Quoting Axb <axb.li...@gmail.com>:
On 06/10/2014 05:11 PM, Patrick Domack wrote:
There are all kinds of way to use the infomation. I just don't
understand why people are so against it, cause it's not 100% foolproof.
Nobody is against the idea, problem is scalability and trust.
To make domain age usable, the BLs I mentioned make use of it as
well as many other daata points to gain trust that a listing won'
tbite the globe, as well as they can.
Consider certain factors wich *can* contribute to delay in listings
produce a positive hit,for example, mirror lag due to rsync,
negative TTL, etc. as reasosn why you seem to see these domains
being listed after you got the spams.
(If your size/budget permits, datafeeds would probably help a lot)
For a small site doing a few whois lookups/hour it may work, but
what if suddenly an ISP/ASP doing many thousands of msgs/sec would
implement this?
I did consider those factors, and they where not the problem.
I do rsync the data feeds locally, and feeds did not contain the
lookups till hours later.
It wasn't a negative ttl issue, as the ttl is non-existant for these lookups
I fail to understand why you would be doing thousands of whois lookups
per second. You see that many new domain names per second?
Mostly it's the same domain names over and over again, and a few new
ones per day.
Domains don't expire, moved around, and updated a lot, and even if it
did, that isn't really much a concern. To cache this infomation for
atleast 3 years, would be fine, likely even longer.
Also, the point of having a central body do this, would cause the
cached results to be even better, and less lookups needed.
I'm not a huge isp, but I don't seem to be any where near as tiny as
you suggest.
