On Fri, 26 Jun 2015 13:34:59 -0700 PGNd wrote: > > > On Fri, Jun 26, 2015, at 01:23 PM, RW wrote: > > They shouldn't be trusted unless there is a chain of trust. They > > don't matter anyway since they are from the original relay before > > the email was forwarded. > > I thought that 'chain of trust' was established by their inclusion in > the internal_networks/trusted_networks. Apparently not ... > > What's the correct means/place to establish that chain of trust?
The received headers are parsed top to bottom; once an untrusted server is identified the chain of trust is broken and nothing below that can be trusted. Spammers can and do forge headers. > If they "don't matter anyway" since they're from prior hop, should > they not be ignored, rather than parsed & identified as untrusted? They might be useful, otherwise they are informational.
