On Thursday, August 11, 2011 11:28 PM, Les Mikesell wrote:
That conversation would make sense if there were any spam blockers that cared about the collateral damage to unrelated hosts that happen to be
So, in your experience, there aren't *any*, they all block an entire range?
If so, why is that a valid method for blocking spam?
I haven't done extensive research, but there's not really a good way to do it at all, much less correctly.
Man, this is getting to sound more and more like SPAM-L. Outblaze Ltd, before they sold their message business to IBM, did the right thing. Where net blocks are proven to be entirely spew engines, the whole net block gets blocked, persistent abusive ones get firewalled. Said net block would be released a year later for review in case it had been reassigned.
Single mail servers with spammy domains and clean ones get 'whitelisted' in that the ip is not stuffed in the block rules but the domains are.
in an IP range that they don't like. I don't think you'll find any. And it has always been that way since the start of those businesses.
Yes, 15 years ago. I reiterate: it has been *completely* wrong for about 10 years.
It was always wrong. That doesn't mean it won't happen.
Whether it is wrong depends on the black list maintainer imho. Some black lists are very clear in their criteria. Whole country. eg: China. Don't like that? Don't use it. That's what you want? Good for you.
When a black list starts doing things inconsistently, then maybe you can label them wrong. Maybe the Centos mail admins might want to take another look into manitu.net...