlsrqNA) who, within the time span of the checkuser table, edit from the same school/business, and from the same coffee shop/library, and from
nowhere else (no home, no wi-fi, no mobile services). That's a pretty big coincidence if they are unrelated. It also means that on July 6,
172Â
(T-C-L-K-R-D)
spent almost 24 hours at this coffee shop and on July 8, he edited from there at 1 AM local time.
QUOTE(Kelly Martin @ Sun 6th September 2009, 6:09pm)
This is true for most of the major ISPs, but is not true for some smaller community-level ISPs, some of which do not assign public IPs to their customers by default. Instead, they get RFC 1918 private IPs and the ISP translates these onto one or more public IPs owned by the ISP. In this sort of situation, you may end up sharing an IP in real-time with another customer of the ISP.
This used to be more common but most of the smaller ISPs that did this sort of thing have gone under or been bought up by larger ISPs who don't do this. Serving RFC 1918 IPs to end customers is unpopular with customers (for several reasons), and today I'd say that this is likely only to be found in captive-market situations (e.g. nursing homes, dormitories). As far as I know no "large" ISP is doing this.
That's quite informative, thank you. In this case, the ISP is in fact, "large."
This post has been edited by No one of consequence: